Compare commits

..

138 Commits

Author SHA1 Message Date
Viktor Liu
3df2b170ba Merge branch 'main' into peer-acl-multi-source
# Conflicts:
#	client/firewall/nftables/router_linux.go
#	client/firewall/uspfilter/filter.go
2026-07-31 22:07:04 +02:00
Misha Bragin
0780a806f2 [management, proxy] Management-owned LLM pricing: file-backed defaults + (#6965) 2026-07-31 20:52:56 +02:00
Viktor Liu
feecb993f4 [client] Restrict debug bundle log path and upload destinations (#6975) 2026-07-31 19:12:56 +02:00
Maycon Santos
f51fadf8d4 [misc] update contributing guide (#7009)
## Describe your changes
We are updating the contributing guide to require an issue to be opened
before PRs; this will allow discussing changes before code is shipped.

Update the rest of the file because of outdated information and align
the pull request template
## Issue ticket number and link
2026-07-31 17:15:39 +02:00
Pascal Fischer
aad2702a14 [proxy] remove cluster tag from proxy metrics (#6985) 2026-07-31 16:56:41 +02:00
Ben
7516aa6473 [client] Fix expression order in legacy nftables route rules (#7011) 2026-07-31 16:28:42 +02:00
Viktor Liu
aed60a2432 [client] Fix daemon lock order inversion between SetConfig and login (#6978) 2026-07-31 15:53:58 +02:00
Maycon Santos
234abd7a08 [misc] group x package updates and run weekly (#7000) 2026-07-31 05:34:58 +02:00
Maycon Santos
c1f0006012 [misc] Update SECURITY.md (#6981)
## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [x] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Documentation**
* Expanded the security vulnerability reporting policy with private
reporting options and guidance for hosted infrastructure issues.
* Added recommendations for report contents, acknowledgements, severity
assessment, remediation, advisories, and reporter credit.
* Clarified supported versions, advisory distribution, bug bounty
status, and where to report non-security issues.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-30 15:34:08 +02:00
Viktor Liu
cff49237b6 [client] Stop and remove the daemon on netbird-ui cask uninstall (#6977) 2026-07-30 13:11:27 +02:00
Zoltan Papp
1bf54ddd8f [client] Support Andorid session expiry handling (#6945)
## Describe your changes

Adds the session surface the Android client was missing: read the status
label and session deadline, receive the engine's expiry warnings, extend
the session via SSO without dropping the tunnel (cancellable), and
dismiss a warning.

Status() latches NeedsLogin so an engine restart doesn't erase it.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6945"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787842170&installation_model_id=427504&pr_number=6945&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6945&signature=c3279ac9ce68e376c23320aa7c2a317ce92377595c433e72925bc0264ef372a5"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
  * Added Android session status and session expiration details.
* Added listener support for wake-state changes and session-expiry
warnings.
* Added interactive authentication session extension with async
login-success handling, error reporting, warning dismissal, and
cancellation of an in-progress extension.
* **Bug Fixes**
* Improved “login required” state handling after successful sign-in to
prevent stale login-needed prompts.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-30 10:27:52 +02:00
Viktor Liu
0f5d2d91fb [client] Authorize daemon IPC callers by their local identity (#6967) 2026-07-29 20:32:26 +02:00
Pascal Fischer
0b7e6a9f46 [signal] make pprof configurable (#6963) 2026-07-29 16:54:01 +02:00
Viktor Liu
f2c1070f95 [client] Fix UI crash on Windows builds without dark-mode support (#6958) 2026-07-29 13:52:18 +02:00
Brandon Hopkins
df39c2b254 [infrastructure] Deprecate legacy Dex and Zitadel getting-started scripts (#6952)
## Describe your changes

Retire the legacy Dex and Zitadel installation scripts by replacing
their implementations with compatibility notices that:

- Exit unsuccessfully before making any system changes.
- Direct new deployments to `getting-started.sh` and the self-hosting
quickstart.
- Explain that the current installer uses NetBird's embedded Dex-based
identity provider.
- Direct users to configure Zitadel through the NetBird Dashboard or
follow the advanced guide for a standalone identity-provider deployment.
- Clarify that Dex support, Zitadel support, and existing deployments
are not deprecated.
- Announce that the compatibility notices will be removed in NetBird
v0.80.

Replace the legacy Zitadel provisioning workflow with CI checks that
verify both scripts fail, write their notices to stderr, and include the
expected documentation links and removal version.

## Issue ticket number and link

No single issue tracks this deprecation.

Related to:

- Legacy installer failures: #4980, #4698, #3566, #3010, #2834, #2094,
  #2046, and #1966.
- Legacy installer enhancement requests: #4510 and #1785.

## Stack

Standalone PR based on `main`.

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change because the
compatibility notices link to the existing quickstart,
identity-provider, and advanced deployment documentation.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

N/A

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6952"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787867848&installation_model_id=427504&pr_number=6952&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6952&signature=0dcc8cda6a3daa5647fa0d40a0f143fef9ad6258aa9ba07a861e658efd244ad1"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **Breaking Changes**
- Retired the legacy Dex and Zitadel getting-started installation
scripts.
- These scripts now exit with an explanatory deprecation message and
direct users to the current setup flow and documentation.
- Existing workflows now verify that the retired scripts fail as
expected and provide the appropriate migration guidance.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-29 12:20:43 +02:00
Pascal Fischer
dd2bdc0de3 [management] explicit accountID check when deleting a user (#6944) 2026-07-28 18:34:55 +02:00
Zoltan Papp
44fef45c2f [client] Export peer details for Android (#6925)
## Describe your changes

Eexport peer details for Android

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6925"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787768377&installation_model_id=427504&pr_number=6925&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6925&signature=b708ba5d6f9b45ae48edf9597aad8aecac9a93c75b2a5a033e14c1c945c47f1f"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **New Features**
* Enhanced Android peer details with latency, data transfer totals,
connection timestamps, handshake information, relay and security status,
and ICE connection details.
* Added formatted latency values to support clearer connection-quality
indicators.
* **Bug Fixes**
* Peer lists now refresh WireGuard statistics before displaying transfer
and handshake information, ensuring current values are shown.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 18:13:22 +02:00
Riccardo Manfrin
3d1f209ea3 [client] parse NB_LAZY_CONN_INACTIVITY_THRESHOLD as a Go duration (#6947)
## Describe your changes

`NB_LAZY_CONN_INACTIVITY_THRESHOLD` was parsed with `strconv.Atoi`, i.e.
as a bare
integer number of minutes. The documentation, however, states it takes a
Go duration
(e.g. `30m`, `1h`). As a result any documented value such as `30m` or
`5m` failed to
parse and **silently fell back to the 15m default**, so the setting
appeared to have
no effect.

`inactivityThresholdEnv()` now parses the value with
`time.ParseDuration`, matching
the docs. A bare integer is still accepted as a number of minutes for
backwards
compatibility, and an unparseable value logs a warning and falls back to
the default.

Added `TestInactivityThresholdEnv` covering Go-duration values
(`30m`/`1h`/`90s`), the
bare-integer minutes fallback, and zero/negative/garbage inputs.

## Issue ticket number and link

N/A
2026-07-28 18:12:49 +02:00
Zoltan Papp
2ef457be95 [client] Unify route selection in the route manager (#6928)
## Describe your changes

Move route select/deselect handling from the daemon server into exported
routemanager methods (SelectRoutes, DeselectRoutes, SelectAllRoutes,
DeselectAllRoutes) so every consumer shares one implementation: v4/v6
exit-pair expansion, exit-node mutual exclusion, and selection
triggering.

Previously the exit-node exclusivity lived only in the daemon's
SelectNetworks RPC, so the Android and iOS bindings could leave two exit
nodes selected until the next network map reconciliation. Both bindings
now call the shared manager methods and enforce exclusivity at toggle
time, matching the desktop behavior.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6928"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787772098&installation_model_id=427504&pr_number=6928&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6928&signature=31ad59e1483e1582cd447a8db2fe21e5309230e631cbd0cad0f977cd15fb7b9b"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Route selection/deselection is now handled through shared
route-manager APIs for both individual routes and “all routes”.
* Exit-node selections automatically enforce mutual exclusivity while
keeping non-exit routes unaffected.
* **Bug Fixes**
* Unknown or unavailable route IDs now return errors, and exclusivity is
preserved even when some route IDs fail.
* **Tests**
* Added route-selection tests covering exclusivity (including
IPv4/IPv6), select-all behavior, partial errors, and invalid IDs.
* **Refactor / Chores**
* Simplified Android, iOS, and server routing flows to delegate to the
shared manager; updated mocks and removed redundant routing command
logic/dependencies.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 17:39:37 +02:00
Zoltan Papp
63c320b6a9 [client] Serialize iOS tunnel reconfiguration callbacks (#6870)
On iOS the tunnel reconfiguration (setTunnelNetworkSettings) was driven
from three independent Go paths without serialization: route prefix
updates via the route notifier's own delivery loop, interface IP/IPv6
set synchronously from the engine start goroutine, and DNS config
applied from the DNS apply chain through a separate Swift object. The
three sources mutated the shared Swift settings-manager state
concurrently and triggered overlapping updateTunnel() calls, losing or
half-applying route and DNS settings.

Introduce client/internal/tunnelnotifier: a single notifier that
implements both listener.NetworkChangeListener and dns.IosDnsManager,
queues all four callbacks (OnNetworkChanged, SetInterfaceIP,
SetInterfaceIPv6, ApplyDns) in one FIFO and delivers them one-by-one
from a single goroutine, so calls into Swift never overlap and arrive in
order. RunOniOS wraps the two Swift objects into the notifier and closes
it after the run loop exits.

The route notifier keeps its prefix dedup but delegates delivery to the
shared notifier instead of maintaining its own queue and loop; the dedup
check and the enqueue stay under one mutex so queue order matches
state-update order. Setting the interface IP becomes asynchronous, which
is safe: on iOS wgInterface.Create() only uses the TunFd, and the FIFO
preserves the IP -> routes/DNS relative order.

The package is build-tag free so the unit tests run on any platform
under -race. Android is unaffected: it receives all settings atomically
in one configureInterface call and serializes TUN rebuilds on a single
handler thread.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [x] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6870"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787404610&installation_model_id=427504&pr_number=6870&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6870&signature=7db66c4ec4262cc21137b3d3e1fd5e905950eab82ef46236a13ac3d512ff2a16"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Improved iOS tunnel networking by coordinating network, interface,
route, and DNS updates through a unified notifier.
* Updated iOS behavior so route/prefix changes are applied immediately
instead of via queued/background delivery.

* **Bug Fixes**
* Improved reliability by ensuring network and DNS-related callbacks are
invoked in order without overlap.
* Ensured pending updates are drained and handled correctly during
shutdown.

* **Tests**
* Added coverage for FIFO ordering, non-overlapping callback execution,
interleaved DNS/route updates, and graceful shutdown behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 16:51:01 +02:00
Stefan Fast
4acbe2670a [client] Escape dots in interface names for sysctl configuration (#6930) 2026-07-28 16:24:30 +02:00
Zoltan Papp
0fb4c8c423 [client] Build UI release binaries with the production tag (#6898)
## Describe your changes

The goreleaser UI configs passed no build tags, so released netbird-ui
binaries were built on the Wails !production path: DevTools enabled,
browser context menu forced on, and FRONTEND_DEVSERVER_URL still
honored.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6898"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787616895&installation_model_id=427504&pr_number=6898&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6898&signature=c9cda21bd668dc58f307e4ef762a194fb28fdb5698786930d911a55811a2bcaf"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Chores**
* Updated release build configurations to mark Linux, Windows, and macOS
UI builds as production releases.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 16:23:09 +02:00
Zoltan Papp
42e45ff9f9 [client] Expose RenameProfile in the Android profile manager binding (#6926)
## Describe your changes
Expose RenameProfile in the Android profile manager binding

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6926"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787768395&installation_model_id=427504&pr_number=6926&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6926&signature=f4684572fa5b03325c89a2ace1e1eed03df3881365acebdc31066b74fa656bdb"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **New Features**
  * Added the ability to rename Android profiles.
  * Rename operations now provide clear success or failure feedback.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 16:22:48 +02:00
Viktor Liu
9269b56386 [management] Read reverse-proxy service and target columns in Postgres path (#6886) 2026-07-28 14:46:19 +02:00
Viktor Liu
b3f9b82442 [management] Force routing-peer DNS resolution for reverse-proxy domain targets (#6872) 2026-07-28 14:45:57 +02:00
Riccardo Manfrin
8a43f4f943 [client] fix build: add ReapplyMatching to dedicated AllowedIPsRefCounter (#6935)
## Describe your changes

#6799 turned `AllowedIPsRefCounter` from an alias of the generic
`Counter` into a dedicated peer-aware type. A change merged in parallel
— `DefaultManager.ReconcilePeerAllowedIPs` (lazy-connection idle→wake
reconvergence) — calls `allowedIPsRefCounter.ReapplyMatching`, which
only existed on the generic `Counter`. Each PR built alone; the merged
`main` did not:

```
client/internal/routemanager/manager.go: m.allowedIPsRefCounter.ReapplyMatching undefined
```

Add `ReapplyMatching(pred, apply)` to the dedicated type, matching the
generic contract (keyed on the active/`Out` peer): it re-applies every
prefix whose currently installed peer satisfies `pred`, skipping
prefixes with no active peer (reconciled by the next
Increment/Decrement). Also update `reconcile_test.go` to construct the
counter via `refcounter.NewAllowedIPs` — the generic `refcounter.New`
return value is no longer assignable to the dedicated type.

Verified with `cd client && CGO_ENABLED=1 go build .` (the failing CI
step) and the `TestReconcilePeerAllowedIPs` / refcounter / routemanager
tests.

## Issue ticket number and link

No public issue — fixes a `main` build break from a semantic merge
conflict between #6799 and the `ReconcilePeerAllowedIPs` change. Failing
run:
https://github.com/netbirdio/netbird/actions/runs/30330882389/job/90185591366

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal build fix restoring a method on the AllowedIPs refcounter. No
public API, CLI, config, or behavior change.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

N/A

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6935"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787814246&installation_model_id=427504&pr_number=6935&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6935&signature=6a0f08ffe5ed39096a77d94e71a5947f68bad6f2b7abee5f1a4823fd44f35106"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved route reconciliation so all applicable allowed IP prefixes
are reliably re-applied for the correct peer.
* Prevented reconciliation from affecting routes assigned to other
peers.
* Improved handling of errors encountered while restoring multiple
routes, providing more consistent results.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 16:42:57 +09:00
Riccardo Manfrin
2f268c8141 [client] fix stale routing peer on overlapping-prefix network removal (#6799)
## Describe your changes

The AllowedIPs reference counter
([refcounter/types.go#L9](https://github.com/netbirdio/netbird/blob/e1a24376a/client/internal/routemanager/refcounter/types.go#L9))
was keyed only by prefix and stored a single active peer set by
the first registrar, never swapped. When two networks advertised the
same prefix via different routing peers, removing the one whose peer was
installed in WireGuard left the prefix pointing at the removed peer
instead of the
surviving one — traffic kept flowing to the old peer until a manual
`netbird down/up`.

Made the AllowedIPs counter peer-aware: it tracks a per-peer reference
count per prefix plus the installed peer, and swaps WireGuard to a
surviving peer when the active one releases its last reference (removes
the prefix when
none remain). `Decrement` now takes the peer key so the exact
incremented peer is released; the static handler records its selected
routing peer like the dynamic and DNS handlers already did. The generic
`Counter` (routes,
  exclusion, ipset) is unchanged.

  ## Issue ticket number and link

No public issue — reported internally (routes not updating without
`netbird down/up` when two networks share a subnet). Root cause is the
prefix-only key at

[refcounter/types.go#L9](https://github.com/netbirdio/netbird/blob/e1a24376a/client/internal/routemanager/refcounter/types.go#L9).

  ## Stack

  <!-- branch-stack -->

  ### Checklist
  - [x] Is it a bug fix
  - [ ] Is a typo/documentation fix
  - [ ] Is a feature enhancement
  - [ ] It is a refactor
  - [x] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

  ## Documentation
  Select exactly one:

  - [ ] I added/updated documentation for this change
  - [x] Documentation is **not needed** for this change (explain why)

Internal client-side routing fix. No public API, CLI, or config change —
only the WireGuard AllowedIPs hand-off when overlapping-prefix networks
are removed.

  ### Docs PR URL (required if "docs added" is checked)
  Paste the PR link from https://github.com/netbirdio/docs here:

  N/A

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6799"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786796229&installation_id=146802194&pr_number=6799&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6799&signature=322d6950b4f664b1cb3421f2efa039fbbb7946de9b79f2a1b1b131dc182ada2c"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Improved routing behavior when multiple peers share the same Allowed
IP by making Allowed IP reference tracking peer-aware.
* Allowed IPs now correctly decrement using the active peer key and
transfer to another surviving active peer when the current peer is
removed.
* Prevented stale routing and incorrect reference cleanup during route
and DNS-driven teardown.

* **Tests**
* Added/extended coverage for peer handoffs, repeated references,
non-active peer removal, flushing behavior, and self-healing after swap
add/remove failures.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 14:10:58 +09:00
Zoltan Papp
e3c4128164 [client] Exit GUI immediately on Windows session end (#6878)
Wails v3 runs the full app teardown synchronously inside WM_ENDSESSION,
overrunning the 5s end-session budget and triggering the "app is
preventing shutdown" screen with a forced kill. Intercept
WM_QUERYENDSESSION/WM_ENDSESSION and exit at once instead, and suppress
error dialogs, toasts, and the hide-on-close hooks once shutdown or a
tray quit has begun.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6878"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787474408&installation_model_id=427504&pr_number=6878&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6878&signature=f09980d83ad32e37b92aa30a7d4552c402a140dbd4e4662a3ca12d1868e9f04b"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Improved shutdown handling on Windows by properly responding to system
end-session requests.
* Prevented the main window, settings window, and error dialogs from
reopening or interfering while the app is closing.
* Updated tray “Quit” flow to begin shutdown immediately, ensuring
active profile/connection operations complete cleanly.
* Suppressed UI notifications during shutdown to avoid stray messages
after exit starts.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-28 14:09:16 +09:00
Maycon Santos
bab5572a74 [management, proxy] scope agent-network model allowlist per policy/group and provider (#6905)
## Describe your changes

The model-allowlist guardrail was merged into one account-wide union and
enforced flat on every request, ignoring which policy/group/provider
authorised it. With multiple policies — especially a mix of guardrailed
and un-guardrailed ones — this caused:

- **false-allow**: a model allowlisted for one group/provider leaked to
any caller; and
- **false-deny**: an un-guardrailed policy (intended unrestricted) was
blocked by another policy's allowlist.

Enforcement is now policy/group-aware, mirroring `llm_limit_check`:

- **Management (`SelectPolicyForRequest`) is authoritative.** It uses
the request model (already carried in
`CheckLLMPolicyLimitsRequest.model`, previously ignored) to keep only
applicable policies whose guardrails permit the model; no
allowlist-enabled guardrail = unrestricted. Denies
`llm_policy.model_blocked` when policies govern the (provider, groups)
but none permits the model.
- **Proxy `llm_guardrail` becomes a per-provider fail-closed backstop.**
The synthesiser emits an allowlist only for providers every authorising
policy restricts; the middleware keys off the resolved provider id and
keeps unknown-model fail-closed.
2026-07-27 20:43:59 +02:00
Zoltan Papp
9b4a5df925 [client] Use platform installer URL for manual update downloads (#6922)
The Wails UI regressed the non-enforced update download to the plain
GitHub releases page. Restore the old Fyne behavior: the tray update
item and the About card's Get installer button now open
version.DownloadUrl(), which points to the direct installer download
(pkgs.netbird.io) per OS/arch and falls back to the generic install page
where no installer exists.

Also fix the dead brew detection on darwin: exec.Command passed the
whole pipeline as a single argument to brew, so the check always failed.
Query the netbird formula and netbird-ui cask explicitly via exit codes
instead.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6922"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787764680&installation_model_id=427504&pr_number=6922&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6922&signature=3d9d640b25a17589114cc704e1d631895ffcd9f554947ff5790fe90d54410d2f"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added a direct installer download option for manual application
updates.
* Non-enforced updates now open the appropriate platform-specific
installer instead of the general releases page.

* **Bug Fixes**
  * Improved macOS download behavior for Homebrew installations.
* Preserved architecture-specific downloads for Intel and Apple silicon
Macs.
* Updated the tray “About” links to use the GitHub repository and
documentation instead of the releases page.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-27 20:08:35 +02:00
Misha Bragin
1816a020c4 [management, proxy] Add Claude Opus 5 (#6895) 2026-07-27 16:15:36 +02:00
Zoltan Papp
aa13928b76 [client] Export agent version info for iOS (#6918)
## Describe your changes

Export agent version info for iOS

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [x] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6918"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787750541&installation_model_id=427504&pr_number=6918&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6918&signature=c2de74d89c36b01aac5866422b0be0b7525f7c6e26e46174efc280339eb864b6"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->
2026-07-27 15:53:09 +02:00
Viktor Liu
d681670a9d [misc] Restore the rootless-latest docker tag (#6914) 2026-07-27 12:07:58 +02:00
Maycon Santos
4f6247b5c3 [management, proxy] Add prompt-cache token and cost accounting to agent network usage (#6900)
Co-authored-by: braginini <bangvalo@gmail.com>
2026-07-26 21:42:41 +02:00
Riccardo Manfrin
1e5b0a5c89 [client] Make Test_ConnectPeers deterministic under Docker/eBPF kernel / Darwin CI (#6884)
## Describe your changes

Make `Test_ConnectPeers` deterministic. Two issues, both surfaced once
the
privileged suite moved into a `--privileged` Docker container (#6425):

1. The peers used `getLocalIP()` as their WireGuard endpoint, i.e. the
host's
routable NIC IP (the docker bridge IP `172.17.0.2` in CI). That address
might
not hairpin reliably inside the container, so the handshake
intermittently
timed out (flaky). Use loopback (`127.1.0.x`) instead — always
self-reachable.
2. On a Linux runner with the WG kernel module the iface uses the eBPF
proxy
factory. Its manager is a singleton with one shared XDP program +
settings
map, so bringing up the two ifaces makes the second factory overwrite
the
first's `wg_port`/`proxy_port` and the handshake is dropped. The test is
   incompatible with the eBPF factory, so disable it via
`NB_DISABLE_EBPF_WG_PROXY` (peers then handshake directly over
loopback).
Running the suite across all three modes (eBPF / UDP proxy / ICE bind)
would
   need a larger refactor.

Also fixes a typo in the `ErrSharedSockStopped` message (`socked` →
`socket`).

## Issue ticket number and link

No public issue — CI flakiness follow-up to #6871 on `Test_ConnectPeers`

(https://github.com/netbirdio/netbird/blob/main/client/iface/iface_test.go).

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Test-only change (plus a log-string typo). No public API, CLI, config,
or
behavior change.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

N/A

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6884"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787485967&installation_model_id=427504&pr_number=6884&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6884&signature=09fb996715d039bd02775a140e8cc03deca5cb42540790b82a744527264a30ad"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Corrected the “shared socket stopped” error message for clearer
output.
* **Tests**
* Improved peer connection test reliability in privileged CI by
disabling the eBPF WireGuard proxy and using fixed loopback UDP
endpoints for deterministic setup.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-24 15:23:22 +02:00
Riccardo Manfrin
b65ec8b68a [client] Restores lost backup.Reset from #4935 (#6883)
## Describe your changes

Restore the management gRPC client's backoff reset after a stream
connects.

This is a regression: PR #4935 originally added this reset; it was then
lost in
commit `58daa674e` during the `withMgmtStream` refactor (Sync/Job
unification).

Compared to before the reset was done AFTER receiveEvents, whereas 
now it's done before. 

Now should cover for Sync and Job (which didn't exist in #4935. Hold for
long living stream that breaks.

Reset prevents two things

1. long living conns going wrong from waiting a long backoff time window
before to drive reconn
2. past MaxElapsedTime (3months) long lived conns failures from being
treated as "unrecoverable" (hence triggering an full restart of the
engine)

## Issue ticket number and link

Internal support case (customer agents lost data-plane connectivity
during a
management maintenance/release window). No public issue. Regression
introduced
in commit `58daa674e`, which removed the `backOff.Reset()` added by PR
#4935.

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal reconnection-behavior fix in the management gRPC client. No
public
NetBird CLI, API, or configuration surface changes.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

N/A

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6883"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787484622&installation_model_id=427504&pr_number=6883&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6883&signature=525b08edbcb94ee6f4ae4226fb1fe8ddb829caa7d4489e8efa639770af9dccdf"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
  * Improved management stream retry behavior.
* Retry delays now reset after a stream is successfully established,
helping subsequent connection attempts recover more quickly.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-24 13:53:53 +02:00
Zoltan Papp
e13bcdbd44 [client] Fetch FreeBSD port files from GitHub mirror instead of cgit (#6880)
cgit.freebsd.org now sits behind an Anubis anti-bot challenge and
intermittently returns an HTML challenge page with HTTP 200 instead of
the requested file, breaking the FreeBSD port release job. Fetch the
port Makefile and distinfo from the official freebsd-ports GitHub
mirror, retry transient failures, and fail loudly if HTML is returned
instead of the expected file.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6880"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787478670&installation_model_id=427504&pr_number=6880&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6880&signature=503a5d79f1671c21cebedac66516c7d7298d401a6382d9bf483813f9e7ee7591"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Improved reliability when retrieving FreeBSD port metadata during
release and issue preparation.
* Added automatic retries and HTTPS-only redirect handling for safer
downloads.
* Validates fetched content to detect unexpected HTML responses and
avoids processing invalid data.
* Updated port metadata retrieval to use a more reliable mirror,
improving version extraction and the resulting comparisons and
regenerated release information.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-24 13:10:53 +02:00
Riccardo Manfrin
46568f7af8 [client] Reconcile routed allowed IPs when a lazy connection goes idle (#6863)
## Describe your changes

Under lazy connections, when a routing peer goes idle its WireGuard peer
is torn down and re-created with a wake endpoint by
the activity listener, carrying only the overlay /32
(`peerCfg.AllowedIPs`). The routed subnet prefixes are dropped from
  the device on the Connected→Idle transition.

They are meant to be restored by the route watcher, which reacts to the
peer's status change and calls `recalculateRoutes` →
  `AddAllowedIP`. Two things prevent that from healing the peer:

- `AddAllowedIP` uses `update_only`, which is a silent no-op (no error)
when the peer does not exist. While the peer is
being torn down and re-armed with its wake endpoint, it is briefly
absent, so a re-add that lands in that window is lost.
- The allowed-IP refcounter only calls its add function on a prefix's
0→1 transition. The routed prefix stays referenced
across the idle cycle, so once the device entry is gone the refcounter
does not re-push it on its own, and nothing retries.

As a result, traffic to the routed subnet is black-holed while the peer
is idle. Because the wake endpoint only fires when a
packet matches the peer's AllowedIPs, a packet to the subnet is dropped
before reaching the wake endpoint, so it cannot
wake the peer. The peer only recovers when woken by other means (e.g. a
ping to its overlay IP).

  ## Approach

This change keeps the existing Connected→Idle transition as-is and
reconciles the AllowedIPs afterwards, avoiding any
additional locking on the transition path. The peer is torn down and
re-armed with its wake endpoint as today; the routed
prefixes are then re-applied from the route manager's allowed-IP
refcounter once the wake endpoint has been (re)armed.

A single add-only method, `ReconcilePeerAllowedIPs(peerKey)`, re-applies
every routed prefix currently tracked for the peer
in the refcounter (the authoritative store; it already covers static,
dynamic and dnsinterceptor routes). It runs whenever
the peer's wake endpoint is (re)created in the lazy manager — every
point where the activity listener builds it with the
  overlay /32 only:

- **initial registration** (`AddPeer`, cold start): the route manager
may have already pushed the peer's routes before the
wake endpoint existed, so those `AddAllowedIP` calls no-op'd; the
reconcile installs them on the freshly created wake
  endpoint.
- **the two paths into idle** (`DeactivatePeer` on a remote GOAWAY,
`onPeerInactivityTimedOut` on local inactivity): the
peer is torn down and re-armed, so the routed prefixes must be
re-applied.

In every case the routed prefixes end up on the wake endpoint, so
traffic to a routed subnet can wake the peer. Arming the
wake endpoint and reconciling are wrapped in a single
`armActivityListener` helper so the two always happen together.

New helper: `refcounter.Counter.KeysMatching(pred)` to enumerate a
peer's prefixes under the counter lock.

Note on scope: the reconcile restores what the refcounter tracks. All
routed AllowedIPs currently go through it, so this
covers the routed-prefix case; it does not attempt to reconcile
AllowedIPs installed outside the refcounter. The
Idle→Connected (wake) path does not need this: the peer is not removed
there (the listener close leaves it in place and only
the endpoint is updated), so a concurrent `AddAllowedIP` lands normally.

  ## Testing

Reproduced deterministically in a local dev setup (userspace client,
`NB_WG_KERNEL_DISABLED=true`, `B_LAZY_CONN_INACTIVITY_THRESHOLD=1`
inactivity threshold 1
min). A temporary 30s sleep in the tear-down → re-arm window widens the
race so the route watcher's async `AddAllowedIP`
reliably lands while the peer is absent and no-ops (the sleep is a test
aid, not part of the change):

- **without the reconcile:** after the peer goes idle, a ping to any
routed IP — both a pre-existing route and one added
  during the window — black-holes; the peer never wakes.
  - **with the reconcile:** the same ping wakes the peer and passes.

Added unit tests: `ReconcilePeerAllowedIPs` (re-applies all of a peer's
tracked prefixes, scoped to that peer) and
  `refcounter.Counter.KeysMatching`.

Note: `netbird status -d` is not a reliable signal for this —
`AddPeerStateRoute` records the route regardless of whether
the underlying `AddAllowedIP` no-op'd, so it reflects the route
manager's intent rather than device state. The reliable
  signal is functional (ping the subnet from idle).
  
  ## Checklist

  - [x] Is it a bug fix
  - [ ] Is a typo/documentation fix
  - [ ] Is a feature enhancement
  - [ ] It is a refactor
- [x] Created tests that fail without the change (unit tests for the
reconcile + `KeysMatching`)

  ## Documentation

- [x] Documentation is **not needed** for this change (internal client
behavior, no API / gRPC / CLI / flag change)

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6863"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787330960&installation_model_id=427504&pr_number=6863&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6863&signature=f3d6a97d7db82e92b3939fdd0f159c5ee74913ff88f4eb82e41e88fcb787aff4"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Routed IP assignments are automatically reconciled and restored
whenever a peer’s lazy wake endpoint is armed or re-armed.
* Routed allowed IPs are re-applied after inactivity transitions and
monitoring re-initialization.
* If reconciliation can’t be performed, the client safely skips it; if
reconciliation encounters issues, failures are logged without stopping
connection monitoring.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-23 18:40:54 +02:00
Riccardo Manfrin
3358138ccc [client] Fix flaky Test_ConnectPeers busy-loop handshake wait (#6871)
## Describe your changes

  `Test_ConnectPeers` in `client/iface` is flaky in CI, timing out with
  `waiting for peer handshake timeout after 30s`.

  The wait loop polled `getPeer()` in a tight busy-loop with no sleep
(a `select` with a `default` branch), pegging a CPU core. The peers run
  userspace WireGuard (stdnet transport), so the spin starved the
wireguard-go goroutines that actually process the handshake, making the
  30s wait flaky under CI load.

Poll on a 500ms ticker instead so the CPU is yielded between checks, and
  check the handshake state before waiting. Same logic, no busy-spin.

  ## Issue ticket number and link

  N/A — CI flakiness fix.

  ### Checklist
  - [x] Is it a bug fix
  - [ ] Is a typo/documentation fix
  - [ ] Is a feature enhancement
  - [ ] It is a refactor
  - [ ] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature

  ## Documentation

  - [x] Documentation is **not needed** for this change (test-only fix)

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6871"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787405146&installation_model_id=427504&pr_number=6871&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6871&signature=a68daf1747583280b5d1d083c6b45ade4c0d441a38fd25ab9a2ca01f87511ad9"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with [code]smith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>@codesmith-bot</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Tests**
* Improved peer connection test waiting behavior to avoid busy spinning.
* Added a timeout and periodic checks, with clearer failure handling
when a handshake does not complete.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-23 15:48:16 +02:00
Viktor Liu
6d15d0729a [client] Clear stale UDP checksum in eBPF XDP proxy after port rewrite (#6861) 2026-07-23 14:37:52 +02:00
Maycon Santos
0936918d24 [management, proxy] add Kimi (Moonshot AI) provider to Agent Network (#6853) 2026-07-23 13:32:32 +02:00
Bethuel Mmbaga
d4a4418969 [misc] Simplify enterprise bootstrap (#6869) 2026-07-23 12:41:13 +03:00
Pascal Fischer
31ed241a1a [management] component types (#6866) 2026-07-23 11:37:09 +02:00
Maycon Santos
178e6a8530 [misc] Probe the agent-network endpoint with a GET instead of getent (#6867)
## Describe your changes

Use http get to trigger lazy connections as e2e will have single proxy
and status checks would fail without proper lazy wake-up
2026-07-23 15:30:40 +09:00
Maycon Santos
96963b6751 [proxy] fix proxy multistage build (#6864) 2026-07-22 21:24:44 +02:00
Maycon Santos
9770814f39 [client] warm lazy connections from the DNS resolver (#6854)
## Describe your changes

Supersedes #6767 (same change, moved to an unprefixed branch; review
feedback from there is addressed here).

With lazy connections enabled, a peer is not dialed until on-demand
traffic arrives, so the first request to a peer resolved by name (e.g.
the agent-network reverse-proxy) races — or loses to — the WireGuard
handshake. Activation was previously reactive only: a data-path packet
or an inbound signal.

This adds a proactive, DNS-time trigger. When the local resolver answers
for an overlay name, it now warms the lazy connection to the peer(s) the
answer points at and waits briefly for one to connect before returning
the response — so by the time the client sends its first packet the
tunnel is already up.

- `dns/local`: new `PeerActivator` capability + `SetPeerActivator`
setter (mirrors the existing `PeerConnectivity`/`SetPeerConnectivity`
injection). `ServeDNS` warms on the pre-filter answer, so activating a
lazily-idle peer also lets it survive the disconnected-peer filter.
Warm-up is scoped to match-only (non-authoritative) zones — the
synthesized private-service zones and user-created zones — so plain
peer-name lookups in the account's authoritative peer zone never wake
idle peers. No-op when no activator is wired (lazy off) or the answer
carries no peer IPs. Budget is `NB_DNS_LAZY_WARMUP_TIMEOUT` (default 2s,
parsed once at construction, invalid values logged); on timeout the
answer is returned anyway (never SERVFAIL).
- The resolver-facing interfaces (`PeerActivator`, `PeerConnectivity`)
take `netip.Addr` instead of string IPs; record addresses are extracted
as `netip.Addr` (v4-mapped forms unmapped) and converted to string only
at the `peer.Status` boundary.
- `SetPeerActivator` is part of the `dns.Server` interface (no-op on the
mock), so the engine wires it without a type assertion.
- `client/internal`: a small engine-side adapter (`dnsPeerActivator`)
resolves answer IPs to peers via `Status.PeerStateByIP`, activates them
through `ConnMgr.ActivatePeer` (HA fan-out included), and polls
`PeerStateByIP` until one is connected. The activation dial is tied to
the engine's long-lived context so a handshake that outlasts the
per-query wait still completes in the background. `ConnMgr.ActivatePeer`
is safe for concurrent use (the lazy manager pointer is guarded by a
dedicated RWMutex and the manager is internally synchronized), so the
DNS path never contends with network-map processing on `syncMsgMux`.

Scope is overlay-only for free: the trigger lives in the local resolver,
which only answers for NetBird-managed names; public/upstream DNS is
unaffected. Already-connected peers short-circuit, so steady-state DNS
latency is unchanged.

## Issue ticket number and link

N/A — follow-up to the lazy-connection rollout; fixes the agent-network
cold-start observed in the e2e (proxy peer stuck disconnected until
traffic).

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal client behavior; the only knob is the optional
`NB_DNS_LAZY_WARMUP_TIMEOUT` tuning env var with a safe default.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

## Tests

- `dns/local`: warm-up invokes the activator with the answer's peer
address in match-only zones; authoritative-zone answers never trigger
warm-up; no-activator path unchanged; no-answer queries don't invoke the
activator; `NB_DNS_LAZY_WARMUP_TIMEOUT` parsing
(valid/invalid/non-positive); `extractRecordAddr` unmaps v4-mapped
record data.
- `client/internal`: `dnsPeerActivator` skips
connected/unknown/conn-less peers with no wait, returns as soon as a
pending peer connects, and releases the DNS response at the budget when
the peer stays idle; `ConnMgr.ActivatePeer` races the manager lifecycle
cleanly under `-race`.
- Agent-network e2e green on this change (with lazy connections
enabled): https://github.com/netbirdio/netbird/actions/runs/29891467544
2026-07-22 18:47:48 +02:00
Nicolas Frati
b0c1ed31b8 [infrastructure] Add unified admin CLI for self-hosted helpers (#6507)
## Summary

Adds a unified `admin` CLI for self-hosted instance administrators in
both the management and combined binaries.

## User Management

### `admin user change-password`
- Changes a local embedded IdP user's password.
- Selects the user with `--email` or `--user-id`.
- Reads the new password from `--password` or `--password-file`.
- Clears the user's local authentication session so the new password is
required on the next login.
- **Alias:** `admin user set-password`.

### `admin user reset-mfa`
- Resets a local embedded IdP user's MFA enrollment.
- Selects the user with `--email` or `--user-id`.
- Clears TOTP/WebAuthn enrollment data and removes the local
authentication session.
- The user will re-enroll MFA on the next login.

## MFA Management

### `admin mfa status`
- Shows whether local MFA is enabled in the account settings.
- Checks the embedded IdP client configuration and reports whether MFA
is enabled there.

### `admin mfa enable`
- Enables local MFA for embedded IdP users.
- Updates embedded IdP clients and saves the account MFA setting.
- Records an audit event on a best-effort basis.

### `admin mfa disable`
- Disables local MFA for embedded IdP users.
- Updates embedded IdP clients and saves the account MFA setting.
- Records an audit event on a best-effort basis.

## Reverse Proxy Tokens

### `admin token create --name <name> [--expires-in <duration>]`
- Creates a reverse proxy access token.
- Prints the plaintext token once, along with the token ID.
- `--expires-in` supports values such as `24h`, `30d`, or `365d`. If
omitted, the token never expires.

### `admin token list`
- Lists reverse proxy access tokens.
- Shows the token ID, name, creation date, expiration, last-used time,
and revocation status.
- **Alias:** `admin token ls`.

### `admin token revoke <token-id>`
- Revokes a reverse proxy access token.
- Revoked tokens can no longer authenticate reverse proxy instances.

## Reverse Proxy Management

### `admin proxy disconnect-all`
- Lists registered reverse proxy instances and force-marks all connected
instances as disconnected.
- Useful for repairing stale proxy state after an unclean management
server shutdown.
- Prompts for confirmation by default.
- `--dry-run` previews the changes without applying them.
- `--force` skips the confirmation prompt.
- Live proxies may appear again after their next heartbeat, reconnect,
or re-registration.

## Compatibility Commands

### `token ...`
- Deprecated top-level compatibility path.
- Behaves the same as `admin token ...`.
- Retained so existing scripts using `token create`, `token list`, or
`token revoke` continue to work.

## Changes

- Adds reusable `management/cmd/admin` command package.
- Wires `admin` into `netbird-mgmt` and `combined`.
- Adds local user password reset with existing password strength
validation.
- Adds local MFA enrollment reset by clearing Dex TOTP/WebAuthn
credentials and local auth sessions.
- Adds local MFA enable/disable/status helpers for embedded IdP
deployments.
- Moves proxy access token commands under `admin token` for a single
admin-focused CLI entry point.
- Exports `server.ValidatePassword` for reuse by CLI helpers.

## Tests

```bash
go test ./management/cmd/...
go test ./management/cmd/admin ./management/cmd ./combined/cmd
go test ./management/server -run TestValidatePassword
```

Pre-push lint also passed.


### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [x] I added/updated documentation for this change
- [ ] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/832

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

## Release Notes

* **New Features**
* Added self-hosted admin CLI commands for changing passwords, resetting
MFA (including WebAuthn), and managing embedded IdP client MFA
(enable/disable/status).
* Introduced a unified admin command entry point and improved
data-directory handling for embedded IdP storage.

* **Refactor**
* Centralized password strength validation into a shared exported
validator.

* **Tests**
* Added a comprehensive admin command test suite covering password
input, selectors, MFA reset, and client MFA state handling.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-22 18:47:19 +02:00
dmitri-netbird
8435682ac8 [client, management] offload client config generation to the client (#6711)
Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>
Co-authored-by: crn4 <vladimir@netbird.io>
Co-authored-by: pascal <pascal@netbird.io>
2026-07-22 18:20:27 +02:00
Zoltan Papp
ed682fad87 [client] Run pnpm install with --ignore-scripts in frontend CI (#6859)
Prevent dependency lifecycle scripts (preinstall/postinstall/prepare)
from executing during install in the UI frontend CI job, closing the
most common npm supply-chain vector at build time. The frontend build
does not rely on any dependency install scripts.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6859"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787319533&installation_model_id=427504&pr_number=6859&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6859&signature=f8285bf8f6381eebd80d2dc801748b2e1dedea6385f3ced7da6bcdc788eb0580"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Chores**
* Updated the UI frontend installation workflow to skip package
installation scripts while preserving the locked dependency versions.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-22 16:01:33 +02:00
Zoltan Papp
a12a3e4603 [client] Keep session deadline visible across reconnects (#6847)
## Describe your changes

The tray session row flapped on every transient engine restart and went
blank at the exact moment the session expired:

- Watcher.Close zeroed the server-scoped recorder deadline on every
engine restart (sleep/wake, network change, stream errors), hiding the
tray row until the next successful login re-seeded it. Close now only
stops the timers; the client run loop clears the recorder when it exits
for real.
- Status.GetSessionExpiresAt masked past deadlines as zero, so the
status snapshot dropped the field right when the UI should have shown
"session expired". The getter now returns the raw value and the tray
renders past deadlines as "Session expired — sign in again", routing the
click to the login flow.
- The watcher rejected recently-expired deadlines (30s skew) delivered
by login/sync, clearing the recorder instead of surfacing the expiry.
Deadlines up to 30 days past are now recorded as expired; only older
values are rejected as garbage.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6847"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787220767&installation_model_id=427504&pr_number=6847&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6847&signature=77977ad47c5eda51e1d76e59bc736203b4592462bf94076811a20ea0f7f23aee"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Tray “session expires” and countdown now accurately reflect expired
sessions and remaining time near boundary moments (including improved
rounding).
* **Bug Fixes**
* Recently expired deadlines are retained for correct status reporting,
while very old past deadlines are rejected and cleared.
* Session expiry “expires at” is preserved when the session watcher
closes during reconnect scenarios.
* Clicking an already-expired session in the tray now routes the user to
the login screen.
* **Tests**
* Updated and expanded coverage for recent/ancient past handling and
watcher close behavior.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-22 15:22:16 +02:00
Riccardo Manfrin
dc89b471fa [client] checks/enforce MDM disableAutostart on every GUI launch, not just fresh installs (#6782)
## Describe your changes

`applyAutostartDefault` gated all MDM enforcement behind the one-time
`AutostartInitialized` marker, so MDM `disableAutostart` only affected
fresh installs — a policy pushed after autostart had been enabled could
not
revoke the OS login-item. 

The PR adds follow-up MDM enforcements at the top of
the function: if at any time MDM sets `disableAutostart=true` and the OS
registration is
present, force `SetEnabled(false)` to align it.

Trade-off: once the admin lifts the policy, autostart stays off until
the user re-toggles
in Settings — consistent with "MDM always wins" behavior of the other
managed keys.

## Issue ticket number and link

Follow-up to PR https://github.com/netbirdio/netbird/pull/6738
(introduced
the `disableAutostart` MDM key with fresh-install-only semantics).

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [x] I added/updated documentation for this change
- [ ] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

<https://github.com/netbirdio/docs/pull/855>


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **New Features**
- Added an administrative policy to disable client autostart (“Disable
Autostart”).
- Added support for configuring this policy via macOS MDM, Windows Group
Policy (ADMX/ADML), and registry settings.
- When enforced, the client prevents new autostart registration on fresh
installs and removes existing autostart on the next GUI launch, keeping
it disabled until the policy is lifted.

- **Bug Fixes**
- Improved enforcement logic for managed autostart defaults so policy
state is applied consistently during startup and first-run setup.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-22 11:34:13 +02:00
Zoltan Papp
0e520ee9f5 [client] Copy trustedproxy package into Docker build context (#6851)
## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6851"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787231318&installation_model_id=427504&pr_number=6851&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6851&signature=dcbd1f845098d270ded2fec67182952868fbcb9f5838ef1fe5f3b6f6406f01d4"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Updated the build process to include trusted proxy configuration in
the application image.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 16:21:49 +02:00
Zoltan Papp
9620890b65 [client] Always connect on profile selection except in manage profiles (#6838)
Selecting a profile from the header dropdown or the tray submenu now
always brings the connection up after the switch, regardless of the
previous daemon state. Switching from the manage-profiles screen
(including profile creation) never connects, leaving a chance to adjust
the management URL first.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6838"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787157550&installation_model_id=427504&pr_number=6838&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6838&signature=8fe9c5f0779df46c76b4135a373591762eb1498d890ff6996626fb75d6433e0a"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added the ability to switch profiles without automatically
establishing a connection.
* Existing profile switching continues to connect when appropriate,
while safely handling active or pending connections during the switch.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 15:28:11 +02:00
Eduard Gert
69c35e31b4 [client] Fix browser dialog not closing on renew session flow (#6745)
## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* SSO browser-login popups now open centered on the display where the
cursor is located, and they recenter correctly on subsequent opens.
* Programmatic cleanup no longer triggers “login canceled” behavior;
cancel is emitted only when the user closes the active popup.
* **New Features**
* Added a streamlined “close renewal flow” action that tears down the
session-renewal UI by closing both the login and session-expiration
popups.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-21 15:01:31 +02:00
Zoltan Papp
b6cd8944b1 [client] Fix nil context panic in iOS dynamic route resolver (#6848)
## Describe your changes

getIPsFromResolver passed a nil context to ExchangeWithFallback, which
net.Dialer.DialContext rejects with panic("nil context"). On iOS this
crashed the whole network extension (SIGABRT) ~2 seconds after connect
whenever the network map contained a domain-based (dynamic) route, as
the resolver goroutine panicked on its first DNS query.

Passing nil used to be a documented input of ExchangeWithFallback ("If
the passed context is nil, this will use Exchange instead of
ExchangeContext") since #3632. 9ed2e2a5b (#5971) removed the nil-context
branch, but this iOS-only caller was not updated — it never fails CI
since route_ios.go only builds with GOOS=ios. Broken since v0.71.1.

Pass a context bounded by the existing dialTimeout instead, matching the
dnsinterceptor pattern (context.Background() + timeout).

Captured panic (netbird.err):
  panic: nil context
  net.(*Dialer).DialContext -> miekg/dns ExchangeContext ->
nbdns.ExchangeWithFallback(nil, ...) ->
dynamic.(*Route).getIPsFromResolver
  route_ios.go:35


## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6848"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787223752&installation_model_id=427504&pr_number=6848&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6848&signature=881d3b30aa4bd15001562ad417ff946b70c7c108856a1b3808b776c12b7022a9"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Fixed iOS dynamic DNS resolution by ensuring DNS queries use a proper
resolver context instead of a nil one.
* Improved DNS resolution reliability by propagating
cancellation/timeouts through all domain IP lookups, including fallback
system resolver queries.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 13:37:29 +02:00
Zoltan Papp
6fc05efa6c [client] Disconnect daemon on GUI quit via async Down (#6796)
The tray Quit menu now disconnects the daemon before exiting instead of
only tearing down the GUI. A new DownAsync RPC lets the daemon start the
teardown and return immediately: beginDown cancels the connection under
the mutex (so it cannot reconnect), then finishDown (the retry-goroutine
wait and status reset) runs on a background goroutine. handleQuit aborts
any in-flight profile switch first (so a queued Up cannot reconnect
during teardown) and calls DownAsync so quitting never blocks on the
engine shutdown.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6796"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786785261&installation_id=146802194&pr_number=6796&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6796&signature=4fc95ebfed320240170c8318a2bc8750acfa5ad131454e3011c43bc8a5a09ed1"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Improved shutdown reliability by continuing teardown even when
stopping the service fails (stop errors are logged but not returned).
* Refined connection shutdown to return “service not up” errors directly
for clearer, more immediate RPC behavior.
* Prevented shutdown hangs by making the tray Quit disconnect
time-bounded (5 seconds).
* Ensured any in-flight profile switch is cancelled before exiting, with
quit serialized to avoid races.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 13:34:30 +02:00
Eduard Gert
3cda14d7f2 [client] Use menu bar wording on macOS welcome screen (#6810)
## Describe your changes

The post-install welcome step said "Look for NetBird in your tray" on
every OS, but macOS has no system tray — the icon sits in the menu bar.
The tray wording stays correct on Windows and Linux.

- Add `welcome.titleMac` and `welcome.descriptionMac` (following the
existing `settings.advanced.interfaceName.errorMac` key convention) to
`en` and all nine translated bundles (`de`, `es`, `fr`, `hu`, `it`,
`ja`, `pt`, `ru`, `zh-CN`), each using that language's Apple term for
the menu bar (Menüleiste, barra de menús, barre des menus, menüsor,
barra dei menu, メニューバー, barra de menus, строка меню, 菜单栏). The `ja`
bundle landed on main (#6790) after the initial commit and was covered
after merging main back in.
- `WelcomeStepTray.tsx` picks the key via `isMacOS()`, which it already
uses to choose the per-OS screenshot.

Verified: `go test ./client/ui/i18n/...`, `tsc --noEmit`, `eslint`, and
`prettier --check` all pass; key set and placement verified identical
across all ten bundles. Not visually verified in the running app
(headless session) — the welcome dialog only shows on first launch.

## Issue ticket number and link

Fixes NET-1411

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (UI copy fix only;
no behavior, API, or configuration change)

### Docs PR URL (required if "docs added" is checked)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6810"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786865844&installation_id=146802194&pr_number=6810&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6810&signature=c18b58a9624d319bff3010aeb92c7b11cc7e99f82ffa5e15a32053500d620af4"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **New Features**
* Added macOS-specific onboarding text that directs users to find
NetBird in the menu bar.
* Updated localized welcome content across supported languages, while
retaining platform-specific tray guidance where applicable.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-21 11:26:27 +02:00
Eduard Gert
d9392fdbb8 [client] Clarify outdated NetBird client overlay (#6718)
## Describe your changes
Add more information about the current client and GUI versions if the
user is running an older client.
Update the URL to download the latest RC if the user is running any RC
build.

<img width="348" height="497" alt="CleanShot 2026-07-10 at 15 11 54"
src="https://github.com/user-attachments/assets/2579f5f3-5056-4184-96bb-9f93d5860328"
/>

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__
2026-07-21 11:26:16 +02:00
Maycon Santos
82fdfa84b8 [proxy] match Bedrock provider models against the normalized request model (#6773)
## Describe your changes

Native AWS Bedrock requests carry the model in the URL path as a
cross-region inference-profile id (e.g.
`us.anthropic.claude-haiku-4-5`). The request parser normalizes that to
the catalog key (`anthropic.claude-haiku-4-5`) before the router runs,
but the router matched it against the operator's registered provider
models with exact string equality. So a Bedrock provider registered with
the id Bedrock actually uses (`us.anthropic…`) never matched a
normalized request → the request denied with
`llm_policy.model_not_routable` ("no provider configured for model …").
Only a provider registered with the already-stripped catalog id worked,
which is not how Bedrock ids appear.

Fix: introduce a single shared `llm.NormalizeBedrockModel` (the same
ARN/region-prefix/version-suffix stripping the parser already does) and,
in the router's `routeClaimsModel`, normalize a **Bedrock** route's
candidate models before comparing. Now a Bedrock provider registered
with either the raw inference-profile id or the normalized catalog id
matches the request. Non-Bedrock routes keep exact matching.

Surfaced by the new native-Bedrock e2e (`WireBedrock`,
`/model/{id}/invoke`); the old e2e used the Anthropic body shape, which
never normalized either side and so hid this.

The request parser keeps its own identical normalizer for now;
de-duplicating it onto `llm.NormalizeBedrockModel` is a trivial
follow-up.

## Issue ticket number and link

N/A — follow-up to the Agent Network Bedrock support / model-allowlist
work.

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal routing correctness fix; no user-facing surface change.

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

## Tests

- `proxy/internal/llm`: `NormalizeBedrockModel` unit cases (region
prefixes, version suffixes, ARN).
- `proxy/internal/middleware/builtin/llm_router`: `routeClaimsModel`
matches a Bedrock route registered with the raw `us.anthropic…` id
against a normalized request model; non-Bedrock routes still match
exactly.

Note: full through-tunnel e2e verification of this (the native-Bedrock
`TestProvidersMatrix/bedrock`) also needs the DNS lazy-connection
warm-up (separate PR) to get the client past the proxy-peer gate; they
converge once both land.


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved Amazon Bedrock model matching across ARN formats, regional
prefixes, and version or throughput suffixes.
* Bedrock routes now correctly match equivalent model identifiers even
when requests and route configurations use different formats.
  * Non-Bedrock model matching remains exact.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 10:10:12 +02:00
Zoltan Papp
ca80e49aa0 [client] Refresh WireGuard stats in mobile debug bundles (#6814)
The iOS and Android DebugBundle paths built GeneratorDependencies
without setting RefreshStatus, so the bundle's status.txt read the
cached peer state instead of live WireGuard interface stats. When the
periodic health probe had not run yet, connected relayed peers showed
"handshake: -" and "0 B/0 B" even though the interface was passing
traffic.

Wire RefreshStatus to RunHealthProbes on both platforms, matching the
desktop daemon path in client/server/debug.go. The engine reference is
already available in the cc.Engine() block used for client metrics.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6814"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786881440&installation_id=146802194&pr_number=6814&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6814&signature=c07cbcef1028a366312f7743e6479f2fddafd568aac498ab95828a372ac6843c"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved debug bundle generation on Android and iOS by refreshing
connection health status before collecting diagnostic information.
* Ensured debug bundles include more current health-related data for
troubleshooting.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-21 09:25:33 +02:00
Zoltan Papp
51f17bf919 [client] Update wails to v3.0.0-alpha2.117 (#6837)
## Describe your changes

Update wails to v3.0.0-alpha2.117

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6837"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787156447&installation_model_id=427504&pr_number=6837&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6837&signature=600f7cb9ecda9ee75b7fe841ed81149507d00dbf6323c61aa92ac2acdaea2487"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Chores**
  * Updated the application framework dependency to a newer release.
  * Removed an obsolete supporting dependency requirement.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-20 21:12:22 +02:00
Zoltan Papp
724c6a06e6 [relay] only trust X-Real-Ip headers from configured trusted proxies (#6833)
The WS listener unconditionally trusted X-Real-Ip/X-Real-Port headers,
letting any client forge the source address the relay logs. Gate header
trust behind a trusted-proxy allowlist; ignore the headers unless the
immediate peer matches a configured prefix. Defaults to never trusting
the headers when the allowlist is empty.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6833"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1787141580&installation_id=146802194&pr_number=6833&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6833&signature=9cf182cc7be248e457dfdb56e8a047401276d8cc567ed8ae715ec1cc809f1b6a"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Added `--trusted-proxies` to configure a comma-separated allowlist of
trusted upstream proxy IPs/CIDRs.
* **Behavior Changes**
* Relay WebSocket now uses `X-Real-Ip` / `X-Real-Port` only when the
immediate peer is from the configured trusted set; otherwise it falls
back to the direct remote address.
* Proxy client IP resolution is now consistent and honors
`X-Forwarded-For` only through trusted hops.
* **Operational**
  * Invalid `--trusted-proxies` values fail fast on startup.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-20 18:15:03 +02:00
Maycon Santos
d64e9542eb [proxy] Bedrock cost-allocation metadata + per-provider metadata_disabled (#6791) 2026-07-20 14:45:50 +02:00
Viktor Liu
3fb26d458e [relay] Remove deprecated Hello handshake and gob token decode (#6783) 2026-07-20 14:13:40 +02:00
s-shimizu-clpl
a411fd300c [client] Add Japanese (ja) UI translation (#6790)
## Describe your changes
Adds a Japanese (ja) locale for the desktop UI, following the procedure
in `client/ui/i18n/TRANSLATING.md`.

- New `client/ui/i18n/locales/ja/common.json` — all 441 keys from the
`en` source bundle, `message` only, same key order.
- Registered in `client/ui/i18n/locales/_index.json`: `{"code": "ja",
"displayName": "日本語", "englishName": "Japanese"}`.

Translation notes:
- Placeholders (`{version}`, `{count}`, `{name}`, `{remaining}`, ...),
`\n`, trailing `…`/`...`, the leading space in
`notify.update.enforcedSuffix`, and the `<docs>…</docs>` inline-link
tags are all preserved verbatim.
- Brands kept as-is: NetBird, WireGuard® (® preserved), Rosenpass,
GitHub, NetBird Cloud. Acronyms kept: SSO, DNS, IP/IPv6, ACL, SSH, JWT,
TTL, SFTP, MTU, PSK, LAN, P2P, ICE, IdP.
- Formal register (です・ます); short labels for buttons/tray; quoted UI
labels use Japanese brackets 「」.

No code changes are required: the React frontend auto-loads every
`locales/*/common.json` via `import.meta.glob`, the tray/Go side embeds
the tree via `//go:embed all:i18n/locales`, and the language picker
lists whatever `_index.json` declares.

### Verification
- `go test ./client/ui/i18n/` passes.
- JSON valid; key set and order identical to `en`; no `description`
fields; no empty messages; placeholder/`\n` counts match `en`.
- Built the Windows UI and confirmed the General, Network, Security, and
Troubleshooting settings tabs plus the main window render correctly in
Japanese (language picker shows "日本語 (Japanese)"), with no text
truncation, overflow, or leaked placeholders.

## Issue ticket number and link
N/A

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— adding a shipped UI locale is the documented extension path in
`client/ui/i18n/TRANSLATING.md`.

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change — UI translation
strings only; no user-facing product docs are affected.

### Docs PR URL (required if "docs added" is checked)
N/A

🤖 Generated with [Claude Code](https://claude.com/claude-code)

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6790"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786742881&installation_id=146802194&pr_number=6790&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6790&signature=a1dcd7faf63ae10071ca2da3639b4daf6505eeb0258790e622222e742a8ebcb6"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
  * Added Japanese as a supported interface language.
* **Localization**
* Provided full Japanese translations for the UI, covering navigation,
settings, onboarding, desktop/tray notifications, connection and status
messaging, troubleshooting, update/about screens, and
authentication/error dialogs.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: Eduard Gert <kontakt@eduardgert.de>
2026-07-20 14:12:13 +02:00
Viktor Liu
92a5ed19d3 [client] Merge 0.74.7 to main (#6819)
## Describe your changes

Merge 0.74.7 to main

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6819"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786890471&installation_id=146802194&pr_number=6819&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6819&signature=6078105908fabf6520f504b594cc394dc2a2d749a709c88c22f2d187301f6c71"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Firewall now safely evaluates fragmented inbound traffic while
blocking unsupported, malformed, overlapping, or unapproved fragments.
* SOCKS5 listener address and port can be configured through environment
variables, defaulting safely to loopback.
* **Bug Fixes**
* Strengthened token validation for malformed authentication responses.
* Prevented invalid hostnames from injecting settings into generated SSH
configurations.
* Improved handling of command-line-like account names and concurrent
QUIC connections.
* **Tests**
* Added coverage and performance benchmarks for fragmentation, listener
configuration, and security validation.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-18 00:07:44 +09:00
Viktor Liu
be6777427d Merge remote-tracking branch 'origin/main' into 0.74.7-branch-sync
# Conflicts:
#	client/internal/peer/conn.go
#	client/internal/peer/conn_test.go
#	client/internal/peer/wg_watcher.go
#	client/internal/peer/wg_watcher_test.go
#	go.mod
#	go.sum
2026-07-17 16:34:58 +02:00
Viktor Liu
a1c9427d80 [client] Evaluate IP fragments against firewall ACLs (#6781) 2026-07-17 16:05:32 +02:00
Pascal Fischer
a59d7fba95 [management] propagate auth grant types for combined server (#6817) 2026-07-17 15:41:57 +02:00
Zoltan Papp
41d7bf4bbd [client] Diagnose empty vs corrupt state (#6816)
## Describe your changes
When loadStateFile fails to unmarshal the state file, log whether the
file is empty (0 bytes) or has malformed content, including the byte
size.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [x] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
aste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved state-file loading warnings by distinguishing empty files
from files containing malformed content.
  * Preserved existing recovery behavior for corrupted state files.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-17 15:31:07 +02:00
Riccardo Manfrin
b7b0d5796e [client] Bind netstack SOCKS5 proxy to 127.0.0.1 by default (#6812)
## Describe your changes

In netstack mode the SOCKS5 proxy bridges local host applications into
the
userspace WireGuard stack (`client/iface/netstack/proxy.go`), so it only
needs
to be reachable from the same machine. It was binding to `0.0.0.0`,
making an
unauthenticated proxy reachable from the network — any host able to
reach the
port could relay traffic through the client into its NetBird overlay.

Bind to `127.0.0.1` by default. Add `NB_SOCKS5_LISTENER_ADDRESS` to
override the
bind host for the rare case the proxy must be reachable from other hosts
(e.g. a
container gateway); it is validated as an IP and falls back to loopback.
`ListenAddr` is split into `listenHost`/`listenPort` helpers, with
tests.

Behavior change: setups that relied on reaching the netstack SOCKS5
proxy from
another host must now set `NB_SOCKS5_LISTENER_ADDRESS=0.0.0.0`
explicitly.

## Issue ticket number and link

Internal security hardening of the netstack SOCKS5 listener bind address

([client/iface/netstack/env.go](https://github.com/netbirdio/netbird/blob/main/client/iface/netstack/env.go)).

## Stack

- `0.74.7-branch` - ⚠️ No PR associated with branch <!--
branch-stack -->
  - \#6812 :point\_left:

### Checklist

- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation

Select exactly one:

- [x] I added/updated documentation for this change
- [ ] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)

Paste the PR link from <https://github.com/netbirdio/docs> here:

<https://github.com/netbirdio/docs/pull/860>

<!-- codesmith:footer -->

***

<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6812"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786872199&installation_id=146802194&pr_number=6812&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6812&signature=6e78df77b784915baee9647b555483a8f7550a604d5393edcc501dd8f58b1395"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->

<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **New Features**
- Added configuration options for the SOCKS5 listener’s bind address and
port.
- SOCKS5 now defaults to listening only on the local machine for
improved security.
- Valid address and port overrides are supported, with safe defaults
used for invalid values.

- **Tests**
- Added coverage for default settings and valid or invalid address and
port configurations.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-17 15:09:09 +02:00
Pascal Fischer
21fc5b81f6 [management] allow disabling device code flow when using dex (#6809) 2026-07-17 12:41:08 +02:00
dmitri-netbird
9906b9b1a1 [management] fix a flake in account_test (#6811)
## Describe your changes
In
"TestDefaultAccountManager_UpdateAccountSettings_NetworkRangePreserved",
in the beginning of the test, during account creation a random /16
subnet from 10.64.0./10 network is used. Later in the test a new range
(10.99.0.0/16) is assigned to the account, but it's one of the possible
subnets used during account creation, which sometimes leads to a
collision and failed test.
Using a network outside of the range of networks used during account
creation fixes the issue.

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6811"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786868463&installation_id=146802194&pr_number=6811&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6811&signature=c0ded1a8ecc05686b6ba19eb366519852ac35e170da4f8522fb7aec2062d4a3e"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Tests**
* Updated account network range test coverage to verify peer IP
reallocation with a distinct network range.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>
2026-07-17 11:21:11 +02:00
Riccardo Manfrin
3f8c447378 [client] Rename isValidAccessToken to reflect audience-only check (#6806)
## Describe your changes

`isValidAccessToken` only decodes the JWT payload and checks the
audience
claim, but its name suggested full token validation. Rename it to
`validateTokenAudience` and document what it does: a client-side
audience/shape
check on a token just obtained from the IdP over TLS. Token authenticity
is
enforced server-side by the management server, which verifies the
signature
against the IdP's JWKS (`shared/auth/jwt/validator.go`) on every
request.

Also harden the parser: a non-empty token lacking the three-part JWT
structure
caused an index-out-of-range panic (`strings.Split(token, ".")[1]`); the
shape
is now validated first. `parseEmailFromIDToken` is documented as
best-effort UX
data (login hint/display), never used for authorization. Added tests for
audience matching, malformed tokens, and the panic regression.

Changes:

- Rename `isValidAccessToken` → `validateTokenAudience`; document that
it does
not verify the signature and that authenticity is enforced server-side.
- Fix an index-out-of-range panic on a non-empty token lacking JWT
structure
(`strings.Split(token, ".")[1]`) by validating the three-part shape
first.
- Document `parseEmailFromIDToken` as best-effort/unverified, used only
for the
  login-hint/display UX, never for an authorization decision.
- Add `util_test.go` covering audience matching (string and array),
missing
  audience, malformed payloads, and the panic regression.

## Issue ticket number and link

Internal cleanup: rename a misleadingly-named client-side helper and
harden JWT
parsing against malformed input (`client/internal/auth/util.go`).

## Stack

- `0.74.7-branch` - ⚠️ No PR associated with branch <!--
branch-stack -->
  - \#6806 :point\_left:

### Checklist

- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation

Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal client-side helper rename plus a panic hardening fix. No public
API,
gRPC, CLI/service flag, or configuration change; token authenticity
enforcement
(server-side JWKS verification) is unchanged.

### Docs PR URL (required if "docs added" is checked)

Paste the PR link from <https://github.com/netbirdio/docs> here:

N/A

<!-- codesmith:footer -->

***

<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6806"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786811124&installation_id=146802194&pr_number=6806&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6806&signature=fe45ce9f6df46a609594f84037aaab2a21893621f29d2be48d7f8b347c3c8776"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->

<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **Bug Fixes**
- Improved access-token audience validation during device and PKCE
authentication flows.
- Malformed tokens now return clear validation errors instead of risking
runtime failures.
  - Added support for validating both string and array audience claims.

- **Tests**
- Added coverage for malformed tokens, invalid claims, missing
audiences, and panic prevention.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-17 11:00:55 +02:00
Viktor Liu
6e3f4d8722 [client] Disable gVisor TCP RACK loss detection on Windows (#6808) 2026-07-17 10:45:30 +02:00
dmitri-netbird
877e889250 [management] fix fetching of missing settings in GetAccount call (#6800)
## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6800"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786797566&installation_id=146802194&pr_number=6800&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6800&signature=36a7053e1029e5de1d496fbd9c428a1745d0ec22b3e6479d64aafb1b8350b0ef"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **Bug Fixes**
* Ensure account settings are fully preserved through save/load,
including automatic update and peer exposure preferences.

* **Tests**
* Added coverage to verify account settings remain unchanged after
database persistence and retrieval (skipped on Windows due to SQLite
limitations).
* Introduced deterministic test-data population helpers to reliably set
struct fields for deeper settings verification.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>
2026-07-17 10:38:43 +02:00
Riccardo Manfrin
099ae4bc6c [client] Sanitize peer FQDN/hostname in generated SSH config (#6805)
## Describe your changes

Validate peer-supplied FQDN and hostname before they are written into
the
generated NetBird SSH client config (`client/ssh/config/manager.go`).
These
values originate from remote peers and were previously written verbatim
into
the config; malformed values (e.g. containing unexpected characters)
could
produce a broken or unintended config. FQDN/hostname are now checked
with
`domain.IsValidDomainNoWildcard`, and invalid, non-empty values are
dropped
with a warning. IPs are unaffected (already validated `netip.Addr`).
Added a
test covering malformed hostnames.

## Issue ticket number and link

Internal input-validation hardening for peer-supplied hostnames in the
generated SSH client config (`client/ssh/config/manager.go`).

## Stack

- \#6726 <!-- branch-stack -->
  - \#6805 :point\_left:

### Checklist

- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation

Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Internal client SSH config generation. No public API, gRPC, CLI/service
flag,
or configuration change — only input validation on peer-supplied
hostnames
before they are written to the generated ssh\_config.

### Docs PR URL (required if "docs added" is checked)

Paste the PR link from <https://github.com/netbirdio/docs> here:

N/A
2026-07-17 10:10:22 +02:00
Viktor Liu
63d60ba490 [client] Reject leading hyphen in getent input to prevent flag injection (#6787) 2026-07-17 09:28:33 +02:00
Zoltan Papp
d15830a2d0 [client] Sync 0.74.6 fix/ios-relogin (#6795)
## sync 0.74.6 fix/ios-relogin

NewAuth built a fresh in-memory config on every call via
CreateInMemoryConfig, which generates a new WireGuard private key when
none is set. The iOS Swift layer calls this on interactive re-login and
writes the resulting config back to the profile's netbird.cfg, so each
re-auth replaced the peer's persisted private key with a new one. A new
key means a new public key, so the management server registered a
brand-new peer on every re-authentication — named after the fallback
hostname.

Load the existing config with DirectUpdateOrCreateConfig when a config
file is already present so re-login reuses the peer's persisted private
key (and its identity). Only fall back to a fresh in-memory config for
the first-time login when no config file exists yet (or after logout,
which deletes the file). DirectUpdateOrCreateConfig uses non-atomic
writes so it also works inside the tvOS App Group sandbox. This matches
what Run() and LoginForMobile() already do.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See

[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License

Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked) Paste the PR link
from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__

<!-- codesmith:footer -->
---
<a
href="https://app.blacksmith.sh/netbirdio/codesmith/netbird/pr/6795"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-light-v2.svg"><img
alt="View with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/view-with-codesmith-dark-v2.svg"></picture></a>
<a
href="https://backend.blacksmith.sh/track/enable-autofix?expires=1786784867&installation_id=146802194&pr_number=6795&repository=netbirdio%2Fnetbird&return_to=https%3A%2F%2Fgithub.com%2Fnetbirdio%2Fnetbird%2Fpull%2F6795&signature=01bd94ea58128257c33f4d430ab3b71ce21eac02577ce90c009105b8c0027d9c"><picture><source
media="(prefers-color-scheme: dark)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"><source
media="(prefers-color-scheme: light)"
srcset="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-light.svg"><img
alt="Autofix with Codesmith"
src="https://pr-comments-assets.blacksmith.sh/codesmith/autofix-with-codesmith-dark.svg"></picture></a>
<sup>Need help on this PR? Tag <code>/codesmith</code> with what you
need. Autofix is disabled.</sup>

<!-- codesmith:autofix:disabled -->
<!-- /codesmith:footer -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **Bug Fixes**
- Improved iOS login handling when a configuration location is provided.
- Existing WireGuard keys can now be reused across subsequent logins,
helping avoid unnecessary key regeneration.
- Login continues to support temporary in-memory configuration when no
persistent location is available.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-16 14:38:01 +02:00
Zoltan Papp
141f3d0390 [client] Fix DNS probe listener impossible panic on unparseable local address (#6797)
generateFreePort used netip.MustParseAddrPort on the OS-produced
LocalAddr().String(), which panics on address strings that don't parse.
Eliminate the parsing entirely by reading the port from the concrete
*net.UDPAddr that net.ListenUDP returns, and construct the bind address
directly. The probe listener is bound with udp4 so only an IPv4 wildcard
address is ever used.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

## Summary by CodeRabbit

* **Bug Fixes**
  * Improved reliability when selecting an ephemeral UDP port.
  * Avoided potential failures when determining the assigned port.
* Preserved existing error handling and diagnostic logging for listener
operations.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-16 14:37:27 +02:00
Pascal Fischer
e1a24376ab [management] build routes for peer cache on network map components (#6780) 2026-07-15 18:24:48 +02:00
Viktor Liu
62fc8d254e [relay] Handle QUIC connections concurrently to prevent handshake head-of-line blocking (#6784) 2026-07-15 18:18:47 +02:00
Zoltan Papp
3a2f773d65 [client] preserve WireGuard key on interactive re-login (#6777)
NewAuth built a fresh in-memory config on every call via
CreateInMemoryConfig, which generates a new WireGuard private key when
none is set. The iOS Swift layer calls this on interactive re-login and
writes the resulting config back to the profile's netbird.cfg, so each
re-auth replaced the peer's persisted private key with a new one. A new
key means a new public key, so the management server registered a
brand-new peer on every re-authentication — named after the fallback
hostname.

Load the existing config with DirectUpdateOrCreateConfig when a config
file is already present so re-login reuses the peer's persisted private
key (and its identity). Only fall back to a fresh in-memory config for
the first-time login when no config file exists yet (or after logout,
which deletes the file). DirectUpdateOrCreateConfig uses non-atomic
writes so it also works inside the tvOS App Group sandbox. This matches
what Run() and LoginForMobile() already do.

## Describe your changes

## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **New Features**
* Added support for loading or creating persistent configuration when a
configuration file path is provided.
* Continued support for in-memory configuration for temporary or
first-time use.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-15 14:12:12 +02:00
Pascal Fischer
8f901f8899 [management] enable pprof via env var (#6778) 2026-07-15 12:05:40 +02:00
Maycon Santos
c6bf5fbbfb [management,client] 0.74.5 branch sync (#6769)
## Describe your changes
* [proxy] enforce model allowlist for URL-routed providers
(Bedrock/Vertex) by @mlsmaycon in
https://github.com/netbirdio/netbird/pull/6764
* [management] Remove proxy peer stale deduplication logic by @mlsmaycon
in https://github.com/netbirdio/netbird/pull/6768
## Issue ticket number and link

## Stack

<!-- branch-stack -->

### Checklist
- [ ] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation
Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)
Paste the PR link from https://github.com/netbirdio/docs here:

https://github.com/netbirdio/docs/pull/__


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **New Features**
- Added model-allowlist guardrails for path-routed providers, including
Bedrock and Vertex.
  - Added Bedrock request support for chat interactions.
  - Added guardrail management capabilities.

- **Bug Fixes**
- Requests with missing or blank model identifiers are now denied when a
model allowlist is configured, improving fail-closed protection.
- Corrected provider-specific request handling and session tracking for
Bedrock interactions.

- **Tests**
- Expanded coverage for allowlist enforcement and provider routing
scenarios.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Theodor Midtlien <theodor@midtlien.com>
Co-authored-by: blaugrau90 <61945343+blaugrau90@users.noreply.github.com>
Co-authored-by: Viktor Liu <17948409+lixmal@users.noreply.github.com>
2026-07-14 21:22:40 +02:00
Maycon Santos
f0eed7564f [management] Remove proxy peer stale deduplication logic (#6768)
## Describe your changes

Removing a leftover from an initial implementation. We ended up
resolving it on the client with status checks on the DNS response

## Issue ticket number and link

## Stack

- \#6726 <!-- branch-stack -->
  - \#6768 :point\_left:

### Checklist

- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation

Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

### Docs PR URL (required if "docs added" is checked)

Paste the PR link from <https://github.com/netbirdio/docs> here:

<https://github.com/netbirdio/docs/pull/>\_\_

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **New Features**
- Added support for Bedrock-native request routing in agent network
scenarios.
- Added guardrail management capabilities for creating and removing
model allowlists.

- **Bug Fixes**
- Model allowlists now reject requests when the model is missing or
blank.
- Improved Rosenpass and WireGuard recovery after repeated handshake
failures.
- Improved relay connection handling so status and cleanup operations
remain responsive during stalled connections.
- Updated private service DNS zones to avoid unintended search-domain
behavior.

- **Tests**
- Expanded coverage for model allowlists, handshake recovery, relay
concurrency, and Bedrock routing.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
2026-07-14 20:13:00 +02:00
Maycon Santos
277d8e4c53 [proxy] enforce model allowlist for URL-routed providers (Bedrock/Vertex) (#6764)
## Describe your changes

The Agent Network policy Guardrail "Model Allowlist" was not enforced
for providers whose model travels in the URL/path rather than the JSON
body — most visibly AWS Bedrock (reported in netbirdio/netbird#6751),
and the same class applies to Google Vertex.

Root cause: the `llm_guardrail` allowlist check **failed open**.
`evaluateAllowlist` returned allow whenever the request model was absent
from the metadata bag (`middleware.go`, `if !modelPresent { return nil
}`). The model is stamped upstream by `llm_request_parser`; for
body-routed providers (OpenAI/Anthropic) it comes from the JSON body,
but for path-routed providers the model is recovered only when the
request matches a recognized path shape (Bedrock
`/model/{id}/{invoke|converse|...}`, Vertex
`/v1/projects/.../publishers/.../models/...`). Any shape the parser did
not recognize reached the guardrail with no model and was allowed
regardless of the allowlist.

Fix (provider-agnostic): **fail closed**. When an allowlist is
configured and the model cannot be determined (absent or empty), the
request is denied `403` with a distinct `llm_policy.model_unknown`
reason. This closes the bypass for Bedrock, Vertex, and any future
URL-routed provider in one place. When no allowlist is configured,
behavior is unchanged.

The model allowlist is enforced solely in the proxy `llm_guardrail`;
management's `CheckLLMPolicyLimits` handles only token/budget caps, so
no management change is required.

## Issue ticket number and link

<https://github.com/netbirdio/netbird/discussions/6751>

## Stack

- \#6726 <!-- branch-stack -->
  - \#6764 :point\_left:

### Checklist

- [x] Is it a bug fix
- [ ] Is a typo/documentation fix
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [x] Created tests that fail without the change (if possible)
- [x] This change does **not** modify the public API, gRPC protocols,
functionality behavior, CLI / service flags, or introduce a new feature
— **OR** I have discussed it with the NetBird team beforehand (link the
issue / Slack thread in the description). See
[CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).

> By submitting this pull request, you confirm that you have read and
agree to the terms of the [Contributor License
Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

## Documentation

Select exactly one:

- [ ] I added/updated documentation for this change
- [x] Documentation is **not needed** for this change (explain why)

Bug fix that restores the documented allowlist behavior; no user-facing
surface changes.

### Docs PR URL (required if "docs added" is checked)

Paste the PR link from <https://github.com/netbirdio/docs> here:

<https://github.com/netbirdio/docs/pull/>\_\_

## Tests

- `llm_guardrail`: absent/empty model under a configured allowlist now
denies (`model_unknown`); empty allowlist still allows a missing model
(fail-closed only applies when a list is set); existing
allow/deny/case-insensitive cases retained.
- `llm_request_parser`: new parser→guardrail integration test drives
real **Bedrock** (`/model/{id}/invoke`) and **Vertex**
(`/v1/projects/.../models/...`) URL shapes and asserts allowed→200,
disallowed→403 (`model_blocked`), and an unrecognized Bedrock action→403
(`model_unknown`, the #6751 regression guard).

Note: a full through-tunnel e2e for the allowlist is intentionally
deferred — the agent-network e2e (`WaitProxyPeer`) is currently red on
`main`/`0.74.x` for an unrelated lazy-connection reason; it will be
added once that harness gate is fixed.
2026-07-14 19:03:01 +02:00
David Fry
e70a69bbcf [client] Restore residual state in foreground mode before login (#6707)
* Improved residual state restoration during foreground startup and
foreground login, ensuring consistent recovery with stale states.
* Foreground flows now initialize advanced routing so stale routes 
are bypassed during login.
2026-07-14 17:43:59 +02:00
Riccardo Manfrin
a48618c074 [client] Fix forwarder peers never excluded from lazy connections (#6674)
* [client] Extract peerRoutesAddr helper in toExcludedLazyPeers

Refactor: pull the AllowedIPs match into a named
peerRoutesAddr helper and document why forward-target peers are excluded
from lazy connections. No behavior change; the existing address match is
preserved as-is.

* [client] Add failing test for lazy-conn forward-target exclusion

toExcludedLazyPeers compares AllowedIPs (CIDR) against the unmasked
TranslatedAddress, so forward-target peers are never excluded. This test
asserts the peer is excluded and fails on the current behavior; the fix
follows.

* [client] Fix lazy-conn exclusion for ingress forward peers

peerRoutesAddr compared AllowedIPs (CIDR, e.g. a peer's overlay IP as /32)
against the unmasked TranslatedAddress string, so the match never fired and
forward-target peers were never excluded from lazy connections. Use prefix
containment so a routed address matches the peer's AllowedIP

* [client] Reuse parsed AllowedIPs from peerStore in lazy exclusion

Instead of re-parsing the network map AllowedIPs strings, look up the
already-parsed []netip.Prefix from peerStore.AllowedIPs (the same typed
value the lazy manager itself consumes). A down/lazy peer still has its
conn in the store, so exclusion is unaffected by connection state. Extract
a pure prefixesContain helper and unit-test it.
2026-07-14 12:12:37 +02:00
Riccardo Manfrin
39193396f5 [client] Fix WGWatcher silently failing to restart on fast disconnect/reconnect (#6664)
* Stick new watcher creation to actual existence of af the conn

and its removal to the removal of such same conn.
Avoid debouncing and cross lock dead locking

* Discriminate not updated from timeout handshakes

* [Recheck watcher ctx cancellation under conn.mu in onWGDisconnected

onWGDisconnected only checked conn.ctx (the engine-scoped context), never
the watcher's own context. disableWgWatcherIfNeeded cancels the wgWatcherCtx,
not conn.ctx, so a disabled watcher's timeout callback did not see the
cancellation.

handshakeCheck runs lock-free, so between the ctx check in periodicHandshakeCheck
and acquiring conn.mu a fast disconnect/reconnect can slip in: the stale watcher
then acquires the lock and tears down the *new*, healthy connection based on the
old timeout, forcing the guard into an unnecessary reconnect (flap).

Recheck watcherCtx.Err() under conn.mu so a superseded watcher exits without
touching the connection that replaced it.

* Remove verbose comments

* Fixup merge conflict leftovers

* Fixup context brought by onWGDisconnected
2026-07-14 10:21:59 +02:00
Riccardo Manfrin
5343402385 [client, relay] Increase early-message buffer cap to 10000 to avoid dropping relayed handshakes 2026-07-13 22:46:03 +02:00
Pascal Fischer
62703ca23e [management] add logs to ephemeral delete (#6747) 2026-07-13 19:06:22 +02:00
Viktor Liu
cc64a93953 [client] Include system events in ToProtoFullStatus conversion (#6746) 2026-07-13 18:49:17 +02:00
Maycon Santos
831325d6e2 [management] require dashboard_features.agent_network when enabling agent_network_only (#6750)
Adds a settings constraint: enabling `agent_network_only` requires `dashboard_features.agent_network` to be `true` in the same account update. Without the Agent Network menu flag, a focused account that later turns the focused view off would lose access to the Agent Network menu entirely, so the two must be set together.

The check runs in `updateAccountRequestSettings` against the parsed request state: if the resulting settings have `agent_network_only == true` but `dashboard_features.agent_network` is not `true`, the update is rejected with `status.InvalidArgument` (HTTP 422) before anything is persisted.

The OpenAPI field descriptions for `agent_network_only` and `dashboard_features.agent_network` document the requirement. Only the descriptions changed — `required` and the schema `$ref` are untouched — and `types.gen.go` was regenerated from the spec (diff is the two comment lines).
2026-07-13 17:28:19 +02:00
Maycon Santos
8f64173574 [client] Enable launch-on-login by default on fresh GUI installs (#6738)
* [client] Add autostart preference marker and MDM disableAutostart key

Adds the autostartInitialized marker to the Wails UI preferences store so
the one-time autostart default decision can persist per OS user, and a
UI-only disableAutostart MDM policy key that suppresses the default and
flows into GetConfigResponse.mDMManagedFields like disableAutoConnect.

* [client] Enable launch-on-login by default on fresh GUI installs

On the first interactive run the GUI persists the autostartInitialized
marker before any enable attempt, then enables autostart only when the
platform supports it, MDM policy does not disable it, the process was not
relaunched by an installer/updater (--post-update), and the installer's
fresh-install breadcrumb is present. Upgrading users have no breadcrumb,
so an update can never write login items, and a user's disable in
Settings is never overridden.

* [release] Write fresh-install breadcrumb from installers

Installers write a .fresh-install breadcrumb on fresh installs only and
delete stale breadcrumbs on upgrade; none of them writes login items or
registry Run keys. Windows NSIS detects upgrades via the uninstall
registry entry or an existing installed executable; the macOS pkg via the
previous pkgutil receipt; Linux deb/rpm via the standard postinstall
arguments. Post-update GUI relaunches (macOS open, Linux
ui-post-install.sh) pass --post-update so the first-run autostart default
cannot fire on updates.

* Revert installer breadcrumb changes

The real Windows installer does uninstall-then-install and deletes
$INSTDIR, so a breadcrumb written there cannot survive or discriminate
a fresh install from an upgrade. Restore the three installer files to
their main versions; no installer or updater writes an autostart entry.

* Detect fresh install from NetBird footprint instead of installer breadcrumb

Replace the installer-written breadcrumb discriminator with a GUI-side
check. netbirdFootprintExists inspects the daemon config/state files
(default.json, legacy config.json, state.json) under profilemanager's
default config dir; combined with whether the UI preferences file already
existed, this tells a genuinely fresh machine from an existing or
upgrading user. Only the signed GUI, via Wails, ever enables
launch-on-login, and a user's later manual disable is never overridden.
The preferences store now exposes ExistedAtLoad and the --post-update
flag is dropped.

* Update tests for footprint-based autostart default

Table tests for shouldEnableAutostartDefault now cover supported,
mdmDisabled, and priorInstall guards plus precedence; breadcrumb and
post-update cases are removed. Add a store test asserting ExistedAtLoad
is false with no file and true after persisting and reopening.
2026-07-13 13:39:57 +02:00
Maycon Santos
76877e83c4 [client] Bring the connection up in Go after SSO login (#6744)
* [client] Bring the connection up in Go after SSO login

The post-login Up ran as a frontend promise continuation after WaitSSOLogin
resolved. During SSO the tray window is hidden and the webview is suspended
(macOS App Nap / hidden-window timer throttling), so that continuation didn't
run until the user woke the window (e.g. hovering the tray icon), leaving the
client not connected for a long time. Combine WaitSSOLogin and Up in a single
Go method so the daemon connects the moment SSO completes, independent of
webview state. The frontend no longer issues a separate Up on the SSO path.

* [client] unexport waitSSOLogin and move below exported methods
2026-07-13 13:32:47 +02:00
Maycon Santos
ecd398d895 [management] Add dashboard_features account setting (#6742)
Introduce a nullable dashboard_features object on account settings, serialized
to a single JSON column so new dashboard sections can be added without schema
changes. Starts with agent_network (show the Agent Network menu for an account
without the deployment flag). Wires the API handler mapping, the pgx GetAccount
loader, and adds store round-trip and handler tests.
2026-07-12 21:16:37 +02:00
Maycon Santos
aa92ad3fb1 [management] Add agent_network_only account setting (#6736)
* [management] Add agent_network_only account setting

* [management] Load agent_network_only in pgx account loader and cover persistence
2026-07-12 14:46:08 +02:00
Sufiyan Khan
fd94fdb42b [management] fix duplicate operationId in OpenAPI spec (#6734) 2026-07-12 14:15:55 +02:00
Maycon Santos
30d15ecc3d [client,management] sync 0.74.4 changes (#6727)
* [management] fix: prevent reverse proxy domain from being pushed as DNS search domain by @blaugrau90 in https://github.com/netbirdio/netbird/pull/6498
* [client] Recover from rosenpass key desync by @lixmal in https://github.com/netbirdio/netbird/pull/6714
* [client] Bump golang.org/x/crypto to v0.54.0 by @lixmal in https://github.com/netbirdio/netbird/pull/6709
* [client] fix MDM managementURL conflict on default-port URL echo by @riccardomanfrin in https://github.com/netbirdio/netbird/pull/6672
* [client] Update gopsutil to v4 by @mlsmaycon in https://github.com/netbirdio/netbird/pull/6688
* [client] Fix hanging status command during relay dial by @theodorsm in https://github.com/netbirdio/netbird/pull/6694

---------

Co-authored-by: Theodor Midtlien <theodor@midtlien.com>
Co-authored-by: blaugrau90 <61945343+blaugrau90@users.noreply.github.com>
Co-authored-by: Viktor Liu <17948409+lixmal@users.noreply.github.com>
2026-07-11 11:03:55 +02:00
Viktor Liu
3d87547d95 [client] Bump golang.org/x/crypto to v0.54.0 and Go toolchain to 1.25.12 (#6709) 2026-07-10 17:42:06 +02:00
Viktor Liu
4d4cc551fd [client] Recover from rosenpass key desync (#6714) 2026-07-10 17:38:29 +02:00
Maycon Santos
8e02154bf5 [client] Add SSO login flow timing instrumentation (#6717)
Users reported long delays between finishing browser authentication and
the client connecting. Logs could not attribute the time: the PKCE and
device flows were silent between issuing the auth URL and returning the
token, and nothing recorded when the GUI issued the Up request after
WaitSSOLogin completed.

Add log lines covering the full chain: PKCE callback arrival and token
exchange duration, device-flow polling and approval timing, GUI-side
brackets around WaitSSOLogin and Up, daemon-side Up arrival and
WaitSSOLogin return, and a frontend stall detector that reports when
webview timers were suspended (macOS App Nap / hidden-window
throttling), which delays the WaitSSOLogin-to-Up handoff.
2026-07-10 16:11:27 +02:00
blaugrau90
08e46aa62f [management] fix: prevent reverse proxy domain from being pushed as DNS search domain (#6498)
SynthesizePrivateServiceZones created CustomZones for private services
without setting SearchDomainDisabled, causing the reverse proxy domain
to be injected as a search domain suffix on all connected peers.

This broke local hostname resolution: short names like 'myserver' were
expanded to 'myserver.app.example.com' (matching the reverse proxy
domain) before local DNS search domains were tried.

Fix: set SearchDomainDisabled: true so the zone is registered as a
match-only supplemental resolver, consistent with the NonAuthoritative
intent already expressed on the same zone.
2026-07-10 12:20:57 +02:00
dmitri-netbird
e0c25ba4ba [client] fix flaky test around event aggregation (#6710)
* fix flaky test around event aggregation: control time.Now() from the test

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* actually use passed in func to generate time

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

---------

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>
2026-07-09 18:17:28 +02:00
Pascal Fischer
2560c6bd6c [management] add traffic filters for source and dest id (#6697) 2026-07-09 14:37:31 +02:00
Viktor Liu
96ac15d292 [client] Fix js relay WebSocket close, raise RDP dial timeout, adjust WASM log levels (#6684) 2026-07-09 13:21:08 +02:00
Misha Bragin
488bbcb22b [doc] Update Agent Network Readme (#6699) 2026-07-08 17:53:55 +02:00
Theodor Midtlien
b7bbb44286 [client] Merge v0.74.x branch (#6700)
* [client] Update gopsutil to v4 (#6688)
* [client] Fix hanging status command during relay dial (#6694)

---------

Co-authored-by: Maycon Santos <mlsmaycon@gmail.com>
2026-07-08 17:52:50 +02:00
Viktor Liu
58a8a69406 Merge remote-tracking branch 'origin/main' into peer-acl-multi-source 2026-07-08 17:20:14 +02:00
Nicolas Frati
58318481e6 [client] fix: re-generate gateway proto files (#6696) 2026-07-08 16:35:37 +02:00
Theodor Midtlien
7cd5c1732b [client] Fix hanging status command during relay dial (#6694)
* Add regression test for relay state lock
* Make connect not hold a lock in openConnVia
2026-07-08 14:36:42 +02:00
Viktor Liu
f13b7738ca Merge remote-tracking branch 'origin/main' into peer-acl-multi-source
# Conflicts:
#	client/internal/acl/manager.go
#	client/ssh/server/test.go
2026-07-08 14:10:58 +02:00
Maycon Santos
816d80602f [client] Update gopsutil to v4 (#6688) 2026-07-08 10:15:31 +02:00
Nicolas Frati
d0d6dd4b0c [client] add json gateway for netbird daemon (#6272)
* add json gateway for netbird daemon

* client: add optional daemon JSON socket gateway
2026-07-07 16:58:48 +02:00
dmitri-netbird
47352e6e45 [client] introduce client-side event aggregation (#6627)
* added an implementation of aggregating memory store

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* initial support for aggregation of events

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* added tcp-aggregation test

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* added manager integration test

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* added tracking of the number of start-, drop, and end-events in an aggregation window

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* fixes based on sonarcube checks

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* regenerated proto files

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* removed inadvertenly added google proto files

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* pacifying linter

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* update test to validate event aggregation over tcp, udp, icmp, and icmpv6

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* updated event aggregation test

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* regenerate protobufs with expected versions of protoc and protoc-gen-go

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* remove protoc/protoc-gen headers from flow_grpc.pb.go

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* updated openapi spec

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* updated openapi NetworkTrafficEvent spec, regenerated types

Signed-off-by: Dmitri <dmitri.external@netbird.io>

* respond to feedback

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* fixed an issue with how we track events that shouldn't be aggregated

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* fixed mapping of events to protobuf

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* icmp code values in aggregated events do not matter

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* regenerate openapi types

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* added a comment re: unbounded unacked events

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* reset aggregated event type to unknown

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* fix event aggregation test

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* used the source port of the earliest event

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* add tracking of window starts and ends

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* updated openapi spec

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* reverted changes to generate.sh

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* cleanup handling of not-aggregated events + test

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* responded to feedback + small fixes

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* small fix in a test

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* another test

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* force setting non-empty rule id on aggregated events

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* fixed a couple of issues flagged by coderabbit

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* fix spelling

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

* handle exhausted retry backoffs

Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>

---------

Signed-off-by: Dmitri <dmitri.external@netbird.io>
Signed-off-by: Dmitri Dolguikh <dmitri.external@netbird.io>
2026-07-06 16:04:26 +02:00
Maycon Santos
91acb8147c [management,client] 0.75.0 release with new desktop UI (#6473)
- **Wails v3 application** (`client/ui`) with a React + TypeScript + Tailwind frontend replacing the Fyne UI: main connection view, exit-node switcher, networks/peers browser with detail panels, profile management, settings (general, network, SSH, security, troubleshooting, appearance), debug-bundle creation, and a first-run welcome flow.
- **Internationalization**: go-i18n bundle with 9 locales (en, de, es, fr, hu, it, pt, ru, zh-CN) shared between the tray and the frontend.
- **New system tray** implementation with per-platform theme-aware icons, including a native XEmbed host for Linux (`xembed_tray_linux.c`) and a Linux theme watcher.
- **Session handling**: auth session watcher (`client/internal/auth/sessionwatch`), pending login flow, session-expiration dialog and tray notifications, and `netbird login` improvements.
- **Daemon API extensions** (`daemon.proto`): status stream subscription, event stream, networks/exit-node selection endpoints, and richer full status — with probe throttling on the daemon side to protect against UI-driven request storms.
- **UI preferences store** persisted per profile, autostart management via the daemon (single source of truth in HKCU on Windows).
- **Build system**: Taskfile-based builds per platform (macOS, Linux, Windows), Docker cross-compilation images, MSIX/NSIS/nfpm/AppImage packaging, and a new `frontend-ui` CI workflow.

Co-authored-by: Zoltan Papp <zoltan.pmail@gmail.com>
Co-authored-by: Eduard Gert <kontakt@eduardgert.de>
Co-authored-by: braginini <bangvalo@gmail.com>
Co-authored-by: Pascal Fischer <32096965+pascal-fischer@users.noreply.github.com>
Co-authored-by: riccardom <riccardomanfrin@gmail.com>
2026-07-06 13:47:16 +02:00
Riccardo Manfrin
c9d387bd0d [client] fix MDM managementURL conflict on default-port URL echo (#6672)
* Adds failing test

* Fixes Management URL normalized compare on MDM
2026-07-06 11:39:20 +02:00
Theodor Midtlien
3aa6c02b93 [client] Fix backoff.Ticker goroutine leak in reconnect guard 2026-07-03 12:23:11 +02:00
Zoltan Papp
f6900fb07c [client] backport enforce a single selected exit node (#6640)
* routemanager: enforce a single selected exit node

Backport of the exit-node exclusivity reconcile from the 0.75.0 line
(upstream commit 966fbec11) onto v0.74.0. Exit nodes are mutually
exclusive, but the RouteSelector stores routes with default-on semantics,
so every available exit node reported as selected at once.

Reconcile exit-node selection on each network map: keep at most one
selected -- the user's persisted pick, else whatever management marks for
auto-apply (SkipAutoApply=false), else none. Never auto-activate an exit
node the map does not request.

Carries over only the manager/routeselector logic and its test; the
desktop-only client/server changes and the BumpNetworksRevision UI-push
feature from the original commit are intentionally excluded.

* routeselector: make exit-node reconciliation atomic

enforceSingleExitNode took the RouteSelector lock three separate times
(IsDeselectAll, then DeselectRoutes, then SelectRoutes), so a concurrent
DeselectAllRoutes could interleave and be silently undone: SelectRoutes on
its deselectAll branch clears the flag and re-selects the preferred exit
node, overriding the user's "all off".

Move the whole reconciliation into a single locked RouteSelector method
(SetExclusiveExitNode) that checks deselectAll inside the critical section,
so a deselect-all either fully precedes the reconcile (left untouched) or
fully follows it (honoured). No interleaving is possible.
2026-07-03 10:31:06 +02:00
Zoltan Papp
4b3dd9103d [client] Fix slow wg operations (#6633)
* [iface] Drop redundant device dump in kernel configure()

wgctrl.ConfigureDevice already returns an error when the interface is
missing, so the preceding wg.Device() existence check is redundant. That
check dumps the entire device (all peers) on every configure() call,
making it O(peers) per call and turning bulk peer insertion into
O(peers^2): inserting N peers one by one re-parsed the whole growing peer
list N times. Removing it keeps each peer write constant-time regardless
of how many peers are already configured.

* [iface] Cache WireGuard stats to collapse per-peer device dumps

Each peer runs a WGWatcher that polls GetStats(), and every call dumps
the whole device, so with N peers the watchers perform O(N) full dumps
per poll cycle (O(N^2) work) while each keeps only its own peer's entry.

Wrap the kernel and userspace configurer GetStats() in a short-TTL cache
with singleflight: the staggered per-peer calls share a single device
dump per window and concurrent misses collapse into one dump. The kernel
and userspace WireGuard APIs have no per-peer stats query (a get always
returns the whole device), so a shared cached snapshot avoids the
repeated full dumps.

* Ignore .claude directory
2026-07-02 20:42:43 +02:00
Riccardo Manfrin
8e3b284f4b [client] Increase mgmt grpc buff size to 16MB (#6641) 2026-07-02 17:50:18 +02:00
Maycon Santos
21aa933584 [misc] Fix GHCR image push after dockers_v2 migration (#6653) 2026-07-02 17:21:06 +02:00
Misha Bragin
1dfa85a917 [management] Add vLLM e2e test (#6649)
* Add vLLM to Agent Network

* Add vllm e2e test
2026-07-02 15:36:51 +02:00
Maycon Santos
859fe19fff [management] return nil when config is not set (#6642)
* [management] return nil when config is not set

* [management] add relay invariant test and enforce config behavior
2026-07-02 14:55:55 +02:00
Misha Bragin
e40cb294f6 [management] Add vLLM to Agent Network (#6643) 2026-07-02 14:45:24 +02:00
Maycon Santos
e203e0f42a [self-hosted] Remove unused server/proxy image override logic in getting-started.sh (#6636) 2026-07-02 14:20:23 +02:00
Viktor Liu
959cfa1804 Merge remote-tracking branch 'origin/main' into peer-acl-multi-source
# Conflicts:
#	.github/workflows/golang-test-linux.yml
#	client/firewall/iptables/manager_linux_test.go
#	client/firewall/iptables/router_linux_test.go
#	client/firewall/nftables/manager_linux_test.go
#	client/firewall/nftables/router_linux_test.go
2026-06-28 17:57:32 +02:00
Viktor Liu
97d9559e6d Fix CI: recognize multi-value port set lookups in tests and correct PeerIP lint suppression 2026-06-10 21:11:43 +02:00
Viktor Liu
ab7639d101 Fix legacy ACL source wildcard and keep rollback tracking on delete failure 2026-06-10 17:44:33 +02:00
Viktor Liu
a5d4373ddc Share the iptables jump rule spec between install and cleanup 2026-06-10 17:44:33 +02:00
Viktor Liu
1b6294f2ff Install nftables peer filter and mangle rules in a single transaction 2026-06-10 17:44:33 +02:00
Viktor Liu
a14586b142 Scope userspace firewall wildcard source rules per address family 2026-06-10 17:44:23 +02:00
Viktor Liu
09c0063d71 Merge branch 'main' into peer-acl-multi-source
# Conflicts:
#	client/firewall/iptables/acl_linux.go
#	client/firewall/iptables/router_linux.go
#	client/firewall/iptables/rulestore_linux.go
2026-06-10 12:29:17 +02:00
Viktor Liu
1c5b84b1a1 Snapshot iptables rule maps before persisting state 2026-06-05 17:07:33 +02:00
Viktor Liu
1ed2067b8b Merge remote-tracking branch 'origin/main' into peer-acl-multi-source
# Conflicts:
#	.github/workflows/golang-test-linux.yml
2026-06-03 22:01:41 +02:00
Viktor Liu
f7e9df6ffa Remove partial userspace firewall mode and open foreign chains via a table-less allower 2026-06-03 22:01:22 +02:00
Viktor Liu
38603c7552 Unify peer and route ACL filtering with multi-source peer rules 2026-06-02 10:50:53 +02:00
888 changed files with 87929 additions and 23201 deletions

18
.coderabbit.yaml Normal file
View File

@@ -0,0 +1,18 @@
# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json
language: en-US
reviews:
profile: chill
request_changes_workflow: false
high_level_summary: true
poem: false
review_status: true
auto_review:
enabled: true
drafts: false
path_filters:
- "!**/*.tsx"
- "!**/*.ts"
- "!**/*.js"
- "!**/*.svg"
chat:
auto_reply: true

View File

@@ -6,7 +6,6 @@ RUN apt-get update && export DEBIAN_FRONTEND=noninteractive \
iptables=1.8.9-2 \
libgl1-mesa-dev=22.3.6-1+deb12u1 \
xorg-dev=1:7.7+23 \
libayatana-appindicator3-dev=0.5.92-1 \
&& apt-get clean \
&& rm -rf /var/lib/apt/lists/* \
&& go install -v golang.org/x/tools/gopls@latest

View File

@@ -3,8 +3,8 @@ updates:
- package-ecosystem: "github-actions"
directory: "/"
schedule:
interval: "daily"
open-pull-requests-limit: 15
interval: "weekly"
open-pull-requests-limit: 3
groups:
actions:
patterns:
@@ -22,9 +22,12 @@ updates:
directories:
- "/"
schedule:
interval: "daily"
interval: "weekly"
open-pull-requests-limit: 15
groups:
golang-x-packages:
patterns:
- "golang.org/x/*"
aws-sdk:
patterns:
- "github.com/aws/aws-sdk-go-v2/*"

View File

@@ -2,6 +2,12 @@
## Issue ticket number and link
<!--
Required for anything that changes behavior. Link the issue (or the validated
discussion it came from) that the NetBird team already agreed on. See
https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#ticket-first-pr-second
-->
## Stack
<!-- branch-stack -->
@@ -12,7 +18,9 @@
- [ ] Is a feature enhancement
- [ ] It is a refactor
- [ ] Created tests that fail without the change (if possible)
- [ ] This change does **not** modify the public API, gRPC protocols, functionality behavior, CLI / service flags, or introduce a new feature — **OR** I have discussed it with the NetBird team beforehand (link the issue / Slack thread in the description). See [CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#discuss-changes-with-the-netbird-team-first).
- [ ] I ran and tested this change locally — I did not rely on CI to find out whether it works
- [ ] This PR has a single purpose (not a fix + refactor + feature in one)
- [ ] This change is a trivial fix, **OR** it links an issue the NetBird team agreed on beforehand. Changes to the public API, gRPC protocols, functionality behavior, CLI / service flags, or new features always need that agreement first. See [CONTRIBUTING.md](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTING.md#ticket-first-pr-second).
> By submitting this pull request, you confirm that you have read and agree to the terms of the [Contributor License Agreement](https://github.com/netbirdio/netbird/blob/main/CONTRIBUTOR_LICENSE_AGREEMENT.md).

View File

@@ -5,6 +5,13 @@ on:
schedule:
- cron: "0 3 * * *"
workflow_dispatch:
inputs:
bedrock_model:
description: >-
Bedrock inference-profile id to drive the matrix with, exactly as
AWS issues it. Leave empty for the Sonnet 4.6 default.
required: false
default: ""
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
@@ -51,6 +58,9 @@ jobs:
# token (and URL, for gateways) is unset, so partial coverage is fine.
OPENAI_TOKEN: ${{ secrets.E2E_OPENAI_TOKEN }}
ANTHROPIC_TOKEN: ${{ secrets.E2E_ANTHROPIC_TOKEN }}
# Moonshot AI platform key (platform.kimi.ai); drives both Kimi wire
# shapes (OpenAI /v1 and Anthropic /anthropic) through kimi_api.
KIMI_TOKEN: ${{ secrets.E2E_KIMI_TOKEN }}
VERCEL_URL: ${{ secrets.E2E_VERCEL_URL }}
VERCEL_TOKEN: ${{ secrets.E2E_VERCEL_TOKEN }}
OPENROUTER_URL: ${{ secrets.E2E_OPENROUTER_URL }}
@@ -59,6 +69,8 @@ jobs:
CLOUDFLARE_TOKEN: ${{ secrets.E2E_CLOUDFLARE_TOKEN }}
AWS_BEARER_TOKEN_BEDROCK: ${{ secrets.E2E_AWS_BEARER_TOKEN_BEDROCK }}
AWS_REGION: ${{ secrets.E2E_AWS_REGION }}
# Bedrock model override: dispatch input wins, then the repo variable, else the test default.
AWS_BEDROCK_MODEL: ${{ inputs.bedrock_model || vars.E2E_AWS_BEDROCK_MODEL }}
# Vertex (Anthropic-on-Vertex): SA + project required; region defaults
# to "global", model to a pinned claude snapshot.
GOOGLE_VERTEX_SA_BASE64: ${{ secrets.E2E_GOOGLE_VERTEX_SA_BASE64 }}

98
.github/workflows/frontend-ui.yml vendored Normal file
View File

@@ -0,0 +1,98 @@
name: UI Frontend
on:
pull_request:
paths:
- "client/ui/frontend/**"
- "client/ui/i18n/**"
- "client/ui/**/*.go"
- ".github/workflows/frontend-ui.yml"
push:
branches:
- main
paths:
- "client/ui/frontend/**"
- "client/ui/i18n/**"
- "client/ui/**/*.go"
permissions:
contents: read
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}-${{ github.head_ref || github.actor_id }}
cancel-in-progress: true
jobs:
lint-and-build:
name: Lint & Build
runs-on: ubuntu-latest
timeout-minutes: 15
defaults:
run:
working-directory: client/ui/frontend
steps:
- name: Checkout repository
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
with:
persist-credentials: false
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: "22"
- name: Set up pnpm
uses: pnpm/action-setup@a3252b78c470c02df07e9d59298aecedc3ccdd6d # v3.0.0
with:
version: 11
# Bindings are generated by wails3 from the Go service definitions and
# are not checked in (see client/ui/frontend/bindings/). Without them,
# typecheck/build fail on missing module imports.
- name: Set up Go
uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
with:
go-version-file: "go.mod"
cache: false
# wails3 CLI links against GTK4 / WebKitGTK 6.0 via its internal/operatingsystem
# package, so the dev libraries must be present before `go install`.
- name: Install Wails Linux system dependencies
run: |
sudo apt-get update
sudo apt-get install -y --no-install-recommends \
pkg-config \
libgtk-4-dev \
libwebkitgtk-6.0-dev
- name: Install wails3 CLI
# Version derived from go.mod so the binding generator always matches
# the wails runtime the daemon links against.
working-directory: ${{ github.workspace }}
run: |
WAILS_VERSION=$(go list -m -f '{{.Version}}' github.com/wailsapp/wails/v3)
go install github.com/wailsapp/wails/v3/cmd/wails3@$WAILS_VERSION
- name: Get pnpm store directory
id: pnpm-store
run: echo "path=$(pnpm store path --silent)" >> "$GITHUB_OUTPUT"
- name: Cache pnpm store
uses: actions/cache@v4
with:
path: ${{ steps.pnpm-store.outputs.path }}
key: ${{ runner.os }}-pnpm-${{ hashFiles('client/ui/frontend/pnpm-lock.yaml') }}
restore-keys: |
${{ runner.os }}-pnpm-
- name: Install dependencies
run: pnpm install --frozen-lockfile --ignore-scripts
- name: Generate Wails bindings
run: pnpm run bindings
- name: Lint, typecheck, format
run: pnpm check
- name: Build
run: pnpm build

View File

@@ -45,7 +45,15 @@ jobs:
run: git --no-pager diff --exit-code
- name: Test
run: NETBIRD_STORE_ENGINE=${{ matrix.store }} CI=true go test -coverprofile=coverage.txt -tags 'devcert privileged' -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE' -timeout 5m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/testutil/privileged)
# Exclude client/ui: its main.go uses //go:embed all:frontend/dist,
# which fails to compile until the frontend has been built. The Wails UI
# has no Go-side unit tests, and its release pipeline runs `pnpm build`
# before goreleaser.
# `go list -e` lets the listing succeed even though the embed fails to
# resolve; the grep then drops the broken package by path. Without -e,
# go list aborts with empty stdout and `go test` falls back to the repo
# root, which has no Go files.
run: NETBIRD_STORE_ENGINE=${{ matrix.store }} CI=true go test -coverprofile=coverage.txt -tags 'devcert privileged' -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE' -timeout 5m -p 1 $(go list -e ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /client/testutil/privileged)
- name: Upload coverage reports to Codecov
uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0

View File

@@ -53,7 +53,7 @@ jobs:
- name: Install dependencies
if: steps.cache.outputs.cache-hit != 'true'
run: sudo apt update && sudo apt install -y -q libgtk-3-dev libayatana-appindicator3-dev libgl1-mesa-dev xorg-dev gcc-multilib libpcap-dev
run: sudo apt update && sudo apt install -y -q libgtk-4-dev libwebkitgtk-6.0-dev libsoup-3.0-dev libgl1-mesa-dev xorg-dev gcc-multilib libpcap-dev
- name: Install 32-bit libpcap
if: steps.cache.outputs.cache-hit != 'true'
@@ -145,7 +145,7 @@ jobs:
${{ runner.os }}-gotest-cache-
- name: Install dependencies
run: sudo apt update && sudo apt install -y -q libgtk-3-dev libayatana-appindicator3-dev libgl1-mesa-dev xorg-dev gcc-multilib libpcap-dev
run: sudo apt update && sudo apt install -y -q libgtk-4-dev libwebkitgtk-6.0-dev libsoup-3.0-dev libgl1-mesa-dev xorg-dev gcc-multilib libpcap-dev
- name: Install 32-bit libpcap
if: matrix.arch == '386'
@@ -158,7 +158,15 @@ jobs:
run: git --no-pager diff --exit-code
- name: Test
run: CGO_ENABLED=1 GOARCH=${{ matrix.arch }} CI=true go test -coverprofile=coverage.txt -tags 'devcert privileged' -exec 'sudo --preserve-env=CI,CGO_ENABLED' -timeout 10m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/testutil/privileged)
# Exclude client/ui: its main.go uses //go:embed all:frontend/dist,
# which fails to compile until the frontend has been built. The Wails UI
# has no Go-side unit tests, and its release pipeline runs `pnpm build`
# before goreleaser.
# `go list -e` lets the listing succeed even though the embed fails to
# resolve; the grep then drops the broken package by path. Without -e,
# go list aborts with empty stdout and `go test` falls back to the repo
# root, which has no Go files.
run: CGO_ENABLED=1 GOARCH=${{ matrix.arch }} CI=true go test -coverprofile=coverage.txt -tags 'devcert privileged' -exec 'sudo --preserve-env=CI,CGO_ENABLED' -timeout 10m -p 1 $(go list -e ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /client/testutil/privileged)
- name: Upload coverage reports to Codecov
if: matrix.arch == 'amd64'
@@ -168,7 +176,6 @@ jobs:
slug: netbirdio/netbird
flags: unit,client
test_client_on_docker:
name: "Client (Docker) / Unit"
needs: [build-cache]
@@ -229,7 +236,7 @@ jobs:
sh -c ' \
apk update; apk add --no-cache \
ca-certificates iptables ip6tables dbus dbus-dev libpcap-dev build-base; \
go test -buildvcs=false -tags "devcert privileged" -v -timeout 10m -p 1 $(go list -buildvcs=false ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /upload-server -e /client/testutil/privileged)
go test -buildvcs=false -tags "devcert privileged" -v -timeout 10m -p 1 $(go list -e -buildvcs=false ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /upload-server -e /client/testutil/privileged)
'
test_relay:

View File

@@ -65,8 +65,15 @@ jobs:
- run: PsExec64 -s -w ${{ github.workspace }} C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe env -w GOCACHE=${{ env.modcache }}
- run: PsExec64 -s -w ${{ github.workspace }} C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe mod tidy
- name: Generate test script
# Exclude client/ui: its main.go uses //go:embed all:frontend/dist,
# which fails to compile until the frontend has been built. The Wails UI
# has no Go-side unit tests, and its release pipeline runs `pnpm build`
# before goreleaser.
# `go list -e` lets the listing succeed even though the embed fails to
# resolve; the Where-Object pipeline then drops the broken package by
# path. Without -e, go list aborts with empty stdout.
run: |
$packages = go list ./... | Where-Object { $_ -notmatch '/management' } | Where-Object { $_ -notmatch '/relay' } | Where-Object { $_ -notmatch '/signal' } | Where-Object { $_ -notmatch '/proxy' } | Where-Object { $_ -notmatch '/combined' }
$packages = go list -e ./... | Where-Object { $_ -notmatch '/management' } | Where-Object { $_ -notmatch '/relay' } | Where-Object { $_ -notmatch '/signal' } | Where-Object { $_ -notmatch '/proxy' } | Where-Object { $_ -notmatch '/combined' } | Where-Object { $_ -notmatch '/client/ui' }
$goExe = "C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe"
$cmd = "$goExe test -tags `"devcert privileged`" -timeout 10m -p 1 $($packages -join ' ') > test-out.txt 2>&1"
Set-Content -Path "${{ github.workspace }}\run-tests.cmd" -Value $cmd

View File

@@ -22,7 +22,15 @@ jobs:
uses: codespell-project/actions-codespell@8f01853be192eb0f849a5c7d721450e7a467c579 # v2.2
with:
ignore_words_list: erro,clienta,hastable,iif,groupd,testin,groupe,cros,ans,deriver,te,userA,ede,additionals,flate,recordin,unparseable
skip: go.mod,go.sum,**/proxy/web/**
# Non-English UI translations trip codespell on real foreign words
# (de: "Sie", "oder", "ist"). Only en/common.json is the source of
# truth that should be spell-checked. List each translated locale
# dir below and add new ones as languages are added under
# client/ui/i18n/locales/. Single-star globs are matched per path
# segment by codespell and behave the same across versions; the
# recursive "**" form did not take effect with the codespell shipped
# by this action.
skip: go.mod,go.sum,*/proxy/web/*,*pnpm-lock.yaml,*package-lock.json,*/locales/de/*,*/locales/es/*,*/locales/fr/*,*/locales/hu/*,*/locales/it/*,*/locales/pt/*,*/locales/ru/*,*/locales/zh-CN/*,*/i18n/TRANSLATING.md
golangci:
strategy:
fail-fast: false
@@ -54,7 +62,16 @@ jobs:
cache: false
- name: Install dependencies
if: matrix.os == 'ubuntu-latest'
run: sudo apt update && sudo apt install -y -q libgtk-3-dev libayatana-appindicator3-dev libgl1-mesa-dev xorg-dev libpcap-dev
run: sudo apt update && sudo apt install -y -q libgtk-4-dev libwebkitgtk-6.0-dev libsoup-3.0-dev libgl1-mesa-dev xorg-dev libpcap-dev
- name: Stub Wails frontend bundle
# client/ui/main.go has //go:embed all:frontend/dist. The
# directory is produced by `pnpm run build` and is gitignored, so
# lint-only runs (no frontend toolchain) need a placeholder file
# for the embed pattern to match.
shell: bash
run: |
mkdir -p client/ui/frontend/dist
touch client/ui/frontend/dist/.embed-placeholder
- name: golangci-lint
uses: golangci/golangci-lint-action@82606bf257cbaff209d206a39f5134f0cfbfd2ee #v9.2.1
with:

View File

@@ -9,7 +9,7 @@ on:
pull_request:
env:
SIGN_PIPE_VER: "v0.1.6"
SIGN_PIPE_VER: "v0.1.8"
GORELEASER_VER: "v2.16.0"
PRODUCT_NAME: "NetBird"
COPYRIGHT: "NetBird GmbH"
@@ -216,9 +216,9 @@ jobs:
- name: Install goversioninfo
run: go install github.com/josephspurrier/goversioninfo/cmd/goversioninfo@233067e
- name: Generate windows syso amd64
run: goversioninfo -icon client/ui/assets/netbird.ico -manifest client/manifest.xml -product-name ${{ env.PRODUCT_NAME }} -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/resources_windows_amd64.syso
run: goversioninfo -icon client/ui/build/windows/icon.ico -manifest client/manifest.xml -product-name ${{ env.PRODUCT_NAME }} -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/resources_windows_amd64.syso
- name: Generate windows syso arm64
run: goversioninfo -arm -64 -icon client/ui/assets/netbird.ico -manifest client/manifest.xml -product-name ${{ env.PRODUCT_NAME }} -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/resources_windows_arm64.syso
run: goversioninfo -arm -64 -icon client/ui/build/windows/icon.ico -manifest client/manifest.xml -product-name ${{ env.PRODUCT_NAME }} -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/resources_windows_arm64.syso
- name: Run GoReleaser
id: goreleaser
uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
@@ -293,8 +293,11 @@ jobs:
${{ steps.goreleaser.outputs.artifacts }}
JSON
# dockers_v2 artifacts have no top-level goarch field, so match the
# per-platform -amd64 tag suffix instead; it works for both the old
# dockers and the new dockers_v2 image naming.
mapfile -t src_images < <(
jq -r '.[] | select(.type == "Docker Image") | select(.goarch == "amd64") | .name | select(startswith("ghcr.io/"))' /tmp/goreleaser-artifacts.json
jq -r '.[] | select(.type == "Docker Image") | .name | select(startswith("ghcr.io/") and endswith("-amd64"))' /tmp/goreleaser-artifacts.json
)
for src in "${src_images[@]}"; do
@@ -394,8 +397,18 @@ jobs:
- name: check git status
run: git --no-pager diff --exit-code
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: '22'
- name: Set up pnpm
uses: pnpm/action-setup@a3252b78c470c02df07e9d59298aecedc3ccdd6d # v3.0.0
with:
version: 11
- name: Install dependencies
run: sudo apt update && sudo apt install -y -q libappindicator3-dev gir1.2-appindicator3-0.1 libxxf86vm-dev gcc-mingw-w64-x86-64
run: sudo apt update && sudo apt install -y -q libgtk-4-dev libwebkitgtk-6.0-dev libsoup-3.0-dev gcc-mingw-w64-x86-64
- name: Decode GPG signing key
if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository
@@ -414,10 +427,16 @@ jobs:
echo "/tmp/llvm-mingw-20250709-ucrt-ubuntu-22.04-x86_64/bin" >> $GITHUB_PATH
- name: Install goversioninfo
run: go install github.com/josephspurrier/goversioninfo/cmd/goversioninfo@233067e
- name: Install wails3 CLI
# Version derived from go.mod so the binding generator always matches
# the wails runtime the binary links against.
run: |
WAILS_VERSION=$(go list -m -f '{{.Version}}' github.com/wailsapp/wails/v3)
go install github.com/wailsapp/wails/v3/cmd/wails3@$WAILS_VERSION
- name: Generate windows syso amd64
run: goversioninfo -64 -icon client/ui/assets/netbird.ico -manifest client/ui/manifest.xml -product-name ${{ env.PRODUCT_NAME }}-"UI" -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/ui/resources_windows_amd64.syso
run: goversioninfo -64 -icon client/ui/build/windows/icon.ico -manifest client/ui/build/windows/wails.exe.manifest -product-name ${{ env.PRODUCT_NAME }}-"UI" -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/ui/resources_windows_amd64.syso
- name: Generate windows syso arm64
run: goversioninfo -arm -64 -icon client/ui/assets/netbird.ico -manifest client/ui/manifest.xml -product-name ${{ env.PRODUCT_NAME }}-"UI" -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/ui/resources_windows_arm64.syso
run: goversioninfo -arm -64 -icon client/ui/build/windows/icon.ico -manifest client/ui/build/windows/wails.exe.manifest -product-name ${{ env.PRODUCT_NAME }}-"UI" -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/ui/resources_windows_arm64.syso
- name: Run GoReleaser
uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
@@ -486,6 +505,20 @@ jobs:
run: go mod tidy
- name: check git status
run: git --no-pager diff --exit-code
- name: Set up Node.js
uses: actions/setup-node@v4
with:
node-version: '22'
- name: Set up pnpm
uses: pnpm/action-setup@a3252b78c470c02df07e9d59298aecedc3ccdd6d # v3.0.0
with:
version: 11
- name: Install wails3 CLI
# Version derived from go.mod so the binding generator always matches
# the wails runtime the binary links against.
run: |
WAILS_VERSION=$(go list -m -f '{{.Version}}' github.com/wailsapp/wails/v3)
go install github.com/wailsapp/wails/v3/cmd/wails3@$WAILS_VERSION
- name: Run GoReleaser
id: goreleaser
uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
@@ -573,23 +606,6 @@ jobs:
- name: Move wintun.dll into dist
run: mv ${{ env.downloadPath }}\wintun\bin\${{ matrix.wintun_arch }}\wintun.dll ${{ github.workspace }}\dist\${{ env.PackageWorkdir }}\
- name: Download Mesa3D (amd64 only)
id: download-mesa3d
if: matrix.arch == 'amd64'
uses: netbirdio/shared-actions/actions/win-download-and-verify@be5df6047383da2236e02243cceb857d8567c27e # v0.0.2
with:
url: https://pkgs.netbird.io/mesa3d/MesaForWindows-x64-20.1.8.7z
destination: ${{ env.downloadPath }}\mesa3d.7z
sha256: 71c7cb64ec229a1d6b8d62fa08e1889ed2bd17c0eeede8689daf0f25cb31d6b9
- name: Extract Mesa3D driver (amd64 only)
if: matrix.arch == 'amd64'
run: 7z x -o"${{ env.downloadPath }}" "${{ env.downloadPath }}/mesa3d.7z"
- name: Move opengl32.dll into dist (amd64 only)
if: matrix.arch == 'amd64'
run: mv ${{ env.downloadPath }}\opengl32.dll ${{ github.workspace }}\dist\${{ env.PackageWorkdir }}\
- name: Download EnVar plugin for NSIS
uses: netbirdio/shared-actions/actions/win-download-and-verify@be5df6047383da2236e02243cceb857d8567c27e # v0.0.2
with:
@@ -612,6 +628,28 @@ jobs:
if: matrix.arch == 'amd64'
run: 7z x -o"${{ github.workspace }}/NSIS_Plugins" "${{ github.workspace }}/ShellExecAsUser_amd64-Unicode.7z"
- name: Set up Go for wails3 CLI
uses: actions/setup-go@v5
with:
go-version-file: "go.mod"
cache: false
- name: Install wails3 CLI
# Version derived from go.mod so the bootstrapper payload always
# matches the wails runtime the binary links against.
shell: bash
run: |
WAILS_VERSION=$(go list -m -f '{{.Version}}' github.com/wailsapp/wails/v3)
go install github.com/wailsapp/wails/v3/cmd/wails3@$WAILS_VERSION
- name: Stage WebView2 bootstrapper for installers
# Both client/installer.nsis and client/netbird.wxs reference
# client/MicrosoftEdgeWebview2Setup.exe. wails3 writes it there.
# The signing pipeline (netbirdio/sign-pipelines) does the same
# step for release builds; this mirrors it for PR sanity testing.
shell: bash
run: wails3 generate webview2bootstrapper -dir client
- name: Build NSIS installer
shell: pwsh
env:

View File

@@ -249,78 +249,35 @@ jobs:
docker compose exec management ls -l /var/lib/netbird/ | grep -i GeoLite2-City_[0-9]*.mmdb
docker compose exec management ls -l /var/lib/netbird/ | grep -i geonames_[0-9]*.db
test-getting-started-script:
test-legacy-getting-started-scripts:
runs-on: ubuntu-latest
steps:
- name: Install jq
run: sudo apt-get install -y jq
- name: Checkout code
uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
with:
persist-credentials: false
- name: run script with Zitadel PostgreSQL
run: NETBIRD_DOMAIN=use-ip bash -x infrastructure_files/getting-started-with-zitadel.sh
- name: test Caddy file gen postgres
run: test -f Caddyfile
- name: test docker-compose file gen postgres
run: test -f docker-compose.yml
- name: test management.json file gen postgres
run: test -f management.json
- name: test turnserver.conf file gen postgres
- name: Verify Dex retirement notice
run: |
set -x
test -f turnserver.conf
grep external-ip turnserver.conf
if infrastructure_files/getting-started-with-dex.sh >stdout.txt 2>stderr.txt; then
echo "Expected the retired Dex installer to fail"
exit 1
fi
test ! -s stdout.txt
grep -Fq "Dex support is not deprecated." stderr.txt
grep -Fq "https://docs.netbird.io/selfhosted/selfhosted-quickstart" stderr.txt
grep -Fq "https://docs.netbird.io/selfhosted/identity-providers/local" stderr.txt
grep -Fq "removed in NetBird v0.80" stderr.txt
- name: test zitadel.env file gen postgres
run: test -f zitadel.env
- name: test dashboard.env file gen postgres
run: test -f dashboard.env
- name: test relay.env file gen postgres
run: test -f relay.env
- name: test zdb.env file gen postgres
run: test -f zdb.env
- name: Postgres run cleanup
- name: Verify Zitadel retirement notice
run: |
docker compose down --volumes --rmi all
rm -rf docker-compose.yml Caddyfile zitadel.env dashboard.env machinekey/zitadel-admin-sa.token turnserver.conf management.json zdb.env
- name: run script with Zitadel CockroachDB
run: bash -x infrastructure_files/getting-started-with-zitadel.sh
env:
NETBIRD_DOMAIN: use-ip
ZITADEL_DATABASE: cockroach
- name: test Caddy file gen CockroachDB
run: test -f Caddyfile
- name: test docker-compose file gen CockroachDB
run: test -f docker-compose.yml
- name: test management.json file gen CockroachDB
run: test -f management.json
- name: test turnserver.conf file gen CockroachDB
run: |
set -x
test -f turnserver.conf
grep external-ip turnserver.conf
- name: test zitadel.env file gen CockroachDB
run: test -f zitadel.env
- name: test dashboard.env file gen CockroachDB
run: test -f dashboard.env
- name: test relay.env file gen CockroachDB
run: test -f relay.env
if bash infrastructure_files/getting-started-with-zitadel.sh >stdout.txt 2>stderr.txt; then
echo "Expected the retired Zitadel installer to fail"
exit 1
fi
test ! -s stdout.txt
grep -Fq "Zitadel support and existing Zitadel deployments are not deprecated." stderr.txt
grep -Fq "https://docs.netbird.io/selfhosted/selfhosted-quickstart" stderr.txt
grep -Fq "https://docs.netbird.io/selfhosted/identity-providers/zitadel" stderr.txt
grep -Fq "https://docs.netbird.io/selfhosted/selfhosted-guide" stderr.txt
grep -Fq "removed in NetBird v0.80" stderr.txt

View File

@@ -27,7 +27,7 @@ jobs:
with:
go-version-file: "go.mod"
- name: Install dependencies
run: sudo apt update && sudo apt install -y -q libgtk-3-dev libayatana-appindicator3-dev libgl1-mesa-dev xorg-dev libpcap-dev
run: sudo apt update && sudo apt install -y -q libgtk-4-dev libwebkitgtk-6.0-dev libsoup-3.0-dev libgl1-mesa-dev xorg-dev libpcap-dev
- name: Install golangci-lint
uses: golangci/golangci-lint-action@82606bf257cbaff209d206a39f5134f0cfbfd2ee #v9.2.1
with:

1
.gitignore vendored
View File

@@ -1,3 +1,4 @@
.claude
.idea
.run
*.iml

View File

@@ -114,6 +114,16 @@ linters:
- linters:
- staticcheck
text: "QF1012"
# client/ui/main.go uses //go:embed all:frontend/dist; the
# directory is populated by `pnpm build` in the release pipeline
# and missing at lint time, so the embed parses to "no matching
# files found" — surfaced by golangci-lint's typecheck pre-pass.
# Suppress just that one diagnostic; the rest of the package
# (services/, tray.go, grpc.go, ...) still gets linted normally.
- linters:
- typecheck
path: client/ui/main\.go
text: "pattern all:frontend/dist"
paths:
- third_party$
- builtin$

View File

@@ -212,6 +212,7 @@ nfpms:
description: Netbird client.
homepage: https://netbird.io/
license: BSD-3-Clause
vendor: NetBird
id: netbird_deb
bindir: /usr/bin
builds:
@@ -226,6 +227,7 @@ nfpms:
description: Netbird client.
homepage: https://netbird.io/
license: BSD-3-Clause
vendor: NetBird
id: netbird_rpm
bindir: /usr/bin
builds:
@@ -271,8 +273,8 @@ dockers_v2:
- netbirdio/netbird
- ghcr.io/netbirdio/netbird
tags:
- "v{{ .Version }}-rootless"
- "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
- "{{ .Version }}-rootless"
- "{{ if eq .Env.SKIP_PUBLISH \"false\" }}rootless-latest{{ end }}"
dockerfile: client/Dockerfile-rootless
extra_files:
- client/netbird-entrypoint.sh

View File

@@ -2,6 +2,15 @@ version: 2
env:
- SKIP_PUBLISH={{ if index .Env "SKIP_PUBLISH" }}{{ .Env.SKIP_PUBLISH }}{{ else }}true{{ end }}
project_name: netbird-ui
before:
hooks:
# Bindings are gitignored; regenerate before the frontend build so
# the @wailsio/runtime Vite plugin can resolve them (vite refuses to
# build without them).
- sh -c 'cd client/ui && wails3 generate bindings -clean=true -ts'
- sh -c 'cd client/ui/frontend && pnpm install --frozen-lockfile && pnpm build'
builds:
- id: netbird-ui
dir: client/ui
@@ -15,6 +24,8 @@ builds:
ldflags:
- -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
mod_timestamp: "{{ .CommitTimestamp }}"
tags:
- production
- id: netbird-ui-windows-amd64
dir: client/ui
@@ -30,6 +41,8 @@ builds:
- -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
- -H windowsgui
mod_timestamp: "{{ .CommitTimestamp }}"
tags:
- production
- id: netbird-ui-windows-arm64
dir: client/ui
@@ -46,6 +59,8 @@ builds:
- -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
- -H windowsgui
mod_timestamp: "{{ .CommitTimestamp }}"
tags:
- production
archives:
- id: linux-arch
@@ -62,6 +77,8 @@ nfpms:
- maintainer: Netbird <dev@netbird.io>
description: Netbird client UI.
homepage: https://netbird.io/
license: BSD-3-Clause
vendor: NetBird
id: netbird_ui_deb
package_name: netbird-ui
builds:
@@ -71,9 +88,9 @@ nfpms:
scripts:
postinstall: "release_files/ui-post-install.sh"
contents:
- src: client/ui/build/netbird.desktop
dst: /usr/share/applications/netbird.desktop
- src: client/ui/assets/netbird.png
- src: client/ui/build/linux/netbird.desktop
dst: /usr/share/applications/org.wails.netbird.desktop
- src: client/ui/build/appicon.png
dst: /usr/share/pixmaps/netbird.png
dependencies:
- netbird
@@ -81,6 +98,8 @@ nfpms:
- maintainer: Netbird <dev@netbird.io>
description: Netbird client UI.
homepage: https://netbird.io/
license: BSD-3-Clause
vendor: NetBird
id: netbird_ui_rpm
package_name: netbird-ui
builds:
@@ -90,12 +109,13 @@ nfpms:
scripts:
postinstall: "release_files/ui-post-install.sh"
contents:
- src: client/ui/build/netbird.desktop
dst: /usr/share/applications/netbird.desktop
- src: client/ui/assets/netbird.png
- src: client/ui/build/linux/netbird.desktop
dst: /usr/share/applications/org.wails.netbird.desktop
- src: client/ui/build/appicon.png
dst: /usr/share/pixmaps/netbird.png
dependencies:
- netbird
rpm:
signature:
key_file: '{{ if index .Env "GPG_RPM_KEY_FILE" }}{{ .Env.GPG_RPM_KEY_FILE }}{{ end }}'

View File

@@ -1,6 +1,15 @@
version: 2
project_name: netbird-ui
before:
hooks:
# Bindings are gitignored; regenerate before the frontend build so
# the @wailsio/runtime Vite plugin can resolve them (vite refuses to
# build without them).
- sh -c 'cd client/ui && wails3 generate bindings -clean=true -ts'
- sh -c 'cd client/ui/frontend && pnpm install --frozen-lockfile && pnpm build'
builds:
- id: netbird-ui-darwin
dir: client/ui
@@ -21,7 +30,7 @@ builds:
- -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
mod_timestamp: "{{ .CommitTimestamp }}"
tags:
- load_wgnt_from_rsrc
- production
universal_binaries:
- id: netbird-ui-darwin

View File

@@ -1,6 +1,6 @@
# Contributing to NetBird
Thanks for your interest in contributing to NetBird.
Thanks for your interest in contributing to NetBird.
There are many ways that you can contribute:
- Reporting issues
@@ -10,12 +10,69 @@ There are many ways that you can contribute:
If you haven't already, join our slack workspace [here](https://docs.netbird.io/slack-url), we would love to discuss topics that need community contribution and enhancements to existing features.
## Ticket first, PR second
**Open a ticket and wait for feedback before you open a pull request.** Every PR
that changes behavior must link to an issue the NetBird team has agreed on. A PR
that arrives without one may be closed and redirected to a discussion, no matter
how good the code is.
Issues in this repository are maintainer-curated work items, so the flow starts
in [Discussions](https://github.com/netbirdio/netbird/discussions):
1. **Open a discussion.** Use
[Issue Triage](https://github.com/netbirdio/netbird/discussions/new?category=issue-triage)
for a bug, regression, or unexpected behavior, and
[Ideas & Feature Requests](https://github.com/netbirdio/netbird/discussions/new?category=ideas-feature-requests)
for a feature, enhancement, or integration idea. Setup and usage questions
belong in
[Q&A / Support](https://github.com/netbirdio/netbird/discussions/new?category=q-a-support).
Never report a security vulnerability in public — follow the
[security policy](https://github.com/netbirdio/netbird/security/policy)
instead.
2. **Wait for feedback.** DevRel validates and reproduces the report, and a
maintainer confirms the direction. We may ask for more detail or propose a
different approach. Validated discussions become issues.
3. **Then write the code**, following the approach agreed in the issue, and open
the PR linking that issue.
Trivial fixes — a typo, a broken link, a documentation correction, or a one-line
fix that already has an issue — can go straight to a PR. Everything else starts
with a ticket. When in doubt, ask in the discussion or on
[Slack](https://docs.netbird.io/slack-url); an hour of conversation up front
regularly saves a week of rework.
### High-risk areas
These always need the design discussed and agreed in the issue **before** you
write code:
- **Public API** — REST / management API, OpenAPI schema, dashboard-facing contracts
- **gRPC protocols** — management, signal, relay, and client daemon protos
- **Functionality behavior** — anything existing deployments would experience differently after an upgrade
- **Peer connectivity** — ICE and NAT traversal, relay selection, WireGuard® and Rosenpass key handling
- **Client system integration** — routing, firewall, DNS, and interface management
- **Authentication and authorization** — IdP integration, tokens, permissions, cryptography
- **CLI / service flags**, configuration file format, and daemon IPC
- **Store and database schema** — models and migrations
- **New features**
These surfaces are NetBird's contract with operators, self-hosters, and
downstream integrators, and changes to them have compatibility, security, and
release-planning implications. Agreeing on the direction early lets the PR
review focus on implementation rather than design.
Typical bug fixes, internal refactors, documentation updates, and tests do not
need a design discussion, but should still be tied to an issue so the work is
visible and nobody duplicates it.
## Contents
- [Contributing to NetBird](#contributing-to-netbird)
- [Ticket first, PR second](#ticket-first-pr-second)
- [High-risk areas](#high-risk-areas)
- [Contents](#contents)
- [Code of conduct](#code-of-conduct)
- [Discuss changes with the NetBird team first](#discuss-changes-with-the-netbird-team-first)
- [Directory structure](#directory-structure)
- [Development setup](#development-setup)
- [Requirements](#requirements)
@@ -24,6 +81,7 @@ If you haven't already, join our slack workspace [here](https://docs.netbird.io/
- [Build and start](#build-and-start)
- [Test suite](#test-suite)
- [Checklist before submitting a PR](#checklist-before-submitting-a-pr)
- [When we close a PR](#when-we-close-a-pr)
- [Other project repositories](#other-project-repositories)
- [Contributor License Agreement](#contributor-license-agreement)
@@ -34,42 +92,66 @@ Conduct which can be found in the file [CODE_OF_CONDUCT.md](CODE_OF_CONDUCT.md).
By participating, you are expected to uphold this code. Please report
unacceptable behavior to community@netbird.io.
## Discuss changes with the NetBird team first
Changes to the **public API**, **gRPC protocols**, **functionality behavior**, **CLI / service flags**, or **new features** should be discussed with the NetBird team before you start the work. These surfaces are part of NetBird's contract with operators, self-hosters, and downstream integrators, and changes to them have compatibility, security, and release-planning implications that benefit from an early conversation.
Open an issue or reach out on [Slack](https://docs.netbird.io/slack-url) to talk through what you have in mind. We'll help shape the change, flag any constraints we know about, and confirm the direction so the PR review can focus on implementation rather than design.
Typical bug fixes, internal refactors, documentation updates, and tests do not need pre-discussion — open the PR directly.
## Directory structure
The NetBird project monorepo is organized to maintain most of its individual dependencies code within their directories, except for a few auxiliary or shared packages.
The NetBird project monorepo keeps most of each component's code within its own
directory, except for a few auxiliary or shared packages. Protocol definitions
and the client-side service clients live under [/shared](/shared), because both
the agent and the services import them.
The most important directories are:
**Agent**
- [/.github](/.github) - Github actions workflow files and issue templates
- [/client](/client) - NetBird agent code
- [/client/cmd](/client/cmd) - NetBird agent cli code
- [/client/cmd](/client/cmd) - NetBird agent CLI code
- [/client/internal](/client/internal) - NetBird agent business logic code
- [/client/proto](/client/proto) - NetBird agent daemon GRPC proto files
- [/client/server](/client/server) - NetBird agent daemon code for background execution
- [/client/ui](/client/ui) - NetBird agent UI code
- [/encryption](/encryption) - Contain main encryption code for agent communication
- [/iface](/iface) - Wireguard® interface code
- [/infrastructure_files](/infrastructure_files) - Getting started files containing docker and template scripts
- [/client/proto](/client/proto) - NetBird agent daemon gRPC proto files
- [/client/iface](/client/iface) - WireGuard® interface code
- [/client/firewall](/client/firewall) - Platform firewall backends (nftables, iptables, pf, WFP, userspace)
- [/client/ssh](/client/ssh) - Built-in SSH server and client
- [/client/ui](/client/ui) - NetBird agent UI code (Wails v3 + React)
- [/client/android](/client/android), [/client/ios](/client/ios) - Mobile platform bindings
- [/client/wasm](/client/wasm) - WebAssembly build of the agent
- [/client/mdm](/client/mdm) - MDM-delivered policy handling
- [/client/system](/client/system) - Host and system information collection
**Control plane services**
- [/management](/management) - Management service code
- [/management/client](/management/client) - Management service client code which is imported by the agent code
- [/management/proto](/management/proto) - Management service GRPC proto files
- [/management/server](/management/server) - Management service server code
- [/management/server/http](/management/server/http) - Management service REST API code
- [/management/server/store](/management/server/store) - Persistence layer and migrations
- [/management/server/idp](/management/server/idp) - Management service IDP management code
- [/release_files](/release_files) - Files that goes into release packages
- [/management/server/peer](/management/server/peer), [/management/server/groups](/management/server/groups), [/management/server/networks](/management/server/networks), [/management/server/posture](/management/server/posture), [/management/server/permissions](/management/server/permissions) - Core domain packages
- [/signal](/signal) - Signal service code
- [/signal/client](/signal/client) - Signal service client code which is imported by the agent code
- [/signal/peer](/signal/peer) - Signal service peer message logic
- [/signal/proto](/signal/proto) - Signal service GRPC proto files
- [/signal/server](/signal/server) - Signal service server code
- [/relay](/relay) - Relay service code
- [/relay/protocol](/relay/protocol) - Relay wire protocol
- [/proxy](/proxy) - Identity-aware proxy used by Agent Network (LLM routing, ACME, access logs)
- [/agent-network](/agent-network) - Agent Network overview and documentation
- [/upload-server](/upload-server) - Debug bundle upload service
**Shared code**
- [/shared/management/proto](/shared/management/proto) - Management service gRPC proto files
- [/shared/management/client](/shared/management/client) - Management service client code which is imported by the agent code
- [/shared/management/http/api](/shared/management/http/api) - OpenAPI specification and generated REST API types
- [/shared/signal/proto](/shared/signal/proto) - Signal service gRPC proto files
- [/shared/signal/client](/shared/signal/client) - Signal service client code which is imported by the agent code
- [/shared/relay](/shared/relay) - Relay client and shared relay types
- [/shared/auth](/shared/auth), [/shared/sshauth](/shared/sshauth) - Shared authentication primitives
- [/encryption](/encryption) - Contain main encryption code for agent communication
- [/dns](/dns), [/route](/route), [/stun](/stun), [/sharedsock](/sharedsock), [/util](/util) - Shared networking and utility primitives
- [/flow](/flow) - Flow event protocol shared by the agent and Management
**Build, test, and packaging**
- [/.github](/.github) - Github actions workflow files, issue templates, and the pull request template
- [/e2e](/e2e) - End-to-end test suites and harness
- [/infrastructure_files](/infrastructure_files) - Getting started files containing docker and template scripts
- [/release_files](/release_files) - Files that goes into release packages
- [/tools](/tools) - Development and maintenance tooling
## Development setup
@@ -79,13 +161,21 @@ dependencies are installed. Here is a short guide on how that can be done.
### Requirements
#### Go 1.21
#### Go 1.25
Follow the installation guide from https://go.dev/
#### UI client - Fyne toolkit
#### UI client - Wails v3 + React
We use the fyne toolkit in our UI client. You can follow its requirement guide to have all its dependencies installed: https://developer.fyne.io/started/#prerequisites
The desktop UI client (`client/ui`) is built with [Wails v3](https://v3.wails.io/) and a React frontend rendered in a WebView. To build it you need:
- Go ≥ 1.25
- Node ≥ 20 and **pnpm** (`corepack enable && corepack prepare pnpm@latest --activate`)
- The `wails3` CLI: `go install github.com/wailsapp/wails/v3/cmd/wails3@latest`
- The `task` runner: `go install github.com/go-task/task/v3/cmd/task@latest`
- Linux only: `libwebkitgtk-6.0-dev`, `libgtk-4-dev`, `libsoup-3.0-dev`
All UI build, dev-loop, and cross-compile commands are described in the [UI client](#ui-client) section below.
#### gRPC
You can follow the instructions from the quickstarter guide https://grpc.io/docs/languages/go/quickstart/#prerequisites and then run the `generate.sh` files located in each `proto` directory to generate changes.
@@ -214,6 +304,49 @@ To start NetBird the client in the foreground:
sudo ./client up --log-level debug --log-file console
```
> On Windows use a powershell with administrator privileges
#### UI client
The desktop UI lives in `client/ui` and is built with Wails v3 (see [Requirements](#ui-client---wails-v3--react)). All commands run from `client/ui`.
Live-reload development (Vite + Go binary + `*.go` watcher):
```
cd client/ui
task dev
```
Pass daemon flags after `--`, pointing the UI at the socket the daemon serves:
```
task dev -- --daemon-addr=unix:///var/run/netbird.sock # Linux, macOS
task dev -- --daemon-addr=npipe://netbird # Windows
```
On Windows the daemon serves a named pipe (`npipe://netbird`). Which path that
ends up being depends on what the daemon may create: as a service or elevated it
serves `\\.\pipe\ProtectedPrefix\Administrators\netbird`, which no unprivileged
process can take from it, and otherwise it falls back to `\\.\pipe\netbird`.
Clients try both and check who owns the pipe before using the plain one. Avoid
`tcp://127.0.0.1:41731`: loopback TCP carries no caller identity, so the daemon
refuses the operations that require an administrator and you will not exercise
those paths.
Production build (frontend assets embedded into the binary, output in `client/ui/bin/`):
```
cd client/ui
task build
```
Cross-compile the Windows binary from Linux (requires the mingw-w64 toolchain, e.g. `sudo apt install gcc-mingw-w64-x86-64`):
```
CGO_ENABLED=1 task windows:build
```
> macOS cross-compile from Linux is not supported (signing and notarization need a real Mac).
#### Signal service
To start NetBird's signal, execute:
@@ -251,10 +384,10 @@ Create dist directory
mkdir -p dist/netbird_windows_amd64
```
UI client
UI client (built with Wails v3 — see the [UI client](#ui-client) section above)
```shell
CC=x86_64-w64-mingw32-gcc CGO_ENABLED=1 GOOS=windows GOARCH=amd64 go build -o netbird-ui.exe -ldflags "-s -w -H windowsgui" ./client/ui
mv netbird-ui.exe ./dist/netbird_windows_amd64/
(cd client/ui && CGO_ENABLED=1 task windows:build)
mv client/ui/bin/netbird-ui.exe ./dist/netbird_windows_amd64/
```
Client
@@ -283,25 +416,172 @@ The installer `netbird-installer.exe` will be created in root directory.
### Test suite
The tests can be started via:
The host-safe unit tests run as a normal user and leave host networking
untouched:
```
cd netbird
go test -exec sudo ./...
make test-unit
```
Tests that need root and mutate host networking (firewall, routing, interface
management) carry the `privileged` build tag and run inside a
`--privileged --cap-add=NET_ADMIN` Docker container:
```
make test-privileged
```
Narrow a privileged run with environment variables:
```
PRIV_RUN=TestNftablesManager PRIV_PKGS=./client/firewall/nftables/... make test-privileged
```
Single packages can be run directly, adding `-race` when the change touches
shared state:
```
go test -race ./client/internal/dns/...
```
> On Windows use a powershell with administrator privileges
> Non-GTK environments will need the `libayatana-appindicator3-dev` (debian/ubuntu) package installed
## Checklist before submitting a PR
As a critical network service and open-source project, we must enforce a few things before submitting the pull-requests:
As a critical network service and open-source project, we must enforce a few
things before submitting a pull request. The
[pull request template](/.github/pull_request_template.md) mirrors this list —
fill it in rather than deleting it.
### Link the issue
The PR description must link the agreed issue (or the validated discussion it
came from). See [Ticket first, PR second](#ticket-first-pr-second).
### Run it locally
**If you can't run it, you can't submit it.** Build the affected components and
exercise the change on a real setup — see [Build and start](#build-and-start).
"CI will tell me" is not acceptable for a VPN agent that runs as root on other
people's machines.
### Green CI, and answer the bots
We do not start reviewing while CI is red. Get the pipeline green first — a
failing build, lint, or test means the PR is not ready for review.
Alongside the test workflows, your PR is reviewed by CodeRabbit and scanned by
SonarCloud, Snyk, and Codecov. Read what they report and either fix it or reply
with why it does not apply; please do not resolve the threads without a
response. They are not always right — this codebase has privileged,
platform-specific, and concurrency-heavy paths that static analysis reads poorly
— so push back when a finding is wrong rather than changing correct code to
silence it. Security and dependency findings are the exception: treat those as
real until shown otherwise. Do not edit workflows, thresholds, or scanner
configuration to make a check pass.
### One PR, one purpose
Bug fix, refactor, feature: separate PRs. Mixed PRs are slow to review, hard to
revert, and may be closed with a request to split them.
### Keep it small
Size is the strongest predictor of how long a PR waits for review. Aim for under
roughly 400 changed lines across under 20 files. Past about 1000 lines or 50
files, expect to be asked to split the change — and large PRs from outside the
core team may be blocked until the scope has been agreed in a ticket. This is
not only about reviewer time: NetBird's agent runs as root on other people's
machines, and a sprawling diff cannot be reviewed with the care that deserves.
Measure by hand-written code, excluding generated output, `go.sum`, and
fixtures. If a change genuinely cannot be small — a protocol migration, a
cross-component rename — agree the split in the issue before you start, and land
it as a series of PRs that each build and make sense on their own.
### Avoid force-pushing during review
Once a PR is open, push new commits instead of rewriting history. A force-push
detaches existing review comments from their lines, throws away the
"changes since your last review" diff, and loses the CI history that showed
which commit broke what. Since we squash on merge, there is nothing to gain from
a tidy branch history.
Force-pushing is sometimes unavoidable — rebasing to clear a real conflict, or
removing a secret or large binary committed by mistake. When that happens, leave
a comment on the PR so reviewers know their anchors moved.
### Quality checks
Run these from the repository root before pushing:
```shell
go fmt ./...
make lint # golangci-lint on files changed against origin/main
make lint-all # full-repository lint, matches CI
make test-unit # host-safe unit tests
```
`make setup-hooks` wires `make lint` into a pre-push hook so the fast lint runs
automatically. If your change touches privileged paths (firewall, routing,
interface management), also run `make test-privileged`, which executes the
`privileged`-tagged suite inside a Docker container with `NET_ADMIN`.
### Code standards
- Keep functions as simple as possible, with a single purpose
- Use private functions and constants where possible
- Comment on any new public functions
- Add unit tests for any new public function
- Comment the **why**, not the **what** — explain non-obvious decisions, invariants, and constraints, not the line below
- Keep comments within 90 characters per line and roughly 250 characters per comment; when a block needs more explanation than that, extract a named function instead of writing a longer comment (see [AGENTS.md](AGENTS.md#length-budget))
### PR title and commits
PR titles must start with a bracketed tag, enforced by
[pr-title-check.yml](/.github/workflows/pr-title-check.yml):
```text
[client] Authorize daemon IPC callers by their local identity
[management,client] Add MDM policy support
```
Use a comma-separated list inside a single pair of brackets when a change spans
components. The `allowedTags` array in
[pr-title-check.yml](/.github/workflows/pr-title-check.yml) is the source of
truth — at the time of writing it accepts `management`, `client`, `signal`,
`proxy`, `relay`, `misc`, `infrastructure`, `self-hosted`, and `doc`.
Commit subjects follow the same convention — keep them short and put the
reasoning in the body, why before what, with no bullet list of files changed.
Keep the PR description itself under 1000 words on top of the template text.
Reviewers read the diff; the description explains what the diff cannot.
> When pushing fixes to the PR comments, please push as separate commits; we will squash the PR before merging, so there is no need to squash it before pushing it, and we are more than okay with 10-100 commits in a single PR. This helps review the fixes to the requested changes.
### Documentation
User-facing changes need a matching PR in
[netbirdio/docs](https://github.com/netbirdio/docs); link it in the PR
description, or state why documentation is not needed.
## When we close a PR
We would rather redirect early than let a PR sit. We may close one if:
- It changes behavior with no linked issue, or the approach was never agreed with a maintainer
- The change was clearly never run or tested locally
- CI has been red without a response
- It mixes unrelated purposes, or the purpose is not clear
- It is far too large to review and the scope was never agreed in a ticket
- The author cannot answer questions about their own change — including PRs that read as unreviewed model output, where review turns into a relay between the maintainer and an LLM. Tooling is fine; unreviewed output is not, you are responsible for the code you sign your name to
- There has been no activity for 14 days after we requested changes
A closed PR is not a rejected idea. Take it back to the
[discussion](https://github.com/netbirdio/netbird/discussions), settle the
approach, and reopen the work from there.
## Other project repositories
NetBird project is composed of 3 main repositories:

View File

@@ -1,12 +1,70 @@
# Security Policy
NetBird's goal is to provide a secure network. If you find a vulnerability or bug, please report it by opening an issue [here](https://github.com/netbirdio/netbird/issues/new?assignees=&labels=&template=bug-issue-report.md&title=) or by contacting us by email.
There has yet to be an official bug bounty program for the NetBird project.
## Supported Versions
- We currently support only the latest version
NetBird's goal is to provide a secure network. The client runs as a privileged service on every machine it is installed on,
so we take reports about it seriously and we publish what we fix.
## Reporting a Vulnerability
Please report security issues to `security@netbird.io`
**Please do not open a public issue for a security vulnerability.** Public issues are visible to everyone, including before
a fix is available.
Report security issues one of these two ways:
- **GitHub private vulnerability reporting** — [open a private report](https://github.com/netbirdio/netbird/security/advisories/new)
on this repository. This is the preferred route: it keeps the discussion, the draft advisory, and the credit in one place.
- **Email** — `security@netbird.io`.
If the finding affects NetBird Cloud or our hosted infrastructure rather than the open-source code, email us rather than
filing a repository report.
### What to include
A report is easier to act on when it contains:
- The affected component (client, management, signal, relay, dashboard) and the version or commit you tested
- The platform and configuration, where relevant — operating system, self-hosted or NetBird Cloud, container or host install
- What an attacker needs before they can exploit it: network position, an account, local access, a specific privilege level
- Steps to reproduce, and a proof of concept if you have one
- The impact you believe it has
Partial reports are still welcome. If you are unsure whether something is a security issue, send it to `security@netbird.io`
and let us make that call.
## What to expect from us
- **We acknowledge your report** and tell you whether we can reproduce it.
- **We work with you on severity and scope.** If we assess it differently than you do, we will explain why rather than
silently downgrade it.
- **We fix and release**, then publish a [GitHub Security Advisory](https://github.com/netbirdio/netbird/security/advisories)
naming the affected version range and the patched version.
- **We credit reporters who want to be credited.** Tell us the name or handle you would like used, or that you would rather
stay anonymous.
- **We keep you in the loop** until the advisory is published.
We ask that you give us a reasonable opportunity to ship a fix before disclosing the issue publicly, and that you avoid
accessing, modifying, or exfiltrating data belonging to other people while testing. Testing against your own installation
or your own account is always fine.
## Supported Versions
We support the latest release. Security fixes ship in the next version rather than as backports to older releases, so
upgrading to the current release is how you get them.
Release notifications are available by watching [releases](https://github.com/netbirdio/netbird/releases).
## Published advisories
Every vulnerability we fix is published as a GitHub Security Advisory on the
[advisories page](https://github.com/netbirdio/netbird/security/advisories), including the affected version range, the
patched version, and the reporter's credit. Advisories for the Go module are also distributed through the Go vulnerability
database, so `govulncheck` will report them against your dependencies.
## Bug bounty
There is no official bug bounty program for the NetBird project. We credit reporters in advisories, and we are grateful for
the work, but we cannot currently offer payment for reports.
## Non-security bugs
For bugs that are not security issues, please use the
[issue tracker](https://github.com/netbirdio/netbird/discussions/new/choose).

View File

@@ -1,16 +1,47 @@
# NetBird Agent Network
Agent Network is NetBird's access control layer for AI agents and the people who run
them. It gives every agent a real identity, tied to your identity provider (IdP), and
governs what it can reach — the LLM APIs and AI gateways it can call, and the internal
resources it can access. Traffic flows only over the encrypted NetBird tunnel, scoped by
policy, with no API keys to leak.
Agent Network is NetBird's access control layer for AI agents and the people who run them.
It gives every agent a real identity, tied to an identity provider (IdP), and governs what it can reach: LLM APIs and
AI gateways it can call, and the internal resources it can access. Traffic flows only over the encrypted NetBird tunnel,
scoped by policy, with no API keys or other credentials to leak. It also gives you control over cost and token usage.
> **Beta.** Agent Network is open source and can be self-hosted on your own
> infrastructure.
Because every LLM request passes through an
identity-aware proxy, you can:
- **Set spending and rate limits** per agent, per user, or per team — with hard caps
that stop requests once a budget is reached.
- **Restrict models and providers** so agents can only call approved (and cost-appropriate)
endpoints, keeping expensive models off-limits unless explicitly allowed.
- **Attribute usage** by tracking token consumption and cost per identity, group, or cost center so every
request is tied back to the agent and person responsible.
- **Reuse your existing AI gateway** — point the proxy at a gateway you already run,
keeping its routing and config in place while it adds identity on top, so you skip
API key distribution.
https://github.com/user-attachments/assets/44d18286-d8ab-49f8-a457-98ccd66f3268
> **Beta.** Agent Network is in beta, but it's stable and already running in
> production environments. It's fully open source and can be self-hosted on your own
> infrastructure, with no vendor lock-in and no data leaving your environment.
## How it works
Say you have a simple use case: your Engineering or IT team needs access to Claude Code or Codex, and you want visibility into usage plus the ability to enforce budgets.
How can you do that without creating a dedicated API key for every team?
With Agent Network you get a private endpoint inside your network, for example: https://mirror.netbird.ai
Teams configure their agents to point to that endpoint instead of using individual API keys directly.
This endpoint is only reachable when users are connected to your NetBird network and authenticated through your IdP. Otherwise, it is not accessible from the public internet.
You can then use this private endpoint to configure your AI agents, whether that is Claude Code, Codex, or another tool.
## Quickstart
Full step-by-step setup:
**https://docs.netbird.io/agent-network/quickstart**
## Architecture
Agent Network is built on two existing NetBird capabilities:
- **Overlay network** — the encrypted WireGuard mesh between peers.
@@ -22,6 +53,9 @@ LLM traffic is routed through the proxy's identity-aware pipeline, while interna
resources (databases, internal APIs, self-hosted models) are reached directly over
peer-to-peer WireGuard tunnels, governed by the same identities and access policies.
<img width="4720" height="2218" alt="image" src="https://github.com/user-attachments/assets/1afa5da1-4b82-4f8a-a7a8-f417efadf1eb" />
## Where the code lives
There is no separate "agent-network" service — it reuses the reverse-proxy and management

View File

@@ -7,6 +7,7 @@ import (
"fmt"
"os"
"slices"
"strings"
"sync"
"time"
@@ -75,6 +76,24 @@ type Client struct {
connectClient *internal.ConnectClient
config *profilemanager.Config
cacheDir string
stateChangeMu sync.Mutex
stateChangeSubID string
eventSub *peer.EventSubscription
// Closed to stop the watch goroutines from delivering buffered items to a
// listener that has been removed or replaced. See stopStateChangeWatchLocked.
stateChangeDone chan struct{}
// Latched "the server wants an interactive login": survives the engine
// restarts that replace the run loop's context state. See Client.Status.
// Guarded by loginRequiredMu together with loginCleared, which counts
// clears so a stale observation cannot re-latch over one.
loginRequiredMu sync.Mutex
loginRequired bool
loginCleared uint64
extendMu sync.Mutex
extendCancel context.CancelFunc
}
func (c *Client) setState(cfg *profilemanager.Config, cacheDir string, cc *internal.ConnectClient) {
@@ -148,11 +167,16 @@ func (c *Client) Run(platformFiles PlatformFiles, urlOpener URLOpener, isAndroid
if err != nil {
return err
}
// todo do not throw error in case of cancelled context
ctx = internal.CtxInitState(ctx)
connectClient := internal.NewConnectClient(ctx, cfg, c.recorder)
c.setState(cfg, cacheDir, connectClient)
// This path runs the interactive SSO flow, so reaching here means the peer
// is authenticated again — release the latch Status() reports from. Clear
// only once the fresh connect client is installed: until then Status()
// still reads the previous run's context state, which holds the NeedsLogin
// that prompted this login, and would re-latch what was just cleared.
c.clearLoginRequired()
return connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
}
@@ -247,6 +271,9 @@ func (c *Client) DebugBundle(platformFiles PlatformFiles, anonymize bool) (strin
deps.SyncResponse = resp
if e := cc.Engine(); e != nil {
deps.RefreshStatus = func() {
e.RunHealthProbes(context.Background(), true)
}
if cm := e.GetClientMetrics(); cm != nil {
deps.ClientMetrics = cm
}
@@ -274,7 +301,7 @@ func (c *Client) DebugBundle(platformFiles PlatformFiles, anonymize bool) (strin
uploadCtx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
defer cancel()
key, err := debug.UploadDebugBundle(uploadCtx, types.DefaultBundleURL, cfg.ManagementURL.String(), path)
key, err := debug.UploadDebugBundle(uploadCtx, types.DefaultBundleURL, cfg.ManagementURL.String(), path, false)
if err != nil {
return "", fmt.Errorf("upload debug bundle: %w", err)
}
@@ -296,6 +323,13 @@ func (c *Client) SetInfoLogLevel() {
// PeersList return with the list of the PeerInfos
func (c *Client) PeersList() *PeerInfoArray {
// The recorder only caches transfer counters and handshake times; nothing
// refreshes them on its own, so without this they read as zero. The desktop
// daemon does the same before serving a full peer status.
if err := c.recorder.RefreshWireGuardStats(); err != nil {
log.Debugf("failed to refresh WireGuard stats: %v", err)
}
fullStatus := c.recorder.GetFullStatus()
peerInfos := make([]PeerInfo, len(fullStatus.Peers))
@@ -306,6 +340,20 @@ func (c *Client) PeersList() *PeerInfoArray {
FQDN: p.FQDN,
ConnStatus: int(p.ConnStatus),
Routes: PeerRoutes{routes: maps.Keys(p.GetRoutes())},
PubKey: p.PubKey,
Latency: formatDuration(p.Latency),
LatencyMs: p.Latency.Milliseconds(),
BytesRx: p.BytesRx,
BytesTx: p.BytesTx,
ConnStatusUpdate: formatTime(p.ConnStatusUpdate),
Relayed: p.Relayed,
RosenpassEnabled: p.RosenpassEnabled,
LastWireguardHandshake: formatTime(p.LastWireguardHandshake),
LocalIceCandidateType: p.LocalIceCandidateType,
RemoteIceCandidateType: p.RemoteIceCandidateType,
LocalIceCandidateEndpoint: p.LocalIceCandidateEndpoint,
RemoteIceCandidateEndpoint: p.RemoteIceCandidateEndpoint,
}
peerInfos[n] = pi
}
@@ -436,10 +484,6 @@ func (c *Client) RemoveConnectionListener() {
c.recorder.RemoveConnectionListener()
}
func (c *Client) toggleRoute(command routeCommand) error {
return command.toggleRoute()
}
func (c *Client) getRouteManager() (routemanager.Manager, error) {
client := c.getConnectClient()
if client == nil {
@@ -459,22 +503,22 @@ func (c *Client) getRouteManager() (routemanager.Manager, error) {
return manager, nil
}
func (c *Client) SelectRoute(route string) error {
func (c *Client) SelectRoute(id string) error {
manager, err := c.getRouteManager()
if err != nil {
return err
}
return c.toggleRoute(selectRouteCommand{route: route, manager: manager})
return manager.SelectRoutes([]route.NetID{route.NetID(id)}, true)
}
func (c *Client) DeselectRoute(route string) error {
func (c *Client) DeselectRoute(id string) error {
manager, err := c.getRouteManager()
if err != nil {
return err
}
return c.toggleRoute(deselectRouteCommand{route: route, manager: manager})
return manager.DeselectRoutes([]route.NetID{route.NetID(id)})
}
// getNetworkDomainsFromRoute extracts domains from a route and enriches each domain
@@ -509,3 +553,28 @@ func exportEnvList(list *EnvList) {
}
}
}
// formatDuration renders a duration for display, trimming the fractional part
// to two digits so latencies read as "12.34ms" rather than "12.345678ms".
func formatDuration(d time.Duration) string {
ds := d.String()
dotIndex := strings.Index(ds, ".")
if dotIndex == -1 {
return ds
}
endIndex := min(dotIndex+3, len(ds))
// Skip the remaining digits so only the unit suffix is appended back.
unitStart := endIndex
for unitStart < len(ds) && ds[unitStart] >= '0' && ds[unitStart] <= '9' {
unitStart++
}
return ds[:endIndex] + ds[unitStart:]
}
// formatTime renders a timestamp in UTC using a fixed layout. The zero time is
// passed through as-is so the UI can recognise it and show "never" instead.
func formatTime(t time.Time) string {
return t.UTC().Format("2006-01-02 15:04:05")
}

View File

@@ -12,12 +12,30 @@ const (
)
// PeerInfo describe information about the peers. It designed for the UI usage
//
// The fields below ConnStatus back the peer detail screen. Durations and times
// are pre-formatted into strings so the UI does not have to know Go's layouts;
// Latency is additionally exposed as LatencyMs for colour coding.
type PeerInfo struct {
IP string
IPv6 string
FQDN string
ConnStatus int
Routes PeerRoutes
PubKey string
Latency string
LatencyMs int64
BytesRx int64
BytesTx int64
ConnStatusUpdate string
Relayed bool
RosenpassEnabled bool
LastWireguardHandshake string
LocalIceCandidateType string
RemoteIceCandidateType string
LocalIceCandidateEndpoint string
RemoteIceCandidateEndpoint string
}
func (p *PeerInfo) GetPeerRoutes() *PeerRoutes {

View File

@@ -189,6 +189,19 @@ func (pm *ProfileManager) LogoutProfile(id string) error {
return nil
}
// RenameProfile changes a profile's display name. The profile ID, and therefore
// its on-disk filename, is left untouched: only the "name" field of the config
// is rewritten. This works for the default profile too, whose config lives in
// netbird.cfg rather than under profiles/.
func (pm *ProfileManager) RenameProfile(id string, newName string) error {
if err := pm.serviceMgr.RenameProfile(profilemanager.ID(id), androidUsername, newName); err != nil {
return fmt.Errorf("failed to rename profile: %w", err)
}
log.Infof("renamed profile %s to: %s", id, newName)
return nil
}
// RemoveProfile deletes a profile
func (pm *ProfileManager) RemoveProfile(id string) error {
// Use ServiceManager (removes profile from profiles/ directory)

View File

@@ -1,70 +0,0 @@
//go:build android
package android
import (
"fmt"
log "github.com/sirupsen/logrus"
"golang.org/x/exp/maps"
"github.com/netbirdio/netbird/client/internal/routemanager"
"github.com/netbirdio/netbird/route"
)
func executeRouteToggle(id string, manager routemanager.Manager,
operationName string,
routeOperation func(routes []route.NetID, allRoutes []route.NetID) error) error {
netID := route.NetID(id)
routes := []route.NetID{netID}
routesMap := manager.GetClientRoutesWithNetID()
routes = route.ExpandV6ExitPairs(routes, routesMap)
log.Debugf("%s with ids: %v", operationName, routes)
if err := routeOperation(routes, maps.Keys(routesMap)); err != nil {
log.Debugf("error when %s: %s", operationName, err)
return fmt.Errorf("error %s: %w", operationName, err)
}
manager.TriggerSelection(manager.GetClientRoutes())
return nil
}
type routeCommand interface {
toggleRoute() error
}
type selectRouteCommand struct {
route string
manager routemanager.Manager
}
func (s selectRouteCommand) toggleRoute() error {
routeSelector := s.manager.GetRouteSelector()
if routeSelector == nil {
return fmt.Errorf("no route selector available")
}
routeOperation := func(routes []route.NetID, allRoutes []route.NetID) error {
return routeSelector.SelectRoutes(routes, true, allRoutes)
}
return executeRouteToggle(s.route, s.manager, "selecting route", routeOperation)
}
type deselectRouteCommand struct {
route string
manager routemanager.Manager
}
func (d deselectRouteCommand) toggleRoute() error {
routeSelector := d.manager.GetRouteSelector()
if routeSelector == nil {
return fmt.Errorf("no route selector available")
}
return executeRouteToggle(d.route, d.manager, "deselecting route", routeSelector.DeselectRoutes)
}

309
client/android/session.go Normal file
View File

@@ -0,0 +1,309 @@
//go:build android
package android
import (
"context"
"fmt"
log "github.com/sirupsen/logrus"
"github.com/netbirdio/netbird/client/internal"
"github.com/netbirdio/netbird/client/internal/auth"
"github.com/netbirdio/netbird/client/internal/auth/sessionwatch"
"github.com/netbirdio/netbird/client/internal/peer"
cProto "github.com/netbirdio/netbird/client/proto"
)
// StateChangeListener receives client state notifications.
//
// OnStateChanged is a payload-free wake-up whenever the state snapshot
// changed: connection state, the run-loop status label (e.g. NeedsLogin) or
// the session deadline. It mirrors the daemon's SubscribeStatus stream
// trigger — on each signal the consumer pulls the fresh values via
// Status() / SessionExpiresAtUnix().
//
// OnSessionExpiring forwards the engine's session-expiry warnings, fired at
// sessionwatch.WarningLead before the deadline and again at FinalWarningLead
// (finalWarning true). The second one is suppressed when the user dismissed
// the first via DismissSessionWarning. The daemon turns the same events into
// its tray notification.
type StateChangeListener interface {
OnStateChanged()
OnSessionExpiring(expiresAtUnix int64, leadMinutes int64, finalWarning bool)
}
// Status returns the connect run-loop's status label — the same value the
// desktop daemon serves in StatusResponse.Status. "NeedsLogin" means the
// management server rejected the peer and an interactive login is required.
//
// The label is latched: the run loop keeps its status in a per-run context
// state, which a restart replaces with a fresh Idle one, so an engine restart
// (network change, always-on) would otherwise erase the fact that the peer
// still needs to log in. Only a successful interactive login or extend clears
// it — see clearLoginRequired.
func (c *Client) Status() string {
latched, generation := c.loginRequiredState()
if latched {
return string(internal.StatusNeedsLogin)
}
cc := c.getConnectClient()
if cc == nil {
return string(internal.StatusIdle)
}
status := cc.Status()
if status == internal.StatusNeedsLogin {
c.latchLoginRequired(generation)
}
return string(status)
}
func (c *Client) loginRequiredState() (bool, uint64) {
c.loginRequiredMu.Lock()
defer c.loginRequiredMu.Unlock()
return c.loginRequired, c.loginCleared
}
// latchLoginRequired records a NeedsLogin observation, unless a clear landed
// while the caller was reading the run loop's status: cc.Status() is read
// outside the lock, so a login or extend completing in that window would
// otherwise be undone by this stale observation, stranding the UI on
// "login required" over a healthy session.
func (c *Client) latchLoginRequired(observedGeneration uint64) {
c.loginRequiredMu.Lock()
defer c.loginRequiredMu.Unlock()
if c.loginCleared != observedGeneration {
return
}
c.loginRequired = true
}
// clearLoginRequired releases the latch after a successful interactive login
// or session extend, and invalidates any observation already in flight.
func (c *Client) clearLoginRequired() {
c.loginRequiredMu.Lock()
defer c.loginRequiredMu.Unlock()
c.loginRequired = false
c.loginCleared++
}
// SessionExpiresAtUnix returns the SSO session deadline as unix seconds, or 0
// when no deadline is known (not SSO-registered, expiry disabled, or the
// engine has not received one yet). A past value means the session expired.
// Mirror of StatusResponse.sessionExpiresAt on the desktop daemon.
func (c *Client) SessionExpiresAtUnix() int64 {
deadline := c.recorder.GetSessionExpiresAt()
if deadline.IsZero() {
return 0
}
return deadline.Unix()
}
// SetStateChangeListener registers the state notification listener.
// Replaces any previously registered listener; remove it with
// RemoveStateChangeListener.
func (c *Client) SetStateChangeListener(listener StateChangeListener) {
c.stateChangeMu.Lock()
defer c.stateChangeMu.Unlock()
c.stopStateChangeWatchLocked()
if listener == nil {
return
}
// Both subscriptions are buffered (one pending tick, ten pending events),
// so unsubscribing is not enough to stop callbacks: the loops would drain
// what is already queued and deliver it to a listener the caller has
// already removed or replaced. Gate every callback on this registration's
// own signal, which is closed before unsubscribing.
done := make(chan struct{})
c.stateChangeDone = done
id, ch := c.recorder.SubscribeToStateChanges()
c.stateChangeSubID = id
// The channel is closed by UnsubscribeFromStateChanges, which ends the
// goroutine. Ticks are coalesced (buffer of one), so a burst of changes
// wakes the listener once.
go func() {
for range ch {
select {
case <-done:
return
default:
}
listener.OnStateChanged()
}
}()
c.eventSub = c.recorder.SubscribeToEvents()
go watchSessionWarnings(c.eventSub, listener, done)
}
// RemoveStateChangeListener unregisters the state notification listener.
func (c *Client) RemoveStateChangeListener() {
c.stateChangeMu.Lock()
defer c.stateChangeMu.Unlock()
c.stopStateChangeWatchLocked()
}
// DismissSessionWarning records the user's "Dismiss" on the first expiry
// warning and suppresses the final one for the current deadline. A refreshed
// deadline re-arms both. No-op while the engine is not running.
func (c *Client) DismissSessionWarning() {
cc := c.getConnectClient()
if cc == nil {
return
}
engine := cc.Engine()
if engine == nil {
return
}
engine.DismissSessionWarning()
}
// ExtendAuthSession runs the interactive SSO flow to obtain a fresh JWT and
// asks the management server to extend the session deadline. The tunnel is
// untouched: no resync, no reconnect. Async; the result arrives on the
// listener. Mirror of the daemon's RequestExtendAuthSession /
// WaitExtendAuthSession RPC pair, with URLOpener playing the "UI opens the
// browser" role.
//
// Only one flow may be in flight: the PKCE step binds a fixed loopback port,
// so a second concurrent flow would fail on that bind. Call
// CancelExtendAuthSession when the user abandons the browser.
func (c *Client) ExtendAuthSession(urlOpener URLOpener, isAndroidTV bool, resultListener ErrListener) {
ctx, err := c.beginExtend()
if err != nil {
resultListener.OnError(err)
return
}
go func() {
defer c.endExtend()
if err := c.extendAuthSession(ctx, urlOpener, isAndroidTV); err != nil {
resultListener.OnError(err)
return
}
resultListener.OnSuccess()
}()
}
// CancelExtendAuthSession aborts an in-flight ExtendAuthSession. The tunnel is
// left alone — unlike the login flow, which cancels the whole client context
// by stopping the engine. Without this the abandoned PKCE wait keeps its
// loopback port for the full flow timeout and blocks every later attempt.
// No-op when no flow is running.
func (c *Client) CancelExtendAuthSession() {
c.extendMu.Lock()
defer c.extendMu.Unlock()
if c.extendCancel != nil {
c.extendCancel()
}
}
func (c *Client) stopStateChangeWatchLocked() {
// Signal first, unsubscribe second: closing the channels only stops new
// items, and the loops would still hand whatever is buffered to a listener
// that is no longer registered.
if c.stateChangeDone != nil {
close(c.stateChangeDone)
c.stateChangeDone = nil
}
if c.stateChangeSubID != "" {
c.recorder.UnsubscribeFromStateChanges(c.stateChangeSubID)
c.stateChangeSubID = ""
}
if c.eventSub != nil {
// Closes the channel, which ends watchSessionWarnings.
c.recorder.UnsubscribeFromEvents(c.eventSub)
c.eventSub = nil
}
}
// watchSessionWarnings forwards the engine's session-expiry warnings to the
// listener. The event stream also carries unrelated traffic — network-map
// updates on every sync, DNS and route errors — so everything but an
// AUTHENTICATION event carrying the session-warning marker is dropped. Exits
// when the subscription is closed by UnsubscribeFromEvents, or earlier when
// done is closed — the stream buffers up to ten events, and a deregistered
// listener must not receive the ones already queued.
func watchSessionWarnings(sub *peer.EventSubscription, listener StateChangeListener, done <-chan struct{}) {
for ev := range sub.Events() {
select {
case <-done:
return
default:
}
if ev.GetCategory() != cProto.SystemEvent_AUTHENTICATION {
continue
}
meta := ev.GetMetadata()
if meta[sessionwatch.MetaSessionWarning] != "true" {
// Other AUTHENTICATION events exist (e.g. a deadline rejected as
// out of range); they carry no warning marker.
continue
}
deadline, err := sessionwatch.ParseExpiresAt(meta[sessionwatch.MetaSessionExpiresAt])
if err != nil {
log.Warnf("session warning event with unparsable deadline: %v", err)
continue
}
lead, err := sessionwatch.ParseLeadMinutes(meta[sessionwatch.MetaSessionLeadMinutes])
if err != nil {
// Informational only — the deadline above is what drives the UI.
lead = 0
}
listener.OnSessionExpiring(deadline.Unix(), int64(lead),
meta[sessionwatch.MetaSessionFinal] == "true")
}
}
func (c *Client) beginExtend() (context.Context, error) {
c.extendMu.Lock()
defer c.extendMu.Unlock()
if c.extendCancel != nil {
return nil, fmt.Errorf("session extend already in progress")
}
ctx, cancel := context.WithCancel(context.Background())
c.extendCancel = cancel
return ctx, nil
}
func (c *Client) endExtend() {
c.extendMu.Lock()
defer c.extendMu.Unlock()
if c.extendCancel != nil {
c.extendCancel()
c.extendCancel = nil
}
}
func (c *Client) extendAuthSession(ctx context.Context, urlOpener URLOpener, isAndroidTV bool) error {
cfg, _, cc := c.stateSnapshot()
if cfg == nil || cc == nil {
return fmt.Errorf("engine is not running")
}
engine := cc.Engine()
if engine == nil {
return fmt.Errorf("engine is not initialized")
}
authClient, err := auth.NewAuth(ctx, cfg.PrivateKey, cfg.ManagementURL, cfg)
if err != nil {
return fmt.Errorf("failed to create auth client: %v", err)
}
defer authClient.Close()
a := &Auth{ctx: ctx, config: cfg}
tokenInfo, err := a.foregroundGetTokenInfo(authClient, urlOpener, isAndroidTV)
if err != nil {
return fmt.Errorf("interactive sso login failed: %v", err)
}
if _, err := engine.ExtendAuthSession(ctx, tokenInfo.GetTokenToUse()); err != nil {
return err
}
c.clearLoginRequired()
go urlOpener.OnLoginSuccess()
return nil
}

View File

@@ -0,0 +1,66 @@
package cmd
import (
"errors"
"fmt"
"strings"
"google.golang.org/genproto/googleapis/rpc/errdetails"
gstatus "google.golang.org/grpc/status"
"github.com/netbirdio/netbird/client/internal/ipcauth"
)
// daemonCallError prepares a daemon error for display. A refusal the daemon
// raised because the operation needs root/administrator is already guidance
// written for the user, so it is surfaced on its own instead of buried under the
// gRPC envelope and the name of the RPC that hit it. Anything else is wrapped
// with context as usual.
func daemonCallError(context string, err error) error {
if guidance, ok := privilegeGuidance(err); ok {
return errors.New(guidance)
}
return fmt.Errorf("%s: %w", context, err)
}
// privilegeGuidance renders the daemon's privilege refusal as a summary and the
// command that performs the operation with the privileges it needs. It reports
// false for any other error.
func privilegeGuidance(err error) (string, bool) {
info, ok := privilegeErrorInfo(err)
if !ok {
return "", false
}
summary := info.GetMetadata()[ipcauth.ErrorMetaSummary]
command := info.GetMetadata()[ipcauth.ErrorMetaCommand]
if summary == "" {
// Detail without a summary: fall back to the status message, which
// carries the same text.
summary = strings.TrimSpace(gstatus.Convert(err).Message())
}
if command == "" {
return summary, true
}
return fmt.Sprintf("%s\n\n %s\n", summary, command), true
}
// privilegeErrorInfo returns the daemon's privilege-refusal detail, if the error
// carries one.
func privilegeErrorInfo(err error) (*errdetails.ErrorInfo, bool) {
if err == nil {
return nil, false
}
for _, detail := range gstatus.Convert(err).Details() {
info, ok := detail.(*errdetails.ErrorInfo)
if !ok {
continue
}
if info.GetReason() == ipcauth.ErrorReasonPrivilegeRequired && info.GetDomain() == ipcauth.ErrorDomain {
return info, true
}
}
return nil, false
}

View File

@@ -29,8 +29,9 @@ const errCloseConnection = "Failed to close connection: %v"
var (
logFileCount uint32
systemInfoFlag bool
uploadBundleFlag bool
uploadBundleURLFlag string
uploadBundleFlag bool
uploadBundleURLFlag string
uploadBundleInsecureFlag bool
)
var debugCmd = &cobra.Command{
@@ -174,10 +175,11 @@ func debugBundle(cmd *cobra.Command, _ []string) error {
}
if uploadBundleFlag {
request.UploadURL = uploadBundleURLFlag
request.UploadInsecure = uploadBundleInsecureFlag
}
resp, err := client.DebugBundle(cmd.Context(), request)
if err != nil {
return fmt.Errorf("failed to bundle debug: %v", status.Convert(err).Message())
return daemonCallError("bundle debug", err)
}
cmd.Printf("Local file:\n%s\n", resp.GetPath())
@@ -373,10 +375,11 @@ func runForDuration(cmd *cobra.Command, args []string) error {
}
if uploadBundleFlag {
request.UploadURL = uploadBundleURLFlag
request.UploadInsecure = uploadBundleInsecureFlag
}
resp, err := client.DebugBundle(cmd.Context(), request)
if err != nil {
return fmt.Errorf("failed to bundle debug: %v", status.Convert(err).Message())
return daemonCallError("bundle debug", err)
}
if needsRestoreUp {
@@ -524,10 +527,12 @@ func init() {
debugBundleCmd.Flags().BoolVarP(&systemInfoFlag, "system-info", "S", true, "Adds system information to the debug bundle")
debugBundleCmd.Flags().BoolVarP(&uploadBundleFlag, "upload-bundle", "U", false, "Uploads the debug bundle to a server")
debugBundleCmd.Flags().StringVar(&uploadBundleURLFlag, "upload-bundle-url", types.DefaultBundleURL, "Service URL to get an URL to upload the debug bundle")
debugBundleCmd.Flags().BoolVar(&uploadBundleInsecureFlag, "upload-bundle-insecure", false, "Allow uploading to an http or untrusted-TLS upload server (self-hosted); requires root")
forCmd.Flags().Uint32VarP(&logFileCount, "log-file-count", "C", 1, "Number of rotated log files to include in debug bundle")
forCmd.Flags().BoolVarP(&systemInfoFlag, "system-info", "S", true, "Adds system information to the debug bundle")
forCmd.Flags().BoolVarP(&uploadBundleFlag, "upload-bundle", "U", false, "Uploads the debug bundle to a server")
forCmd.Flags().StringVar(&uploadBundleURLFlag, "upload-bundle-url", types.DefaultBundleURL, "Service URL to get an URL to upload the debug bundle")
forCmd.Flags().BoolVar(&uploadBundleInsecureFlag, "upload-bundle-insecure", false, "Allow uploading to an http or untrusted-TLS upload server (self-hosted); requires root")
forCmd.Flags().Bool("capture", false, "Capture packets during the debug duration and include in bundle")
}

View File

@@ -17,16 +17,26 @@ import (
"github.com/netbirdio/netbird/client/internal"
"github.com/netbirdio/netbird/client/internal/auth"
"github.com/netbirdio/netbird/client/internal/profilemanager"
nbnet "github.com/netbirdio/netbird/client/net"
"github.com/netbirdio/netbird/client/proto"
"github.com/netbirdio/netbird/client/server"
"github.com/netbirdio/netbird/client/system"
"github.com/netbirdio/netbird/util"
)
// extendSessionFlag drives the `netbird login --extend` flow: refresh the
// SSO session expiry on the management server without tearing down the
// tunnel. Mutually exclusive with setup-key login (a setup-key cannot
// refresh an SSO-tracked peer — see auth.errSetupKeyOnSSOExpiredPeer).
var extendSessionFlag bool
func init() {
loginCmd.PersistentFlags().BoolVar(&noBrowser, noBrowserFlag, false, noBrowserDesc)
loginCmd.PersistentFlags().BoolVar(&showQR, showQRFlag, false, showQRDesc)
loginCmd.PersistentFlags().StringVar(&profileName, profileNameFlag, "", profileNameDesc)
loginCmd.PersistentFlags().StringVarP(&configPath, "config", "c", "", "(DEPRECATED) Netbird config file location")
loginCmd.PersistentFlags().BoolVar(&extendSessionFlag, "extend", false,
"refresh the SSO session expiry without tearing down the tunnel (requires an active connection)")
}
var loginCmd = &cobra.Command{
@@ -61,6 +71,16 @@ var loginCmd = &cobra.Command{
return err
}
if extendSessionFlag {
if providedSetupKey != "" {
return fmt.Errorf("--extend cannot be combined with a setup key; setup keys can only enrol new peers")
}
if err := doExtendSession(ctx, cmd); err != nil {
return fmt.Errorf("extend session failed: %v", err)
}
return nil
}
// workaround to run without service
if util.FindFirstLogPath(logFiles) == "" {
if err := doForegroundLogin(ctx, cmd, providedSetupKey, activeProf); err != nil {
@@ -152,6 +172,65 @@ func doDaemonLogin(ctx context.Context, cmd *cobra.Command, providedSetupKey str
return nil
}
// doExtendSession drives the daemon's RequestExtendAuthSession /
// WaitExtendAuthSession pair. The user is sent through a regular SSO flow
// (browser + verification URL) and the resulting JWT is forwarded to the
// management server's ExtendAuthSession RPC. The tunnel stays up
// throughout — no Down/Up, no network-map resync.
func doExtendSession(ctx context.Context, cmd *cobra.Command) error {
conn, err := DialClientGRPCServer(ctx, daemonAddr)
if err != nil {
//nolint
return fmt.Errorf("failed to connect to daemon error: %v\n"+
"If the daemon is not running please run: "+
"\nnetbird service install \nnetbird service start\n", err)
}
defer conn.Close()
client := proto.NewDaemonServiceClient(conn)
req := &proto.RequestExtendAuthSessionRequest{}
// Pre-fill the IdP login hint from the active profile so the user
// doesn't have to retype their email. Best-effort: we still proceed
// without a hint if the lookup fails.
pm := profilemanager.NewProfileManager()
if active, perr := pm.GetActiveProfile(); perr == nil {
if profState, sperr := pm.GetProfileState(active.ID); sperr == nil && profState.Email != "" {
req.Hint = &profState.Email
}
}
startResp, err := client.RequestExtendAuthSession(ctx, req)
if err != nil {
return fmt.Errorf("start extend session: %v", err)
}
uri := startResp.GetVerificationURIComplete()
if uri == "" {
uri = startResp.GetVerificationURI()
}
openURL(cmd, uri, startResp.GetUserCode(), noBrowser, showQR)
waitResp, err := client.WaitExtendAuthSession(ctx, &proto.WaitExtendAuthSessionRequest{
DeviceCode: startResp.GetDeviceCode(),
UserCode: startResp.GetUserCode(),
})
if err != nil {
return fmt.Errorf("wait for extend session: %v", err)
}
if ts := waitResp.GetSessionExpiresAt(); ts.IsValid() && !ts.AsTime().IsZero() {
deadline := ts.AsTime().Local()
cmd.Printf("Session extended. New expiry: %s\n", deadline.Format("2006-01-02 15:04:05 MST"))
} else {
// Management reported the peer is not eligible (e.g. login
// expiration disabled on the account). Surface that fact
// instead of pretending the call succeeded.
cmd.Println("Session extension call completed, but the management server did not return a new deadline (peer may not be SSO-tracked or login expiration is disabled).")
}
return nil
}
func getActiveProfile(ctx context.Context, pm *profilemanager.ProfileManager, profileName string, username string) (*profilemanager.Profile, error) {
// switch profile if provided
@@ -254,6 +333,14 @@ func doForegroundLogin(ctx context.Context, cmd *cobra.Command, setupKey string,
return fmt.Errorf("read config file %s: %v", configFilePath, err)
}
// Mirror runInForegroundMode: recover residual state (DNS, firewall,
// ssh config, legacy routing) from a previous unclean shutdown and
// enable advanced routing before dialing management.
if err := server.RestoreResidualState(ctx, profilemanager.NewServiceManager(configFilePath).GetStatePath()); err != nil {
log.Warnf("failed to restore residual state: %v", err)
}
nbnet.Init()
err = foregroundLogin(ctx, cmd, config, setupKey, activeProf.ID)
if err != nil {
return fmt.Errorf("foreground login failed: %v", err)

View File

@@ -46,7 +46,7 @@ var logoutCmd = &cobra.Command{
}
if _, err := daemonClient.Logout(ctx, req); err != nil {
return fmt.Errorf("deregister: %v", err)
return daemonCallError("deregister", err)
}
cmd.Println("Deregistered successfully")

View File

@@ -20,7 +20,6 @@ import (
"github.com/spf13/cobra"
"github.com/spf13/pflag"
"google.golang.org/grpc"
"google.golang.org/grpc/credentials/insecure"
daddr "github.com/netbirdio/netbird/client/internal/daemonaddr"
"github.com/netbirdio/netbird/client/internal/profilemanager"
@@ -91,6 +90,7 @@ var (
// Don't resolve for service commands — they create the socket, not connect to it.
if !isServiceCmd(cmd) {
daemonAddr = daddr.ResolveUnixDaemonAddr(daemonAddr)
daemonAddr = daddr.ResolveDaemonAddr(daemonAddr)
}
return nil
},
@@ -143,10 +143,10 @@ func init() {
defaultDaemonAddr := "unix:///var/run/netbird.sock"
if runtime.GOOS == "windows" {
defaultDaemonAddr = "tcp://127.0.0.1:41731"
defaultDaemonAddr = daddr.WindowsPipeAddr
}
rootCmd.PersistentFlags().StringVar(&daemonAddr, "daemon-addr", defaultDaemonAddr, "Daemon service address to serve CLI requests [unix|tcp]://[path|host:port]")
rootCmd.PersistentFlags().StringVar(&daemonAddr, "daemon-addr", defaultDaemonAddr, "Daemon service address to serve CLI requests [unix|tcp|npipe]://[path|host:port|name]")
rootCmd.PersistentFlags().StringVarP(&managementURL, "management-url", "m", "", fmt.Sprintf("Management Service URL [http|https]://[host]:[port] (default \"%s\")", profilemanager.DefaultManagementURL))
rootCmd.PersistentFlags().StringVar(&adminURL, "admin-url", "", fmt.Sprintf("Admin Panel URL [http|https]://[host]:[port] (default \"%s\")", profilemanager.DefaultAdminURL))
rootCmd.PersistentFlags().StringVarP(&logLevel, "log-level", "l", "info", "sets NetBird log level")
@@ -269,12 +269,10 @@ func DialClientGRPCServer(ctx context.Context, addr string) (*grpc.ClientConn, e
ctx, cancel := context.WithTimeout(ctx, time.Second*10)
defer cancel()
return grpc.DialContext(
ctx,
strings.TrimPrefix(addr, "tcp://"),
grpc.WithTransportCredentials(insecure.NewCredentials()),
grpc.WithBlock(),
)
target, opts := daddr.DialTarget(addr)
opts = append(opts, grpc.WithBlock())
return grpc.DialContext(ctx, target, opts...)
}
// WithBackOff execute function in backoff cycle.

View File

@@ -5,6 +5,7 @@ package cmd
import (
"context"
"fmt"
"net/http"
"runtime"
"strings"
"sync"
@@ -22,15 +23,26 @@ var serviceCmd = &cobra.Command{
Short: "Manage the NetBird daemon service",
}
const defaultJSONSocket = "unix:///var/run/netbird-http.sock"
var (
serviceName string
serviceEnvVars []string
serviceName string
serviceEnvVars []string
jsonSocket string
enableJSONSocket bool
)
type program struct {
ctx context.Context
cancel context.CancelFunc
serv *grpc.Server
ctx context.Context
cancel context.CancelFunc
serv *grpc.Server
jsonServ *http.Server
// jsonClient is the gateway's own connection to the daemon. It is held so
// shutting the gateway down also closes it: nothing else references it once
// the handlers are registered, so its transport goroutines would otherwise
// outlive the server.
jsonClient *grpc.ClientConn
jsonServMu sync.Mutex
serverInstance *server.Server
serverInstanceMu sync.Mutex
}
@@ -46,6 +58,8 @@ func init() {
serviceCmd.PersistentFlags().BoolVar(&updateSettingsDisabled, "disable-update-settings", false, "Disables update settings feature. If enabled, the client will not be able to change or edit any settings. To persist this setting, use: netbird service install --disable-update-settings")
serviceCmd.PersistentFlags().BoolVar(&captureEnabled, "enable-capture", false, "Enables packet capture via 'netbird debug capture'. To persist, use: netbird service install --enable-capture")
serviceCmd.PersistentFlags().BoolVar(&networksDisabled, "disable-networks", false, "Disables network selection. If enabled, the client will not allow listing, selecting, or deselecting networks. To persist, use: netbird service install --disable-networks")
serviceCmd.PersistentFlags().BoolVar(&enableJSONSocket, "enable-json-socket", false, "Enables the HTTP/JSON API socket served by grpc-gateway. To persist, use: netbird service install --enable-json-socket")
serviceCmd.PersistentFlags().StringVar(&jsonSocket, "json-socket", defaultJSONSocket, "HTTP/JSON API socket address [unix|tcp]://[path|host:port]. Requires --enable-json-socket to serve. To persist, use: netbird service install --enable-json-socket --json-socket")
rootCmd.PersistentFlags().StringVarP(&serviceName, "service", "s", defaultServiceName, "Netbird system service name")
serviceEnvDesc := `Sets extra environment variables for the service. ` +

View File

@@ -5,9 +5,7 @@ package cmd
import (
"context"
"fmt"
"net"
"os"
"strings"
"runtime"
"time"
"github.com/kardianos/service"
@@ -16,69 +14,157 @@ import (
"github.com/spf13/cobra"
"google.golang.org/grpc"
"github.com/netbirdio/netbird/client/internal/daemonaddr"
"github.com/netbirdio/netbird/client/internal/ipcauth"
"github.com/netbirdio/netbird/client/proto"
"github.com/netbirdio/netbird/client/server"
"github.com/netbirdio/netbird/client/system"
"github.com/netbirdio/netbird/util"
)
func validateJSONSocketFlags() error {
if serviceCmd.PersistentFlags().Changed("json-socket") && !enableJSONSocket {
return fmt.Errorf("--json-socket requires --enable-json-socket to configure the daemon JSON gateway")
}
return nil
}
// daemonServerOptions installs the transport credentials that expose each
// caller's kernel-authenticated identity to the handlers, which is what lets
// the daemon require root/administrator for privileged operations.
//
// The handshake exchanges no bytes, so older CLI and UI binaries still
// interoperate. Callers on a TCP socket carry no identity at all: the daemon
// keeps serving them, and the privileged operations deny them, so a warning is
// logged to make the loss of functionality visible.
func daemonServerOptions(network string) []grpc.ServerOption {
if network == "tcp" {
log.Warnf("daemon is listening on TCP (%s): callers carry no verifiable identity over TCP, "+
"so privileged operations (SSH root login, SSH auth, enabling the SSH server, management URL changes, "+
"deregistration) will be denied. Use a unix socket, or npipe:// on Windows", daemonAddr)
return nil
}
creds := ipcauth.NewTransportCredentials()
if creds == nil {
log.Warnf("daemon IPC has no peer-identity primitive on %s: privileged operations will be denied", runtime.GOOS)
return nil
}
return []grpc.ServerOption{grpc.Creds(creds)}
}
func (p *program) Start(svc service.Service) error {
// Start should not block. Do the actual work async.
log.Info("starting NetBird service") //nolint
if err := validateJSONSocketFlags(); err != nil {
return err
}
// Collect static system and platform information
system.UpdateStaticInfoAsync()
// in any case, even if configuration does not exists we run daemon to serve CLI gRPC API.
p.serv = grpc.NewServer()
split := strings.Split(daemonAddr, "://")
switch split[0] {
case "unix":
// cleanup failed close
stat, err := os.Stat(split[1])
if err == nil && !stat.IsDir() {
if err := os.Remove(split[1]); err != nil {
log.Debugf("remove socket file: %v", err)
}
}
case "tcp":
default:
return fmt.Errorf("unsupported daemon address protocol: %v", split[0])
// A daemon installed before named-pipe support has the loopback TCP address
// persisted. Move it to the named pipe so an upgraded daemon can identify
// its callers instead of silently serving an unauthenticated socket.
if migrated, ok := daemonaddr.MigrateLegacy(daemonAddr); ok {
log.Infof("daemon address %q predates named-pipe support, listening on %q so callers can be identified", daemonAddr, migrated)
daemonAddr = migrated
}
listen, err := net.Listen(split[0], split[1])
network, _, err := parseListenAddress(daemonAddr)
if err != nil {
return fmt.Errorf("listen daemon interface: %w", err)
return fmt.Errorf("parse daemon address: %w", err)
}
// in any case, even if configuration does not exists we run daemon to serve CLI gRPC API.
p.serv = grpc.NewServer(daemonServerOptions(network)...)
daemonListener, jsonListener, err := listenDaemonSockets()
if err != nil {
return err
}
go func() {
defer listen.Close()
if split[0] == "unix" {
if err := os.Chmod(split[1], 0666); err != nil {
log.Errorf("failed setting daemon permissions: %v", split[1])
return
}
}
serverInstance := server.New(p.ctx, util.FindFirstLogPath(logFiles), configPath, profilesDisabled, updateSettingsDisabled, captureEnabled, networksDisabled)
if err := serverInstance.Start(); err != nil {
log.Fatalf("failed to start daemon: %v", err)
}
proto.RegisterDaemonServiceServer(p.serv, serverInstance)
p.serverInstanceMu.Lock()
p.serverInstance = serverInstance
p.serverInstanceMu.Unlock()
log.Printf("started daemon server: %v", split[1])
if err := p.serv.Serve(listen); err != nil {
log.Errorf("failed to serve daemon requests: %v", err)
// Fatal here rather than inside serve, so serve's deferred listener
// closes run before the process exits.
if err := p.serve(daemonListener, jsonListener); err != nil {
log.Fatalf("failed to %v", err)
}
}()
return nil
}
// listenDaemonSockets opens the daemon control socket and, when it is enabled, the
// JSON gateway socket. The control socket is closed again if the second one fails,
// so a failed start leaves nothing listening. The returned JSON listener is nil
// when the socket is disabled.
func listenDaemonSockets() (*socketListener, *socketListener, error) {
daemonListener, err := listenOnAddress(daemonAddr)
if err != nil {
return nil, nil, fmt.Errorf("listen daemon interface: %w", err)
}
if !enableJSONSocket {
removeStaleUnixSocketForAddress(jsonSocket)
return daemonListener, nil, nil
}
jsonListener, err := listenOnAddress(jsonSocket)
if err != nil {
if cerr := daemonListener.Close(); cerr != nil {
log.Debugf("close daemon listener: %v", cerr)
}
return nil, nil, fmt.Errorf("listen daemon JSON interface: %w", err)
}
return daemonListener, jsonListener, nil
}
// serve brings up the daemon server on an already-open control socket and blocks
// until it stops. jsonListener is nil when the JSON socket is disabled. A returned
// error means the daemon cannot run at all and the caller is expected to exit; the
// failures it recovers from on its own are logged here.
func (p *program) serve(daemonListener, jsonListener *socketListener) error {
defer daemonListener.Close()
if jsonListener != nil {
defer jsonListener.Close()
}
// chmodUnixSocket is a no-op for a nil listener and for a non-unix one.
if err := daemonListener.chmodUnixSocket("daemon"); err != nil {
log.Error(err)
return nil
}
if err := jsonListener.chmodUnixSocket("daemon JSON"); err != nil {
log.Error(err)
return nil
}
serverInstance := server.New(p.ctx, util.FindFirstLogPath(logFiles), configPath, profilesDisabled, updateSettingsDisabled, captureEnabled, networksDisabled)
if err := serverInstance.Start(); err != nil {
return fmt.Errorf("start daemon: %w", err)
}
proto.RegisterDaemonServiceServer(p.serv, serverInstance)
p.serverInstanceMu.Lock()
p.serverInstance = serverInstance
p.serverInstanceMu.Unlock()
if jsonListener == nil {
log.Debug("daemon JSON socket disabled")
} else if err := p.startJSONGateway(jsonListener, daemonAddr); err != nil {
return fmt.Errorf("start daemon JSON server: %w", err)
}
log.Printf("started daemon server: %v", daemonListener.address)
if err := p.serv.Serve(daemonListener.Listener); err != nil {
log.Errorf("failed to serve daemon requests: %v", err)
}
return nil
}
func (p *program) Stop(srv service.Service) error {
p.serverInstanceMu.Lock()
if p.serverInstance != nil {
@@ -92,6 +178,25 @@ func (p *program) Stop(srv service.Service) error {
p.cancel()
p.jsonServMu.Lock()
jsonServ, jsonClient := p.jsonServ, p.jsonClient
p.jsonServMu.Unlock()
if jsonClient != nil {
if err := jsonClient.Close(); err != nil {
log.Debugf("close daemon JSON gateway client: %v", err)
}
}
if jsonServ != nil {
shutdownCtx, shutdownCancel := context.WithTimeout(context.Background(), 2*time.Second)
if err := jsonServ.Shutdown(shutdownCtx); err != nil {
log.Errorf("failed to stop daemon JSON server gracefully: %v", err)
if err := jsonServ.Close(); err != nil {
log.Errorf("failed to close daemon JSON server: %v", err)
}
}
shutdownCancel()
}
if p.serv != nil {
p.serv.Stop()
}
@@ -148,6 +253,9 @@ var runCmd = &cobra.Command{
if err != nil {
return err
}
if err := validateJSONSocketFlags(); err != nil {
return err
}
return s.Run()
},
@@ -162,6 +270,9 @@ var startCmd = &cobra.Command{
if err != nil {
return err
}
if err := validateJSONSocketFlags(); err != nil {
return err
}
if err := s.Start(); err != nil {
return fmt.Errorf("start service: %w", err)
@@ -198,6 +309,9 @@ var restartCmd = &cobra.Command{
if err != nil {
return err
}
if err := validateJSONSocketFlags(); err != nil {
return err
}
if err := s.Restart(); err != nil {
return fmt.Errorf("restart service: %w", err)

View File

@@ -67,6 +67,10 @@ func buildServiceArguments() []string {
args = append(args, "--disable-networks")
}
if enableJSONSocket {
args = append(args, "--enable-json-socket", "--json-socket", jsonSocket)
}
return args
}
@@ -106,6 +110,10 @@ func configurePlatformSpecificSettings(svcConfig *service.Config) error {
// Create fully configured service config for install/reconfigure
func createServiceConfigForInstall() (*service.Config, error) {
if err := validateJSONSocketFlags(); err != nil {
return nil, err
}
svcConfig, err := newSVCConfig()
if err != nil {
return nil, fmt.Errorf("create service config: %w", err)

View File

@@ -0,0 +1,150 @@
//go:build !ios && !android
package cmd
import (
"context"
"errors"
"fmt"
"net"
"net/http"
"sync"
"time"
"github.com/grpc-ecosystem/grpc-gateway/v2/runtime"
log "github.com/sirupsen/logrus"
"google.golang.org/grpc"
"github.com/netbirdio/netbird/client/internal/daemonaddr"
"github.com/netbirdio/netbird/client/internal/ipcauth"
"github.com/netbirdio/netbird/client/proto"
)
// jsonPeerIdentity is the context key under which the connecting HTTP client's
// identity is stashed for the lifetime of its connection.
type jsonPeerIdentity struct{}
// jsonPeerIdentityValue pairs the identity with whether it could be read at
// all, so an unreadable identity is forwarded as "unknown" rather than omitted.
type jsonPeerIdentityValue struct {
id ipcauth.Identity
known bool
}
// jsonConnContext reads the identity of the client connecting to the JSON
// socket and stashes it on the connection's context. The gateway re-dials the
// daemon in-process, so the daemon would otherwise see every JSON request as
// coming from the daemon itself.
func jsonConnContext(ctx context.Context, c net.Conn) context.Context {
value := jsonPeerIdentityValue{}
id, err := ipcauth.ConnIdentity(c)
if err != nil {
log.Warnf("json gateway: cannot read HTTP client identity, privileged operations will be denied for this connection: %v", err)
} else {
value.id = id
value.known = true
}
return context.WithValue(ctx, jsonPeerIdentity{}, value)
}
// forwardIdentity stamps the HTTP client's identity onto every call the gateway
// makes to the daemon.
//
// It is an interceptor on the gateway's client connection rather than a
// runtime.WithMetadata annotator because grpc-gateway skips annotators when no
// request header maps to metadata, which an HTTP/1.0 request with no Host header
// over a unix socket achieves. The daemon would then receive no marker, see its own
// identity as the transport peer, and authorize the request as the daemon itself.
// An interceptor runs for every RPC whatever the request looked like.
func forwardIdentity(ctx context.Context) context.Context {
value, ok := ctx.Value(jsonPeerIdentity{}).(jsonPeerIdentityValue)
if !ok {
// No ConnContext ran for this request, so forward an unknown identity:
// the daemon must not mistake its own identity for the client's.
return ipcauth.WithForwardedIdentity(ctx, ipcauth.Identity{}, false)
}
return ipcauth.WithForwardedIdentity(ctx, value.id, value.known)
}
func forwardIdentityUnary(ctx context.Context, method string, req, reply any, cc *grpc.ClientConn, invoker grpc.UnaryInvoker, opts ...grpc.CallOption) error {
return invoker(forwardIdentity(ctx), method, req, reply, cc, opts...)
}
func forwardIdentityStream(ctx context.Context, desc *grpc.StreamDesc, cc *grpc.ClientConn, method string, streamer grpc.Streamer, opts ...grpc.CallOption) (grpc.ClientStream, error) {
return streamer(forwardIdentity(ctx), desc, cc, method, opts...)
}
// reservedHeaderWarning limits the dropped-header warning to the first occurrence.
var reservedHeaderWarning sync.Once
// jsonIncomingHeaderMatcher keeps an HTTP client from supplying the metadata the
// gateway uses to forward its identity. grpc-gateway turns "Grpc-Metadata-<key>"
// headers into gRPC metadata and joins them ahead of what its annotators add, so
// without this filter a JSON client could send its own x-netbird-fwd-uid and the
// daemon would authorize that instead of the client's real identity.
func jsonIncomingHeaderMatcher(key string) (string, bool) {
mapped, ok := runtime.DefaultHeaderMatcher(key)
if !ok {
return "", false
}
if ipcauth.IsReservedForwardKey(mapped) {
// Warn once: any client can send these on every request, so warning each
// time hands it a way to fill the log. The rest are debug-level.
reservedHeaderWarning.Do(func() {
log.Warnf("json gateway: dropping reserved header %q from a request: only the gateway may set the caller's identity", key)
})
log.Debugf("json gateway: dropping reserved header %q", key)
return "", false
}
return mapped, true
}
func (p *program) startJSONGateway(jsonListener *socketListener, daemonEndpoint string) error {
if jsonListener.network == "tcp" {
log.Warnf("daemon JSON socket is listening on TCP (%s): callers carry no verifiable identity over TCP, "+
"so privileged operations will be denied for JSON clients", jsonListener.address)
}
mux := runtime.NewServeMux(runtime.WithIncomingHeaderMatcher(jsonIncomingHeaderMatcher))
// grpc.NewClient does not connect until the first request, so registering
// the handler here cannot block daemon startup.
target, opts := daemonaddr.DialTarget(daemonEndpoint)
opts = append(opts,
grpc.WithChainUnaryInterceptor(forwardIdentityUnary),
grpc.WithChainStreamInterceptor(forwardIdentityStream),
)
conn, err := grpc.NewClient(target, opts...)
if err != nil {
return fmt.Errorf("create daemon client for JSON gateway: %w", err)
}
if err := proto.RegisterDaemonServiceHandler(p.ctx, mux, conn); err != nil {
if cerr := conn.Close(); cerr != nil {
log.Debugf("close daemon client after failed JSON gateway registration: %v", cerr)
}
return err
}
jsonServer := &http.Server{
Handler: mux,
ReadHeaderTimeout: 5 * time.Second,
BaseContext: func(net.Listener) context.Context {
return p.ctx
},
ConnContext: jsonConnContext,
}
p.jsonServMu.Lock()
p.jsonServ = jsonServer
p.jsonClient = conn
p.jsonServMu.Unlock()
go func() {
log.Printf("started daemon JSON server: %v", jsonListener.address)
if err := jsonServer.Serve(jsonListener.Listener); err != nil && !errors.Is(err, http.ErrServerClosed) {
log.Errorf("failed to serve daemon JSON requests: %v", err)
}
}()
return nil
}

View File

@@ -0,0 +1,261 @@
//go:build !windows && !ios && !android
package cmd
import (
"context"
"net"
"net/http"
"path/filepath"
"testing"
"time"
"github.com/grpc-ecosystem/grpc-gateway/v2/runtime"
"google.golang.org/grpc/credentials"
"google.golang.org/grpc/metadata"
"google.golang.org/grpc/peer"
"github.com/netbirdio/netbird/client/internal/ipcauth"
)
// The JSON gateway runs inside the daemon and re-dials it locally, so every JSON
// request reaches a handler with the daemon's own identity as the transport peer.
// The gateway therefore forwards its HTTP client's identity as metadata, and the
// daemon authorizes that instead of itself. These tests drive the real wiring
// (jsonConnContext, forwardIdentity, jsonIncomingHeaderMatcher) and check the
// identity a handler would end up authorizing.
// daemonSideCtx is what a handler sees for a gateway-relayed call. The transport
// peer must be this process's own identity: the gateway is the daemon, so the two
// cannot differ, and hardcoding root here instead would describe a state that
// never occurs.
func daemonSideCtx(t *testing.T, md metadata.MD) context.Context {
t.Helper()
self, err := ipcauth.CurrentProcessIdentity()
if err != nil {
t.Skipf("cannot read this process's identity: %v", err)
}
ctx := peer.NewContext(context.Background(), &peer.Peer{
AuthInfo: ipcauth.AuthInfo{
CommonAuthInfo: credentials.CommonAuthInfo{SecurityLevel: credentials.NoSecurity},
Identity: self,
},
})
return metadata.NewIncomingContext(ctx, md)
}
// gatewayMetadata reproduces what the daemon receives for a JSON request: the
// mux annotates the context from the request's headers, then the interceptor on the
// gateway's client connection stamps the caller's identity. The order matters,
// since the interceptor must win over anything a header put there.
func gatewayMetadata(t *testing.T, req *http.Request, ctx context.Context) metadata.MD {
t.Helper()
mux := runtime.NewServeMux(runtime.WithIncomingHeaderMatcher(jsonIncomingHeaderMatcher))
annotated, err := runtime.AnnotateContext(ctx, mux, req,
"/daemon.DaemonService/SetConfig",
runtime.WithHTTPPathPattern("/daemon.DaemonService/SetConfig"))
if err != nil {
t.Fatalf("annotate: %v", err)
}
md, ok := metadata.FromOutgoingContext(forwardIdentity(annotated))
if !ok {
t.Fatal("the interceptor produced no metadata")
}
return md
}
// clientCtx is the connection context jsonConnContext would have produced for an
// HTTP client whose identity the gateway could read.
func clientCtx(id ipcauth.Identity, known bool) context.Context {
return context.WithValue(context.Background(), jsonPeerIdentity{},
jsonPeerIdentityValue{id: id, known: known})
}
// An HTTP client must not be able to name its own identity. grpc-gateway turns
// Grpc-Metadata-<key> headers into gRPC metadata, so without the header filter and
// the interceptor overwriting the reserved keys, this request would authorize as
// uid 0.
func TestJSONGateway_ForgedIdentityHeaderIsDropped(t *testing.T) {
req, err := http.NewRequest(http.MethodPost, "http://localhost/daemon.DaemonService/SetConfig", nil)
if err != nil {
t.Fatal(err)
}
req.Header.Set("Grpc-Metadata-X-Netbird-Fwd-Uid", "0")
req.Header.Set("Grpc-Metadata-X-Netbird-Fwd-Gid", "0")
req.Header.Set("Grpc-Metadata-X-Netbird-Fwd", "1")
req.Header.Set("Grpc-Metadata-X-Netbird-Fwd-Sid", "S-1-5-18")
caller := ipcauth.Identity{UID: 31000, GID: 31000}
md := gatewayMetadata(t, req, clientCtx(caller, true))
id, ok := ipcauth.CallerIdentity(daemonSideCtx(t, md))
if !ok {
t.Fatal("the forwarded identity should be usable")
}
if id.IsPrivileged() {
t.Errorf("forged header was believed: authorized as %v", id)
}
if id.UID != caller.UID {
t.Errorf("authorized as uid %d, want the real client %d", id.UID, caller.UID)
}
}
// A request with no headers at all (HTTP/1.0 needs no Host, and a unix socket
// yields no host:port) makes grpc-gateway produce no metadata whatsoever and skip
// its annotators: "if len(pairs) == 0 { return ctx, nil, nil }" in
// runtime/context.go. That is why the identity is stamped by an interceptor
// instead. This is the case that previously reached the gate as the daemon itself.
func TestJSONGateway_HeaderlessRequestIsStillMarkedForwarded(t *testing.T) {
req, err := http.NewRequest(http.MethodPost, "http://localhost/daemon.DaemonService/SetConfig", nil)
if err != nil {
t.Fatal(err)
}
req.Header = http.Header{}
req.Host = ""
caller := ipcauth.Identity{UID: 31000, GID: 31000}
ctx := clientCtx(caller, true)
// Pin the skip path itself: if grpc-gateway ever produced a pair here, this
// test would still pass below while no longer covering what it was written for.
mux := runtime.NewServeMux(runtime.WithIncomingHeaderMatcher(jsonIncomingHeaderMatcher))
annotated, err := runtime.AnnotateContext(ctx, mux, req,
"/daemon.DaemonService/SetConfig",
runtime.WithHTTPPathPattern("/daemon.DaemonService/SetConfig"))
if err != nil {
t.Fatalf("annotate: %v", err)
}
if md, ok := metadata.FromOutgoingContext(annotated); ok {
t.Fatalf("grpc-gateway produced metadata %v for a headerless request; "+
"this test no longer covers the annotator-skip path", md)
}
md := gatewayMetadata(t, req, ctx)
id, ok := ipcauth.CallerIdentity(daemonSideCtx(t, md))
if !ok {
t.Fatal("the forwarded identity should be usable")
}
if id.UID != caller.UID || id.IsPrivileged() {
t.Errorf("authorized as %v, want the real client uid %d", id, caller.UID)
}
}
// When the gateway cannot read its client's identity (a TCP JSON socket, say) it
// forwards the marker alone. The daemon must then report "unidentified" so the
// privileged operations refuse, rather than falling back to the gateway's own
// identity.
func TestJSONGateway_UnreadableClientIdentityIsUnidentified(t *testing.T) {
req, err := http.NewRequest(http.MethodPost, "http://localhost/daemon.DaemonService/SetConfig", nil)
if err != nil {
t.Fatal(err)
}
md := gatewayMetadata(t, req, clientCtx(ipcauth.Identity{}, false))
if id, ok := ipcauth.CallerIdentity(daemonSideCtx(t, md)); ok {
t.Errorf("a request with no client identity was authorized as %v", id)
}
}
// A request that never passed through jsonConnContext (no stashed identity) must
// also come out unidentified rather than as the daemon.
func TestJSONGateway_MissingConnContextIsUnidentified(t *testing.T) {
req, err := http.NewRequest(http.MethodPost, "http://localhost/daemon.DaemonService/SetConfig", nil)
if err != nil {
t.Fatal(err)
}
md := gatewayMetadata(t, req, context.Background())
if id, ok := ipcauth.CallerIdentity(daemonSideCtx(t, md)); ok {
t.Errorf("a request with no connection context was authorized as %v", id)
}
}
// End to end over a real unix socket: the gateway reads the connecting client's
// identity from the socket itself, so a client cannot present anything else.
func TestJSONGateway_IdentityComesFromTheSocket(t *testing.T) {
mux := runtime.NewServeMux(runtime.WithIncomingHeaderMatcher(jsonIncomingHeaderMatcher))
type observed struct {
md metadata.MD
}
seen := make(chan observed, 1)
srv := &http.Server{
Handler: http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
ctx, err := runtime.AnnotateContext(r.Context(), mux, r,
"/daemon.DaemonService/SetConfig",
runtime.WithHTTPPathPattern("/daemon.DaemonService/SetConfig"))
if err != nil {
t.Errorf("annotate: %v", err)
return
}
md, _ := metadata.FromOutgoingContext(forwardIdentity(ctx))
seen <- observed{md: md}
w.WriteHeader(http.StatusOK)
}),
ReadHeaderTimeout: 5 * time.Second,
ConnContext: jsonConnContext,
}
sock := filepath.Join(t.TempDir(), "http.sock")
ln, err := net.Listen("unix", sock)
if err != nil {
t.Fatal(err)
}
t.Cleanup(func() {
if err := srv.Close(); err != nil {
t.Logf("close server: %v", err)
}
})
go func() {
if err := srv.Serve(ln); err != nil && err != http.ErrServerClosed {
t.Logf("serve: %v", err)
}
}()
conn, err := net.Dial("unix", sock)
if err != nil {
t.Fatal(err)
}
t.Cleanup(func() {
if err := conn.Close(); err != nil {
t.Logf("close conn: %v", err)
}
})
// Forge the identity headers on the wire as well.
request := "POST /daemon.DaemonService/SetConfig HTTP/1.1\r\n" +
"Host: localhost\r\n" +
"Grpc-Metadata-X-Netbird-Fwd: 1\r\n" +
"Grpc-Metadata-X-Netbird-Fwd-Uid: 0\r\n" +
"Content-Length: 0\r\n\r\n"
if _, err := conn.Write([]byte(request)); err != nil {
t.Fatal(err)
}
select {
case got := <-seen:
self, err := ipcauth.CurrentProcessIdentity()
if err != nil {
t.Skipf("cannot read this process's identity: %v", err)
}
// The socket peer is this test process, so that is the identity the
// gateway must forward, not the uid 0 the request asked for.
if uids := got.md.Get("x-netbird-fwd-uid"); len(uids) != 1 {
t.Fatalf("x-netbird-fwd-uid = %v, want exactly the gateway's own value", uids)
}
id, ok := ipcauth.CallerIdentity(daemonSideCtx(t, got.md))
if !ok {
t.Fatal("the forwarded identity should be usable")
}
if id.UID != self.UID {
t.Errorf("authorized as uid %d, want the socket peer %d", id.UID, self.UID)
}
case <-time.After(5 * time.Second):
t.Fatal("the gateway never handled the request")
}
}

View File

@@ -0,0 +1,176 @@
//go:build !ios && !android
package cmd
import (
"net"
"os"
"path/filepath"
"runtime"
"testing"
"github.com/spf13/cobra"
"github.com/spf13/pflag"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func preserveJSONSocketTestState(t *testing.T) {
t.Helper()
origJSONSocket := jsonSocket
origEnableJSONSocket := enableJSONSocket
origChanged := map[string]bool{}
serviceCmd.PersistentFlags().VisitAll(func(flag *pflag.Flag) {
origChanged[flag.Name] = flag.Changed
})
t.Cleanup(func() {
jsonSocket = origJSONSocket
enableJSONSocket = origEnableJSONSocket
serviceCmd.PersistentFlags().VisitAll(func(flag *pflag.Flag) {
flag.Changed = origChanged[flag.Name]
})
})
}
func TestJSONSocketFlagsArePositiveEnableOnly(t *testing.T) {
assert.NotNil(t, serviceCmd.PersistentFlags().Lookup("enable-json-socket"))
assert.NotNil(t, serviceCmd.PersistentFlags().Lookup("json-socket"))
assert.Nil(t, serviceCmd.PersistentFlags().Lookup("disable-json-socket"))
assert.Equal(t, "false", serviceCmd.PersistentFlags().Lookup("enable-json-socket").DefValue)
}
func TestBuildServiceArgumentsDefaultDisablesJSONSocket(t *testing.T) {
preserveJSONSocketTestState(t)
enableJSONSocket = false
jsonSocket = "tcp://127.0.0.1:8080"
args := buildServiceArguments()
assert.NotContains(t, args, "--enable-json-socket")
assert.NotContains(t, args, "--json-socket")
}
func TestBuildServiceArgumentsIncludesJSONSocketWhenEnabled(t *testing.T) {
preserveJSONSocketTestState(t)
enableJSONSocket = true
jsonSocket = "tcp://127.0.0.1:8080"
args := buildServiceArguments()
enableIndex := indexOfArg(args, "--enable-json-socket")
jsonIndex := indexOfArg(args, "--json-socket")
require.NotEqual(t, -1, enableIndex)
require.NotEqual(t, -1, jsonIndex)
require.Less(t, enableIndex, jsonIndex)
require.Less(t, jsonIndex+1, len(args))
assert.Equal(t, "tcp://127.0.0.1:8080", args[jsonIndex+1])
}
func TestJSONSocketWithoutEnableValidation(t *testing.T) {
preserveJSONSocketTestState(t)
enableJSONSocket = false
require.NoError(t, serviceCmd.PersistentFlags().Set("json-socket", "tcp://127.0.0.1:8080"))
err := validateJSONSocketFlags()
require.Error(t, err)
assert.Contains(t, err.Error(), "--enable-json-socket")
}
func TestJSONSocketWithEnableValidation(t *testing.T) {
preserveJSONSocketTestState(t)
require.NoError(t, serviceCmd.PersistentFlags().Set("enable-json-socket", "true"))
require.NoError(t, serviceCmd.PersistentFlags().Set("json-socket", "tcp://127.0.0.1:8080"))
assert.NoError(t, validateJSONSocketFlags())
}
func TestJSONSocketServiceParamsPersistEnableAndAddress(t *testing.T) {
preserveJSONSocketTestState(t)
serviceCmd.PersistentFlags().VisitAll(func(flag *pflag.Flag) {
flag.Changed = false
})
enableJSONSocket = true
jsonSocket = "tcp://127.0.0.1:8080"
params := currentServiceParams()
require.True(t, params.EnableJSONSocket)
require.Equal(t, "tcp://127.0.0.1:8080", params.JSONSocket)
enableJSONSocket = false
jsonSocket = defaultJSONSocket
applyServiceParams(testServiceEnvCommand(), params)
assert.True(t, enableJSONSocket)
assert.Equal(t, "tcp://127.0.0.1:8080", jsonSocket)
}
func TestRemoveStaleUnixSocketDoesNotRemoveRegularFile(t *testing.T) {
path := filepath.Join(t.TempDir(), "netbird-http.sock")
require.NoError(t, os.WriteFile(path, []byte("not a socket"), 0600))
removeStaleUnixSocket(path)
data, err := os.ReadFile(path)
require.NoError(t, err)
assert.Equal(t, []byte("not a socket"), data)
}
func TestRemoveStaleUnixSocketRemovesSocket(t *testing.T) {
if runtime.GOOS == "windows" {
t.Skip("unix sockets are not available on Windows")
}
path := filepath.Join(t.TempDir(), "netbird-http.sock")
addr := &net.UnixAddr{Name: path, Net: "unix"}
listener, err := net.ListenUnix("unix", addr)
require.NoError(t, err)
listener.SetUnlinkOnClose(false)
require.NoError(t, listener.Close())
_, err = os.Lstat(path)
require.NoError(t, err, "test setup must leave a stale Unix socket path")
removeStaleUnixSocket(path)
_, err = os.Lstat(path)
assert.True(t, os.IsNotExist(err), "expected stale Unix socket to be removed, got %v", err)
}
func TestRemoveStaleUnixSocketDoesNotRemoveLiveSocket(t *testing.T) {
if runtime.GOOS == "windows" {
t.Skip("unix sockets are not available on Windows")
}
path := filepath.Join(t.TempDir(), "netbird-http.sock")
listener, err := net.Listen("unix", path)
require.NoError(t, err)
defer listener.Close()
removeStaleUnixSocket(path)
_, err = os.Lstat(path)
assert.NoError(t, err, "expected live Unix socket to be preserved")
}
func testServiceEnvCommand() *cobra.Command {
cmd := &cobra.Command{}
cmd.Flags().StringSlice("service-env", nil, "")
return cmd
}
func indexOfArg(args []string, arg string) int {
for i, candidate := range args {
if candidate == arg {
return i
}
}
return -1
}

View File

@@ -13,6 +13,7 @@ import (
"github.com/spf13/cobra"
"github.com/netbirdio/netbird/client/configs"
"github.com/netbirdio/netbird/client/internal/daemonaddr"
"github.com/netbirdio/netbird/util"
)
@@ -23,6 +24,7 @@ const serviceParamsFile = "service.json"
type serviceParams struct {
LogLevel string `json:"log_level"`
DaemonAddr string `json:"daemon_addr"`
JSONSocket string `json:"json_socket"`
ManagementURL string `json:"management_url,omitempty"`
ConfigPath string `json:"config_path,omitempty"`
LogFiles []string `json:"log_files,omitempty"`
@@ -30,6 +32,7 @@ type serviceParams struct {
DisableUpdateSettings bool `json:"disable_update_settings,omitempty"`
EnableCapture bool `json:"enable_capture,omitempty"`
DisableNetworks bool `json:"disable_networks,omitempty"`
EnableJSONSocket bool `json:"enable_json_socket,omitempty"`
ServiceEnvVars map[string]string `json:"service_env_vars,omitempty"`
}
@@ -75,6 +78,7 @@ func currentServiceParams() *serviceParams {
params := &serviceParams{
LogLevel: logLevel,
DaemonAddr: daemonAddr,
JSONSocket: jsonSocket,
ManagementURL: managementURL,
ConfigPath: configPath,
LogFiles: logFiles,
@@ -82,6 +86,7 @@ func currentServiceParams() *serviceParams {
DisableUpdateSettings: updateSettingsDisabled,
EnableCapture: captureEnabled,
DisableNetworks: networksDisabled,
EnableJSONSocket: enableJSONSocket,
}
if len(serviceEnvVars) > 0 {
@@ -113,15 +118,29 @@ func applyServiceParams(cmd *cobra.Command, params *serviceParams) {
return
}
// For fields with non-empty defaults (log-level, daemon-addr), keep the
// != "" guard so that an older service.json missing the field doesn't
// clobber the default with an empty string.
// For fields with non-empty defaults, keep the != "" guard so that an older
// service.json missing the field doesn't clobber the default with an empty string.
if !rootCmd.PersistentFlags().Changed("log-level") && params.LogLevel != "" {
logLevel = params.LogLevel
}
if !rootCmd.PersistentFlags().Changed("daemon-addr") && params.DaemonAddr != "" {
daemonAddr = params.DaemonAddr
// An install that predates named-pipe support has the loopback TCP
// address saved. Callers carry no identity over TCP, so move it to the
// pipe instead of restoring a socket the daemon cannot authorize on.
if migrated, ok := daemonaddr.MigrateLegacy(daemonAddr); ok {
cmd.Printf("Moving the saved daemon address from %s to %s so the daemon can identify its callers\n", daemonAddr, migrated)
daemonAddr = migrated
}
}
if !serviceCmd.PersistentFlags().Changed("json-socket") && params.JSONSocket != "" {
jsonSocket = params.JSONSocket
}
if !serviceCmd.PersistentFlags().Changed("enable-json-socket") {
enableJSONSocket = params.EnableJSONSocket
}
// For optional fields where empty means "use default", always apply so

View File

@@ -41,6 +41,8 @@ func TestSaveAndLoadServiceParams(t *testing.T) {
params := &serviceParams{
LogLevel: "debug",
DaemonAddr: "unix:///var/run/netbird.sock",
JSONSocket: "tcp://127.0.0.1:8080",
EnableJSONSocket: true,
ManagementURL: "https://my.server.com",
ConfigPath: "/etc/netbird/config.json",
LogFiles: []string{"/var/log/netbird/client.log", "console"},
@@ -63,6 +65,8 @@ func TestSaveAndLoadServiceParams(t *testing.T) {
assert.Equal(t, params.LogLevel, loaded.LogLevel)
assert.Equal(t, params.DaemonAddr, loaded.DaemonAddr)
assert.Equal(t, params.JSONSocket, loaded.JSONSocket)
assert.Equal(t, params.EnableJSONSocket, loaded.EnableJSONSocket)
assert.Equal(t, params.ManagementURL, loaded.ManagementURL)
assert.Equal(t, params.ConfigPath, loaded.ConfigPath)
assert.Equal(t, params.LogFiles, loaded.LogFiles)
@@ -101,6 +105,8 @@ func TestLoadServiceParams_InvalidJSON(t *testing.T) {
func TestCurrentServiceParams(t *testing.T) {
origLogLevel := logLevel
origDaemonAddr := daemonAddr
origJSONSocket := jsonSocket
origEnableJSONSocket := enableJSONSocket
origManagementURL := managementURL
origConfigPath := configPath
origLogFiles := logFiles
@@ -110,6 +116,8 @@ func TestCurrentServiceParams(t *testing.T) {
t.Cleanup(func() {
logLevel = origLogLevel
daemonAddr = origDaemonAddr
jsonSocket = origJSONSocket
enableJSONSocket = origEnableJSONSocket
managementURL = origManagementURL
configPath = origConfigPath
logFiles = origLogFiles
@@ -120,6 +128,8 @@ func TestCurrentServiceParams(t *testing.T) {
logLevel = "trace"
daemonAddr = "tcp://127.0.0.1:9999"
jsonSocket = "tcp://127.0.0.1:8080"
enableJSONSocket = true
managementURL = "https://mgmt.example.com"
configPath = "/tmp/test-config.json"
logFiles = []string{"/tmp/test.log"}
@@ -131,6 +141,8 @@ func TestCurrentServiceParams(t *testing.T) {
assert.Equal(t, "trace", params.LogLevel)
assert.Equal(t, "tcp://127.0.0.1:9999", params.DaemonAddr)
assert.Equal(t, "tcp://127.0.0.1:8080", params.JSONSocket)
assert.True(t, params.EnableJSONSocket)
assert.Equal(t, "https://mgmt.example.com", params.ManagementURL)
assert.Equal(t, "/tmp/test-config.json", params.ConfigPath)
assert.Equal(t, []string{"/tmp/test.log"}, params.LogFiles)
@@ -142,6 +154,8 @@ func TestCurrentServiceParams(t *testing.T) {
func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
origLogLevel := logLevel
origDaemonAddr := daemonAddr
origJSONSocket := jsonSocket
origEnableJSONSocket := enableJSONSocket
origManagementURL := managementURL
origConfigPath := configPath
origLogFiles := logFiles
@@ -151,6 +165,8 @@ func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
t.Cleanup(func() {
logLevel = origLogLevel
daemonAddr = origDaemonAddr
jsonSocket = origJSONSocket
enableJSONSocket = origEnableJSONSocket
managementURL = origManagementURL
configPath = origConfigPath
logFiles = origLogFiles
@@ -162,6 +178,8 @@ func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
// Reset all flags to defaults.
logLevel = "info"
daemonAddr = "unix:///var/run/netbird.sock"
jsonSocket = defaultJSONSocket
enableJSONSocket = false
managementURL = ""
configPath = "/etc/netbird/config.json"
logFiles = []string{"/var/log/netbird/client.log"}
@@ -184,6 +202,8 @@ func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
saved := &serviceParams{
LogLevel: "debug",
DaemonAddr: "tcp://127.0.0.1:5555",
JSONSocket: "tcp://127.0.0.1:8080",
EnableJSONSocket: true,
ManagementURL: "https://saved.example.com",
ConfigPath: "/saved/config.json",
LogFiles: []string{"/saved/client.log"},
@@ -201,6 +221,8 @@ func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
// All other fields were not Changed, so they should use saved values.
assert.Equal(t, "tcp://127.0.0.1:5555", daemonAddr)
assert.Equal(t, "tcp://127.0.0.1:8080", jsonSocket)
assert.True(t, enableJSONSocket)
assert.Equal(t, "https://saved.example.com", managementURL)
assert.Equal(t, "/saved/config.json", configPath)
assert.Equal(t, []string{"/saved/client.log"}, logFiles)
@@ -212,14 +234,17 @@ func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
func TestApplyServiceParams_BooleanRevertToFalse(t *testing.T) {
origProfilesDisabled := profilesDisabled
origUpdateSettingsDisabled := updateSettingsDisabled
origEnableJSONSocket := enableJSONSocket
t.Cleanup(func() {
profilesDisabled = origProfilesDisabled
updateSettingsDisabled = origUpdateSettingsDisabled
enableJSONSocket = origEnableJSONSocket
})
// Simulate current state where booleans are true (e.g. set by previous install).
profilesDisabled = true
updateSettingsDisabled = true
enableJSONSocket = true
// Reset Changed state so flags appear unset.
serviceCmd.PersistentFlags().VisitAll(func(f *pflag.Flag) {
@@ -238,6 +263,7 @@ func TestApplyServiceParams_BooleanRevertToFalse(t *testing.T) {
assert.False(t, profilesDisabled, "saved false should override current true")
assert.False(t, updateSettingsDisabled, "saved false should override current true")
assert.False(t, enableJSONSocket, "saved false should override current true")
}
func TestApplyServiceParams_ClearManagementURL(t *testing.T) {
@@ -530,6 +556,7 @@ func fieldToGlobalVar(field string) string {
m := map[string]string{
"LogLevel": "logLevel",
"DaemonAddr": "daemonAddr",
"JSONSocket": "jsonSocket",
"ManagementURL": "managementURL",
"ConfigPath": "configPath",
"LogFiles": "logFiles",
@@ -537,6 +564,7 @@ func fieldToGlobalVar(field string) string {
"DisableUpdateSettings": "updateSettingsDisabled",
"EnableCapture": "captureEnabled",
"DisableNetworks": "networksDisabled",
"EnableJSONSocket": "enableJSONSocket",
"ServiceEnvVars": "serviceEnvVars",
}
if v, ok := m[field]; ok {

View File

@@ -0,0 +1,14 @@
//go:build !windows
package cmd
import (
"fmt"
"net"
)
// listenNamedPipe is Windows-only: no other platform serves the daemon on a
// named pipe.
func listenNamedPipe(string) (net.Listener, string, error) {
return nil, "", fmt.Errorf("named pipes are only supported on Windows")
}

View File

@@ -0,0 +1,41 @@
//go:build windows
package cmd
import (
"errors"
"fmt"
"net"
"github.com/Microsoft/go-winio"
log "github.com/sirupsen/logrus"
"github.com/netbirdio/netbird/client/internal/daemonaddr"
"github.com/netbirdio/netbird/client/internal/ipcauth"
)
// listenNamedPipe creates the daemon control pipe and reports the path it ended
// up on. The security descriptor lets any local caller connect, as a Unix socket
// at 0666 does, and the privileged operations are authorized separately from the
// caller's token.
//
// The protected name comes first so that an unprivileged process cannot take the
// name before the service does. Creating it requires being an administrator or
// LocalSystem, so a daemon an ordinary user runs themselves, as in netstack mode,
// falls back to the plain name; clients try both and check who serves them.
func listenNamedPipe(name string) (net.Listener, string, error) {
var errs []error
for _, path := range daemonaddr.PipePaths(name) {
listener, err := winio.ListenPipe(path, &winio.PipeConfig{
SecurityDescriptor: ipcauth.DefaultPipeSDDL(),
})
if err != nil {
log.Debugf("not serving the daemon on %s: %v", path, err)
errs = append(errs, fmt.Errorf("%s: %w", path, err))
continue
}
return listener, path, nil
}
return nil, "", errors.Join(errs...)
}

View File

@@ -0,0 +1,119 @@
//go:build !ios && !android
package cmd
import (
"errors"
"fmt"
"net"
"os"
"strings"
"syscall"
"time"
log "github.com/sirupsen/logrus"
)
type socketListener struct {
net.Listener
network string
address string
}
func listenOnAddress(addr string) (*socketListener, error) {
network, address, err := parseListenAddress(addr)
if err != nil {
return nil, err
}
if network == "npipe" {
listener, path, err := listenNamedPipe(address)
if err != nil {
return nil, err
}
return &socketListener{Listener: listener, network: network, address: path}, nil
}
if network == "unix" {
removeStaleUnixSocket(address)
}
listener, err := net.Listen(network, address)
if err != nil {
return nil, err
}
return &socketListener{Listener: listener, network: network, address: address}, nil
}
func parseListenAddress(addr string) (string, string, error) {
network, address, ok := strings.Cut(addr, "://")
if !ok || network == "" || address == "" {
return "", "", fmt.Errorf("address must be in [unix|tcp|npipe]://[path|host:port|name] format: %q", addr)
}
switch network {
case "unix", "tcp", "npipe":
return network, address, nil
default:
return "", "", fmt.Errorf("unsupported daemon address protocol: %v", network)
}
}
func removeStaleUnixSocket(path string) {
stat, err := os.Lstat(path)
if err != nil {
if !os.IsNotExist(err) {
log.Debugf("stat socket file: %v", err)
}
return
}
if stat.Mode()&os.ModeSocket == 0 {
return
}
if !isStaleUnixSocket(path) {
return
}
if err := os.Remove(path); err != nil {
log.Debugf("remove socket file: %v", err)
}
}
func isStaleUnixSocket(path string) bool {
conn, err := net.DialTimeout("unix", path, 100*time.Millisecond)
if err == nil {
if closeErr := conn.Close(); closeErr != nil {
log.Debugf("close unix socket probe: %v", closeErr)
}
return false
}
if os.IsNotExist(err) || os.IsPermission(err) || os.IsTimeout(err) {
log.Debugf("not removing unix socket %s after probe error: %v", path, err)
return false
}
return errors.Is(err, syscall.ECONNREFUSED)
}
func removeStaleUnixSocketForAddress(addr string) {
network, address, err := parseListenAddress(addr)
if err != nil || network != "unix" {
return
}
removeStaleUnixSocket(address)
}
func (l *socketListener) chmodUnixSocket(description string) error {
if l == nil || l.network != "unix" {
return nil
}
if err := os.Chmod(l.address, 0666); err != nil {
return fmt.Errorf("failed setting %s permissions for %s: %w", description, l.address, err)
}
return nil
}

View File

@@ -6,6 +6,7 @@ import (
"net"
"net/netip"
"strings"
"time"
"github.com/spf13/cobra"
"google.golang.org/grpc/status"
@@ -115,6 +116,11 @@ func statusFunc(cmd *cobra.Command, args []string) error {
// manager only knows the active profile ID, not its display name.
profName := getActiveProfileName(ctx)
var sessionExpiresAt time.Time
if ts := resp.GetSessionExpiresAt(); ts.IsValid() {
sessionExpiresAt = ts.AsTime().UTC()
}
var outputInformationHolder = nbstatus.ConvertToStatusOutputOverview(resp.GetFullStatus(), nbstatus.ConvertOptions{
Anonymize: anonymizeFlag,
DaemonVersion: resp.GetDaemonVersion(),
@@ -125,6 +131,7 @@ func statusFunc(cmd *cobra.Command, args []string) error {
IPsFilter: ipsFilterMap,
ConnectionTypeFilter: connectionTypeFilter,
ProfileName: profName,
SessionExpiresAt: sessionExpiresAt,
})
var statusOutputString string
switch {

View File

@@ -22,6 +22,8 @@ import (
"github.com/netbirdio/netbird/client/internal/peer"
"github.com/netbirdio/netbird/client/internal/profilemanager"
"github.com/netbirdio/netbird/client/proto"
nbnet "github.com/netbirdio/netbird/client/net"
"github.com/netbirdio/netbird/client/server"
"github.com/netbirdio/netbird/client/system"
"github.com/netbirdio/netbird/shared/management/domain"
"github.com/netbirdio/netbird/util"
@@ -229,6 +231,24 @@ func runInForegroundMode(ctx context.Context, cmd *cobra.Command, activeProf *pr
_, _ = profilemanager.UpdateOldManagementURL(ctx, config, configFilePath)
// Restore residual state left by a previous run that did not shut down
// cleanly, mirroring what the daemon does before connecting: it recovers
// DNS config (a stale resolv.conf takeover can make the management
// hostname unresolvable), firewall rules, ssh config and legacy routing.
// Route cleanup itself happens at engine start; nbnet.Init() below lets
// the management dial bypass a leftover fwmark rule until then.
// Foreground mode is particularly exposed in containers: a crashed
// container restarts inside the same (pod) network namespace, so stale
// state survives while the process does not.
if err := server.RestoreResidualState(ctx, profilemanager.NewServiceManager(configPath).GetStatePath()); err != nil {
log.Warnf("failed to restore residual state: %v", err)
}
// Enable advanced routing (as the daemon does on startup) so the
// management dial bypasses a leftover fwmark rule instead of being
// shunted into a stale routing table.
nbnet.Init()
err = foregroundLogin(ctx, cmd, config, providedSetupKey, activeProf.ID)
if err != nil {
return fmt.Errorf("foreground login failed: %v", err)
@@ -305,7 +325,7 @@ func runInDaemonMode(ctx context.Context, cmd *cobra.Command, pm *profilemanager
if st, ok := gstatus.FromError(err); ok && st.Code() == codes.Unavailable {
log.Warnf("setConfig method is not available in the daemon: %s", st.Message())
} else {
return fmt.Errorf("call service setConfig method: %v", err)
return daemonCallError("call service setConfig method", err)
}
}
@@ -359,7 +379,7 @@ func doDaemonUp(ctx context.Context, cmd *cobra.Command, client proto.DaemonServ
}
if loginErr != nil {
return fmt.Errorf("login failed: %v", loginErr)
return daemonCallError("login failed", loginErr)
}
if loginResp.NeedsSSOLogin {
@@ -372,7 +392,7 @@ func doDaemonUp(ctx context.Context, cmd *cobra.Command, client proto.DaemonServ
ProfileName: &profileID,
Username: &username,
}); err != nil {
return fmt.Errorf("call service up method: %v", err)
return daemonCallError("call service up method", err)
}
return nil

View File

@@ -6,6 +6,11 @@ import (
"runtime"
)
// UILogFile is the file name the desktop UI writes its log to. It is defined
// here so the UI (writer), the daemon's RegisterUILog validation, and the debug
// bundle collector all share one definition.
const UILogFile = "gui-client.log"
var StateDir string
func init() {

View File

@@ -470,7 +470,7 @@ func (c *Client) Status() (peer.FullStatus, error) {
if connect != nil {
engine := connect.Engine()
if engine != nil {
_ = engine.RunHealthProbes(false)
_ = engine.RunHealthProbes(context.Background(), false)
}
}

View File

@@ -0,0 +1,11 @@
//go:build android || (!linux && !windows)
package firewall
import "github.com/netbirdio/netbird/client/firewall/uspfilter"
// interfaceAllower returns no allower: these platforms have no host firewall to
// open for the interface.
func interfaceAllower(IFaceMapper, uint16) uspfilter.InterfaceAllower {
return nil
}

View File

@@ -0,0 +1,10 @@
//go:build windows
package firewall
import "github.com/netbirdio/netbird/client/firewall/uspfilter"
// interfaceAllower returns the Windows netsh-based interface allower.
func interfaceAllower(iface IFaceMapper, _ uint16) uspfilter.InterfaceAllower {
return uspfilter.NewWindowsInterfaceAllower(iface)
}

View File

@@ -6,8 +6,6 @@ import (
"fmt"
"runtime"
log "github.com/sirupsen/logrus"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
"github.com/netbirdio/netbird/client/firewall/uspfilter"
nftypes "github.com/netbirdio/netbird/client/internal/netflow/types"
@@ -21,13 +19,11 @@ func NewFirewall(iface IFaceMapper, _ *statemanager.Manager, flowLogger nftypes.
}
// use userspace packet filtering firewall
fm, err := uspfilter.Create(iface, disableServerRoutes, flowLogger, mtu)
if err != nil {
return nil, err
}
err = fm.AllowNetbird()
if err != nil {
log.Warnf("failed to allow netbird interface traffic: %v", err)
}
return fm, nil
return uspfilter.Create(uspfilter.Config{
IFace: iface,
DisableServerRoutes: disableServerRoutes,
FlowLogger: flowLogger,
MTU: mtu,
InterfaceAllower: interfaceAllower(iface, mtu),
})
}

View File

@@ -16,6 +16,7 @@ import (
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnftables "github.com/netbirdio/netbird/client/firewall/nftables"
"github.com/netbirdio/netbird/client/firewall/uspfilter"
"github.com/netbirdio/netbird/client/iface/netstack"
nftypes "github.com/netbirdio/netbird/client/internal/netflow/types"
"github.com/netbirdio/netbird/client/internal/statemanager"
)
@@ -29,47 +30,107 @@ const (
NFTABLES
)
// SKIP_NFTABLES_ENV is the environment variable to skip nftables check
const SKIP_NFTABLES_ENV = "NB_SKIP_NFTABLES_CHECK"
// SkipNftablesEnv is the environment variable to skip nftables check
const SkipNftablesEnv = "NB_SKIP_NFTABLES_CHECK"
// errNoFirewallManager indicates no kernel firewall backend is present,
// as opposed to a backend that exists but failed to create or initialize.
var errNoFirewallManager = errors.New("no firewall manager found")
// FWType is the type for the firewall type
type FWType int
func NewFirewall(iface IFaceMapper, stateManager *statemanager.Manager, flowLogger nftypes.FlowLogger, disableServerRoutes bool, mtu uint16) (firewall.Manager, error) {
// We run in userspace mode and force userspace firewall was requested. We don't attempt native firewall.
if iface.IsUserspaceBind() && forceUserspaceFirewall() {
log.Info("forcing userspace firewall")
return createUserspaceFirewall(iface, nil, disableServerRoutes, flowLogger, mtu)
// Userspace firewall without a native counterpart: routing is handled
// entirely in userspace. The interface is opened in the kernel's foreign
// filter chains via a table-less allower, except in netstack mode where no
// kernel interface exists.
if netstack.IsEnabled() || (iface.IsUserspaceBind() && forceUserspaceFirewall()) {
if netstack.IsEnabled() {
log.Info("netstack mode, using userspace firewall")
} else {
log.Info("forcing userspace firewall")
}
cfg := uspfilter.Config{
IFace: iface,
DisableServerRoutes: disableServerRoutes,
FlowLogger: flowLogger,
MTU: mtu,
InterfaceAllower: interfaceAllower(iface, mtu),
}
return uspfilter.Create(cfg)
}
// Use native firewall for either kernel or userspace, the interface appears identical to netfilter
fm, err := createNativeFirewall(iface, stateManager, disableServerRoutes, mtu)
// Kernel cannot fall back to anything else, need to return error
if !iface.IsUserspaceBind() {
return fm, err
}
// Fall back to the userspace packet filter if native is unavailable
if err != nil {
log.Warnf("failed to create native firewall: %v. Proceeding with userspace", err)
return createUserspaceFirewall(iface, nil, disableServerRoutes, flowLogger, mtu)
}
// Native firewall handles packet filtering, but the userspace WireGuard bind
// needs a device filter for DNS interception hooks. Install a minimal
// hooks-only filter that passes all traffic through to the kernel firewall.
if err := iface.SetFilter(&uspfilter.HooksFilter{}); err != nil {
log.Warnf("failed to set hooks filter, DNS via memory hooks will not work: %v", err)
fm, err := createNativeFirewall(iface, stateManager, mtu)
switch {
case err == nil && !iface.IsUserspaceBind():
// Nothing to do, fall through
case err == nil && iface.IsUserspaceBind():
// Native firewall handles packet filtering, but the userspace WireGuard bind
// needs a device filter for DNS interception hooks. Install a minimal
// hooks-only filter that passes all traffic through to the kernel firewall.
if err := iface.SetFilter(&uspfilter.HooksFilter{}); err != nil {
log.Warnf("failed to set hooks filter, DNS via memory hooks will not work: %v", err)
}
case err != nil && !iface.IsUserspaceBind():
// Kernel cannot fall back to anything else, need to return error
return nil, err
case err != nil && iface.IsUserspaceBind():
// Fall back to the userspace packet filter if native is unavailable
logNativeFirewallUnavailable(err)
return uspfilter.Create(uspfilter.Config{
IFace: iface,
DisableServerRoutes: disableServerRoutes,
FlowLogger: flowLogger,
MTU: mtu,
InterfaceAllower: interfaceAllower(iface, mtu),
})
}
return fm, nil
}
func createNativeFirewall(iface IFaceMapper, stateManager *statemanager.Manager, routes bool, mtu uint16) (firewall.Manager, error) {
// interfaceAllower selects how the userspace firewall opens the interface in
// foreign kernel chains: nftables when available (which also opens foreign nft
// tables), else iptables (the legacy fallback, filter INPUT only), else nil.
// firewalld trust is applied separately by the manager. Netstack has no kernel
// interface to open.
func interfaceAllower(iface IFaceMapper, mtu uint16) uspfilter.InterfaceAllower {
if netstack.IsEnabled() {
return nil
}
nftAllower, err := nbnftables.NewInterfaceAllower(iface, mtu)
if err == nil {
return nftAllower
}
log.Infof("no nftables interface allower: %v", err)
iptAllower, err := nbiptables.NewInterfaceAllower(iface)
if err == nil {
return iptAllower
}
log.Infof("no iptables interface allower: %v", err)
return nil
}
// logNativeFirewallUnavailable logs the fallback to userspace at info level
// when no kernel firewall backend exists, and at warn level otherwise.
func logNativeFirewallUnavailable(err error) {
if errors.Is(err, errNoFirewallManager) {
log.Infof("no native firewall backend available: %v. Proceeding with userspace", err)
} else {
log.Warnf("failed to create native firewall: %v. Proceeding with userspace", err)
}
}
func createNativeFirewall(iface IFaceMapper, stateManager *statemanager.Manager, mtu uint16) (firewall.Manager, error) {
fm, err := createFW(iface, mtu)
if err != nil {
return nil, fmt.Errorf("create firewall: %s", err)
return nil, fmt.Errorf("create firewall: %w", err)
}
if err = fm.Init(stateManager); err != nil {
@@ -88,29 +149,10 @@ func createFW(iface IFaceMapper, mtu uint16) (firewall.Manager, error) {
log.Info("creating an nftables firewall manager")
return nbnftables.Create(iface, mtu)
default:
log.Info("no firewall manager found, trying to use userspace packet filtering firewall")
return nil, errors.New("no firewall manager found")
return nil, errNoFirewallManager
}
}
func createUserspaceFirewall(iface IFaceMapper, fm firewall.Manager, disableServerRoutes bool, flowLogger nftypes.FlowLogger, mtu uint16) (firewall.Manager, error) {
var errUsp error
if fm != nil {
fm, errUsp = uspfilter.CreateWithNativeFirewall(iface, fm, disableServerRoutes, flowLogger, mtu)
} else {
fm, errUsp = uspfilter.Create(iface, disableServerRoutes, flowLogger, mtu)
}
if errUsp != nil {
return nil, fmt.Errorf("create userspace firewall: %s", errUsp)
}
if err := fm.AllowNetbird(); err != nil {
log.Errorf("failed to allow netbird interface traffic: %v", err)
}
return fm, nil
}
// check returns the firewall type based on common lib checks. It returns UNKNOWN if no firewall is found.
func check() FWType {
useIPTABLES := false
@@ -132,35 +174,38 @@ func check() FWType {
}
}
nf := nftables.Conn{}
if chains, err := nf.ListChains(); err == nil && os.Getenv(SKIP_NFTABLES_ENV) != "true" {
if !useIPTABLES {
return NFTABLES
}
// search for chains where table is filter
// if we find one, we assume that nftables manager can be used with iptables
for _, chain := range chains {
if chain.Table.Name == "filter" {
// Honor the skip env before probing nftables at all.
if os.Getenv(SkipNftablesEnv) != "true" {
nf := nftables.Conn{}
if chains, err := nf.ListChains(); err == nil {
if !useIPTABLES {
return NFTABLES
}
}
// check tables for the following constraints:
// 1. there is no chain in nftables for the filter table and there is at least one chain in iptables, we assume that nftables manager can not be used
// 2. there is no tables or more than one table, we assume that nftables manager can be used
// 3. there is only one table and its name is filter, we assume that nftables manager can not be used, since there was no chain in it
// 4. if we find an error we log and continue with iptables check
nbTablesList, err := nf.ListTables()
switch {
case err == nil && len(iptablesChains) > 0:
return IPTABLES
case err == nil && len(nbTablesList) != 1:
return NFTABLES
case err == nil && len(nbTablesList) == 1 && nbTablesList[0].Name == "filter":
return IPTABLES
case err != nil:
log.Errorf("failed to list nftables tables on fw manager discovery: %s", err)
// search for chains where table is filter
// if we find one, we assume that nftables manager can be used with iptables
for _, chain := range chains {
if chain.Table.Name == "filter" {
return NFTABLES
}
}
// check tables for the following constraints:
// 1. there is no chain in nftables for the filter table and there is at least one chain in iptables, we assume that nftables manager can not be used
// 2. there is no tables or more than one table, we assume that nftables manager can be used
// 3. there is only one table and its name is filter, we assume that nftables manager can not be used, since there was no chain in it
// 4. if we find an error we log and continue with iptables check
nbTablesList, err := nf.ListTables()
switch {
case err == nil && len(iptablesChains) > 0:
return IPTABLES
case err == nil && len(nbTablesList) != 1:
return NFTABLES
case err == nil && len(nbTablesList) == 1 && nbTablesList[0].Name == "filter":
return IPTABLES
case err != nil:
log.Errorf("failed to list nftables tables on fw manager discovery: %s", err)
}
}
}
@@ -176,15 +221,21 @@ func isIptablesClientAvailable(client *iptables.IPTables) bool {
return err == nil
}
// forceUserspaceFirewall reports whether the userspace firewall is forced.
// NB_FORCE_USERSPACE_ROUTER is an alias: forcing userspace routing implies the
// userspace firewall, since the two are no longer separable.
func forceUserspaceFirewall() bool {
val := os.Getenv(EnvForceUserspaceFirewall)
return envForceBool(EnvForceUserspaceFirewall) || envForceBool(uspfilter.EnvForceUserspaceRouter)
}
func envForceBool(name string) bool {
val := os.Getenv(name)
if val == "" {
return false
}
force, err := strconv.ParseBool(val)
if err != nil {
log.Warnf("failed to parse %s: %v", EnvForceUserspaceFirewall, err)
log.Warnf("failed to parse %s: %v", name, err)
return false
}
return force

View File

@@ -1,560 +0,0 @@
package iptables
import (
"errors"
"fmt"
"maps"
"net"
"slices"
"github.com/coreos/go-iptables/iptables"
"github.com/google/uuid"
ipset "github.com/lrh3321/ipset-go"
log "github.com/sirupsen/logrus"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
"github.com/netbirdio/netbird/client/internal/statemanager"
nbnet "github.com/netbirdio/netbird/client/net"
)
const (
tableName = "filter"
// rules chains contains the effective ACL rules
chainNameInputRules = "NETBIRD-ACL-INPUT"
// mangleFwdKey is the entries map key for mangle FORWARD guard rules that prevent
// external DNAT from bypassing ACL rules.
mangleFwdKey = "MANGLE-FORWARD"
)
type aclEntries map[string][][]string
type entry struct {
spec []string
position int
}
type aclManager struct {
iptablesClient *iptables.IPTables
wgIface iFaceMapper
entries aclEntries
optionalEntries map[string][]entry
ipsetStore *ipsetStore
v6 bool
stateManager *statemanager.Manager
}
func newAclManager(iptablesClient *iptables.IPTables, wgIface iFaceMapper) (*aclManager, error) {
return &aclManager{
iptablesClient: iptablesClient,
wgIface: wgIface,
entries: make(map[string][][]string),
optionalEntries: make(map[string][]entry),
ipsetStore: newIpsetStore(),
v6: iptablesClient.Proto() == iptables.ProtocolIPv6,
}, nil
}
func (m *aclManager) init(stateManager *statemanager.Manager) error {
m.stateManager = stateManager
m.seedInitialEntries()
m.seedInitialOptionalEntries()
if err := m.cleanChains(); err != nil {
return fmt.Errorf("clean chains: %w", err)
}
if err := m.createDefaultChains(); err != nil {
return fmt.Errorf("create default chains: %w", err)
}
m.updateState()
return nil
}
func (m *aclManager) AddPeerFiltering(
id []byte,
ip net.IP,
protocol firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
ipsetName string,
) ([]firewall.Rule, error) {
chain := chainNameInputRules
ipsetName = transformIPsetName(ipsetName, sPort, dPort, action)
if m.v6 && ipsetName != "" {
ipsetName += "-v6"
}
proto := protoForFamily(protocol, m.v6)
specs := filterRuleSpecs(ip, proto, sPort, dPort, action, ipsetName)
mangleSpecs := slices.Clone(specs)
mangleSpecs = append(mangleSpecs,
"-i", m.wgIface.Name(),
"-m", "addrtype", "--dst-type", "LOCAL",
"-j", "MARK", "--set-xmark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected),
)
specs = append(specs, "-j", actionToStr(action))
if ipsetName != "" {
if ipList, ipsetExists := m.ipsetStore.ipset(ipsetName); ipsetExists {
if err := m.addToIPSet(ipsetName, ip); err != nil {
return nil, fmt.Errorf("add IP to ipset: %w", err)
}
// if ruleset already exists it means we already have the firewall rule
// so we need to update IPs in the ruleset and return new fw.Rule object for ACL manager.
ipList.addIP(ip.String())
return []firewall.Rule{&Rule{
ruleID: uuid.New().String(),
ipsetName: ipsetName,
ip: ip.String(),
chain: chain,
specs: specs,
v6: m.v6,
}}, nil
}
if err := m.flushIPSet(ipsetName); err != nil {
if errors.Is(err, ipset.ErrSetNotExist) {
log.Debugf("flush ipset %s before use: %v", ipsetName, err)
} else {
log.Errorf("flush ipset %s before use: %v", ipsetName, err)
}
}
if err := m.createIPSet(ipsetName); err != nil {
return nil, fmt.Errorf("create ipset: %w", err)
}
if err := m.addToIPSet(ipsetName, ip); err != nil {
return nil, fmt.Errorf("add IP to ipset: %w", err)
}
ipList := newIpList(ip.String())
m.ipsetStore.addIpList(ipsetName, ipList)
}
ok, err := m.iptablesClient.Exists(tableFilter, chain, specs...)
if err != nil {
return nil, fmt.Errorf("failed to check rule: %w", err)
}
if ok {
return nil, fmt.Errorf("rule already exists")
}
// Insert DROP rules at the beginning, append ACCEPT rules at the end
if action == firewall.ActionDrop {
// Insert at the beginning of the chain (position 1)
err = m.iptablesClient.Insert(tableFilter, chain, 1, specs...)
} else {
err = m.iptablesClient.Append(tableFilter, chain, specs...)
}
if err != nil {
return nil, err
}
if err := m.iptablesClient.Append(tableMangle, chainRTPRE, mangleSpecs...); err != nil {
log.Errorf("failed to add mangle rule: %v", err)
mangleSpecs = nil
}
rule := &Rule{
ruleID: uuid.New().String(),
specs: specs,
mangleSpecs: mangleSpecs,
ipsetName: ipsetName,
ip: ip.String(),
chain: chain,
v6: m.v6,
}
m.updateState()
return []firewall.Rule{rule}, nil
}
// DeletePeerRule from the firewall by rule definition
func (m *aclManager) DeletePeerRule(rule firewall.Rule) error {
r, ok := rule.(*Rule)
if !ok {
return fmt.Errorf("invalid rule type")
}
shouldDestroyIpset := false
if ipsetList, ok := m.ipsetStore.ipset(r.ipsetName); ok {
// delete IP from ruleset IPs list and ipset
if _, ok := ipsetList.ips[r.ip]; ok {
ip := net.ParseIP(r.ip)
if ip == nil {
return fmt.Errorf("parse IP %s", r.ip)
}
if err := m.delFromIPSet(r.ipsetName, ip); err != nil {
return fmt.Errorf("delete ip from ipset: %w", err)
}
delete(ipsetList.ips, r.ip)
}
// if after delete, set still contains other IPs,
// no need to delete firewall rule and we should exit here
if len(ipsetList.ips) != 0 {
return nil
}
// we delete last IP from the set, that means we need to delete
// set itself and associated firewall rule too
m.ipsetStore.deleteIpset(r.ipsetName)
shouldDestroyIpset = true
}
if err := m.iptablesClient.Delete(tableName, r.chain, r.specs...); err != nil {
return fmt.Errorf("failed to delete rule: %s, %v: %w", r.chain, r.specs, err)
}
if r.mangleSpecs != nil {
if err := m.iptablesClient.Delete(tableMangle, chainRTPRE, r.mangleSpecs...); err != nil {
log.Errorf("failed to delete mangle rule: %v", err)
}
}
if shouldDestroyIpset {
if err := m.destroyIPSet(r.ipsetName); err != nil {
if errors.Is(err, ipset.ErrBusy) || errors.Is(err, ipset.ErrSetNotExist) {
log.Debugf("destroy empty ipset: %v", err)
} else {
log.Errorf("destroy empty ipset: %v", err)
}
}
}
m.updateState()
return nil
}
func (m *aclManager) Reset() error {
if err := m.cleanChains(); err != nil {
return fmt.Errorf("clean chains: %w", err)
}
m.updateState()
return nil
}
// todo write less destructive cleanup mechanism
func (m *aclManager) cleanChains() error {
ok, err := m.iptablesClient.ChainExists(tableName, chainNameInputRules)
if err != nil {
log.Debugf("failed to list chains: %s", err)
return err
}
if ok {
for _, rule := range m.entries["INPUT"] {
err := m.iptablesClient.DeleteIfExists(tableName, "INPUT", rule...)
if err != nil {
log.Errorf("failed to delete rule: %v, %s", rule, err)
}
}
for _, rule := range m.entries["FORWARD"] {
err := m.iptablesClient.DeleteIfExists(tableName, "FORWARD", rule...)
if err != nil {
log.Errorf("failed to delete rule: %v, %s", rule, err)
}
}
err = m.iptablesClient.ClearAndDeleteChain(tableName, chainNameInputRules)
if err != nil {
log.Debugf("failed to clear and delete %s chain: %s", chainNameInputRules, err)
return err
}
}
ok, err = m.iptablesClient.ChainExists("mangle", "PREROUTING")
if err != nil {
return fmt.Errorf("list chains: %w", err)
}
if ok {
for _, rule := range m.entries["PREROUTING"] {
err := m.iptablesClient.DeleteIfExists("mangle", "PREROUTING", rule...)
if err != nil {
log.Errorf("failed to delete rule: %v, %s", rule, err)
}
}
}
for _, rule := range m.entries[mangleFwdKey] {
if err := m.iptablesClient.DeleteIfExists(tableMangle, chainFORWARD, rule...); err != nil {
log.Errorf("failed to delete mangle FORWARD guard rule: %v, %s", rule, err)
}
}
for _, ipsetName := range m.ipsetStore.ipsetNames() {
if err := m.flushIPSet(ipsetName); err != nil {
if errors.Is(err, ipset.ErrSetNotExist) {
log.Debugf("flush ipset %q during reset: %v", ipsetName, err)
} else {
log.Errorf("flush ipset %q during reset: %v", ipsetName, err)
}
}
if err := m.destroyIPSet(ipsetName); err != nil {
if errors.Is(err, ipset.ErrBusy) || errors.Is(err, ipset.ErrSetNotExist) {
log.Debugf("destroy ipset %q during reset: %v", ipsetName, err)
} else {
log.Errorf("destroy ipset %q during reset: %v", ipsetName, err)
}
}
m.ipsetStore.deleteIpset(ipsetName)
}
return nil
}
func (m *aclManager) createDefaultChains() error {
// chain netbird-acl-input-rules
if err := m.iptablesClient.NewChain(tableName, chainNameInputRules); err != nil {
log.Debugf("failed to create '%s' chain: %s", chainNameInputRules, err)
return err
}
for chainName, rules := range m.entries {
// mangle FORWARD guard rules are handled separately below
if chainName == mangleFwdKey {
continue
}
for _, rule := range rules {
if err := m.iptablesClient.InsertUnique(tableName, chainName, 1, rule...); err != nil {
log.Debugf("failed to create input chain jump rule: %s", err)
return err
}
}
}
for chainName, entries := range m.optionalEntries {
for _, entry := range entries {
if err := m.iptablesClient.InsertUnique(tableName, chainName, entry.position, entry.spec...); err != nil {
log.Errorf("failed to insert optional entry %v: %v", entry.spec, err)
continue
}
m.entries[chainName] = append(m.entries[chainName], entry.spec)
}
}
clear(m.optionalEntries)
// Insert mangle FORWARD guard rules to prevent external DNAT bypass.
for _, rule := range m.entries[mangleFwdKey] {
if err := m.iptablesClient.AppendUnique(tableMangle, chainFORWARD, rule...); err != nil {
log.Errorf("failed to add mangle FORWARD guard rule: %v", err)
}
}
return nil
}
// seedInitialEntries adds default rules to the entries map, rules are inserted on pos 1, hence the order is reversed.
// We want to make sure our traffic is not dropped by existing rules.
// The existing FORWARD rules/policies decide outbound traffic towards our interface.
// In case the FORWARD policy is set to "drop", we add an established/related rule to allow return traffic for the inbound rule.
func (m *aclManager) seedInitialEntries() {
established := getConntrackEstablished()
m.appendToEntries("INPUT", []string{"-i", m.wgIface.Name(), "-j", "DROP"})
m.appendToEntries("INPUT", []string{"-i", m.wgIface.Name(), "-j", chainNameInputRules})
m.appendToEntries("INPUT", append([]string{"-i", m.wgIface.Name()}, established...))
// Inbound is handled by our ACLs, the rest is dropped.
// For outbound we respect the FORWARD policy. However, we need to allow established/related traffic for inbound rules.
m.appendToEntries("FORWARD", []string{"-i", m.wgIface.Name(), "-j", "DROP"})
m.appendToEntries("FORWARD", []string{"-o", m.wgIface.Name(), "-j", chainRTFWDOUT})
m.appendToEntries("FORWARD", []string{"-i", m.wgIface.Name(), "-j", chainRTFWDIN})
// Mangle FORWARD guard: when external DNAT redirects traffic from the wg interface, it
// traverses FORWARD instead of INPUT, bypassing ACL rules. ACCEPT rules in filter FORWARD
// can be inserted above ours. Mangle runs before filter, so these guard rules enforce the
// ACL mark check where it cannot be overridden.
m.appendToEntries(mangleFwdKey, []string{
"-i", m.wgIface.Name(),
"-m", "conntrack", "--ctstate", "RELATED,ESTABLISHED",
"-j", "ACCEPT",
})
m.appendToEntries(mangleFwdKey, []string{
"-i", m.wgIface.Name(),
"-m", "conntrack", "--ctstate", "DNAT",
"-m", "mark", "!", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected),
"-j", "DROP",
})
}
func (m *aclManager) seedInitialOptionalEntries() {
m.optionalEntries["FORWARD"] = []entry{
{
spec: []string{"-m", "mark", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected), "-j", "ACCEPT"},
position: 2,
},
}
}
func (m *aclManager) appendToEntries(chainName string, spec []string) {
m.entries[chainName] = append(m.entries[chainName], spec)
}
func (m *aclManager) updateState() {
if m.stateManager == nil {
return
}
var currentState *ShutdownState
if existing := m.stateManager.GetState(currentState); existing != nil {
if existingState, ok := existing.(*ShutdownState); ok {
currentState = existingState
}
}
if currentState == nil {
currentState = &ShutdownState{}
}
currentState.Lock()
defer currentState.Unlock()
// Clone the maps so the persisted state holds a private snapshot. The
// live maps keep being mutated by subsequent rule operations while the
// state manager marshals the state from its periodic-save goroutine.
// Sharing them by reference races the two and aborts the process with a
// concurrent map iteration and write.
if m.v6 {
currentState.ACLEntries6 = maps.Clone(m.entries)
currentState.ACLIPsetStore6 = m.ipsetStore.clone()
} else {
currentState.ACLEntries = maps.Clone(m.entries)
currentState.ACLIPsetStore = m.ipsetStore.clone()
}
if err := m.stateManager.UpdateState(currentState); err != nil {
log.Errorf("failed to update state: %v", err)
}
}
// filterRuleSpecs returns the specs of a filtering rule
// protoForFamily translates ICMP to ICMPv6 for ip6tables.
// ip6tables requires "ipv6-icmp" (or "icmpv6") instead of "icmp".
func protoForFamily(protocol firewall.Protocol, v6 bool) string {
if v6 && protocol == firewall.ProtocolICMP {
return "ipv6-icmp"
}
return string(protocol)
}
func filterRuleSpecs(ip net.IP, protocol string, sPort, dPort *firewall.Port, action firewall.Action, ipsetName string) (specs []string) {
// don't use IP matching if IP is 0.0.0.0
matchByIP := !ip.IsUnspecified()
if matchByIP {
if ipsetName != "" {
specs = append(specs, "-m", "set", "--match-set", ipsetName, "src")
} else {
specs = append(specs, "-s", ip.String())
}
}
if protocol != "all" {
specs = append(specs, "-p", protocol)
}
specs = append(specs, applyPort("--sport", sPort)...)
specs = append(specs, applyPort("--dport", dPort)...)
return specs
}
func actionToStr(action firewall.Action) string {
if action == firewall.ActionAccept {
return "ACCEPT"
}
return "DROP"
}
func transformIPsetName(ipsetName string, sPort, dPort *firewall.Port, action firewall.Action) string {
if ipsetName == "" {
return ""
}
actionSuffix := ""
if action == firewall.ActionDrop {
actionSuffix = "-drop"
}
switch {
case sPort != nil && dPort != nil:
return ipsetName + "-sport-dport" + actionSuffix
case sPort != nil:
return ipsetName + "-sport" + actionSuffix
case dPort != nil:
return ipsetName + "-dport" + actionSuffix
default:
return ipsetName + actionSuffix
}
}
func (m *aclManager) createIPSet(name string) error {
opts := ipset.CreateOptions{
Replace: true,
}
if m.v6 {
opts.Family = ipset.FamilyIPV6
}
if err := ipset.Create(name, ipset.TypeHashNet, opts); err != nil {
return fmt.Errorf("create ipset %s: %w", name, err)
}
log.Debugf("created ipset %s with type hash:net", name)
return nil
}
func (m *aclManager) addToIPSet(name string, ip net.IP) error {
cidr := uint8(32)
if ip.To4() == nil {
cidr = 128
}
entry := &ipset.Entry{
IP: ip,
CIDR: cidr,
Replace: true,
}
if err := ipset.Add(name, entry); err != nil {
return fmt.Errorf("add IP to ipset %s: %w", name, err)
}
return nil
}
func (m *aclManager) delFromIPSet(name string, ip net.IP) error {
cidr := uint8(32)
if ip.To4() == nil {
cidr = 128
}
entry := &ipset.Entry{
IP: ip,
CIDR: cidr,
}
if err := ipset.Del(name, entry); err != nil {
return fmt.Errorf("delete IP from ipset %s: %w", name, err)
}
return nil
}
func (m *aclManager) flushIPSet(name string) error {
return ipset.Flush(name)
}
func (m *aclManager) destroyIPSet(name string) error {
return ipset.Destroy(name)
}

View File

@@ -0,0 +1,346 @@
//go:build !android
package iptables
import (
"fmt"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnet "github.com/netbirdio/netbird/client/net"
)
func (r *family) createContainers() error {
for _, chainInfo := range []struct {
chain string
table string
}{
{chainRTFwdIn, tableFilter},
{chainRTFwdOut, tableFilter},
{chainRTPre, tableMangle},
{chainRTNAT, tableNat},
{chainRTRdr, tableNat},
{chainRTMSSClamp, tableMangle},
} {
// Fallback: clear chains that survived an unclean shutdown.
if ok, _ := r.iptablesClient.ChainExists(chainInfo.table, chainInfo.chain); ok {
if err := r.iptablesClient.ClearAndDeleteChain(chainInfo.table, chainInfo.chain); err != nil {
log.Warnf("clear stale chain %s in %s: %v", chainInfo.chain, chainInfo.table, err)
}
}
if err := r.iptablesClient.NewChain(chainInfo.table, chainInfo.chain); err != nil {
return fmt.Errorf("create chain %s in table %s: %w", chainInfo.chain, chainInfo.table, err)
}
}
if err := r.insertEstablishedRule(chainRTFwdIn); err != nil {
return fmt.Errorf("insert established rule: %w", err)
}
if err := r.insertEstablishedRule(chainRTFwdOut); err != nil {
return fmt.Errorf("insert established rule: %w", err)
}
if err := r.addPostroutingRules(); err != nil {
return fmt.Errorf("add static nat rules: %w", err)
}
if err := r.addJumpRules(); err != nil {
return fmt.Errorf("add jump rules: %w", err)
}
if err := r.addMSSClampingRules(); err != nil {
log.Errorf("failed to add MSS clamping rules: %s", err)
}
return nil
}
func (r *family) addJumpRules() error {
// Jump to nat chain
natRule := jumpRuleSpec(chainRTNAT)
if err := r.iptablesClient.Insert(tableNat, chainPostrouting, 1, natRule...); err != nil {
return fmt.Errorf("add nat postrouting jump rule: %w", err)
}
r.rules[jumpNATPost] = natRule
// Jump to mangle prerouting chain
preRule := jumpRuleSpec(chainRTPre)
if err := r.iptablesClient.Insert(tableMangle, chainPrerouting, 1, preRule...); err != nil {
return fmt.Errorf("add mangle prerouting jump rule: %w", err)
}
r.rules[jumpManglePre] = preRule
// Jump to nat prerouting chain
rdrRule := jumpRuleSpec(chainRTRdr)
if err := r.iptablesClient.Insert(tableNat, chainPrerouting, 1, rdrRule...); err != nil {
return fmt.Errorf("add nat prerouting jump rule: %w", err)
}
r.rules[jumpNATPre] = rdrRule
return nil
}
func (r *family) setupDataPlaneMark() error {
var merr *multierror.Error
preRule := []string{
"-i", r.wgIface.Name(),
"-m", "conntrack", "--ctstate", "NEW",
"-j", "CONNMARK", "--set-mark", fmt.Sprintf("%#x", nbnet.DataPlaneMarkIn),
}
if err := r.iptablesClient.AppendUnique(tableMangle, chainPrerouting, preRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add mangle prerouting rule: %w", err))
} else {
r.rules[markManglePre] = preRule
}
postRule := []string{
"-o", r.wgIface.Name(),
"-m", "conntrack", "--ctstate", "NEW",
"-j", "CONNMARK", "--set-mark", fmt.Sprintf("%#x", nbnet.DataPlaneMarkOut),
}
if err := r.iptablesClient.AppendUnique(tableMangle, chainPostrouting, postRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add mangle postrouting rule: %w", err))
} else {
r.rules[markManglePost] = postRule
}
return nberrors.FormatErrorOrNil(merr)
}
// seedInitialEntries adds default rules to the entries map. Rules are
// inserted at position 1, so the order here is reversed.
//
// Existing FORWARD policy decides outbound traffic towards our
// interface. If FORWARD policy is "drop", we add an
// established/related rule to allow return traffic for inbound rules.
func (r *family) seedInitialEntries() {
established := getConntrackEstablished()
r.appendToEntries(chainInput, []string{"-i", r.wgIface.Name(), "-j", "DROP"})
r.appendToEntries(chainInput, []string{"-i", r.wgIface.Name(), "-j", chainACLInput})
r.appendToEntries(chainInput, append([]string{"-i", r.wgIface.Name()}, established...))
r.appendToEntries(chainForward, []string{"-i", r.wgIface.Name(), "-j", "DROP"})
r.appendToEntries(chainForward, []string{"-o", r.wgIface.Name(), "-j", chainRTFwdOut})
r.appendToEntries(chainForward, []string{"-i", r.wgIface.Name(), "-j", chainRTFwdIn})
// Mangle FORWARD guard: when external DNAT redirects traffic from
// the wg interface, it traverses FORWARD instead of INPUT,
// bypassing ACL rules. ACCEPT rules in filter FORWARD can be
// inserted above ours. Mangle runs before filter, so these guard
// rules enforce the ACL mark check where it cannot be overridden.
r.appendToEntries(mangleForwardKey, []string{
"-i", r.wgIface.Name(),
"-m", "conntrack", "--ctstate", "RELATED,ESTABLISHED",
"-j", "ACCEPT",
})
r.appendToEntries(mangleForwardKey, []string{
"-i", r.wgIface.Name(),
"-m", "conntrack", "--ctstate", "DNAT",
"-m", "mark", "!", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected),
"-j", "DROP",
})
}
func (r *family) seedInitialOptionalEntries() {
r.optionalEntries[chainForward] = []entry{
{
spec: []string{"-m", "mark", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected), "-j", "ACCEPT"},
position: 2,
},
}
}
func (r *family) appendToEntries(chain chainKey, spec ruleSpec) {
r.entries[chain] = append(r.entries[chain], spec)
}
func (r *family) createDefaultChains() error {
if err := r.iptablesClient.NewChain(tableFilter, chainACLInput); err != nil {
return fmt.Errorf("create %s chain: %w", chainACLInput, err)
}
for chain, rules := range r.entries {
// mangle FORWARD guard rules are handled separately below
if chain == mangleForwardKey {
continue
}
for _, rule := range rules {
if err := r.iptablesClient.InsertUnique(tableFilter, string(chain), 1, rule...); err != nil {
return fmt.Errorf("insert jump rule into %s: %w", chain, err)
}
}
}
for chain, entries := range r.optionalEntries {
for _, entry := range entries {
if err := r.iptablesClient.InsertUnique(tableFilter, string(chain), entry.position, entry.spec...); err != nil {
log.Errorf("failed to insert optional entry %v: %v", entry.spec, err)
continue
}
r.entries[chain] = append(r.entries[chain], entry.spec)
}
}
clear(r.optionalEntries)
// Insert mangle FORWARD guard rules to prevent external DNAT bypass.
for _, rule := range r.entries[mangleForwardKey] {
if err := r.iptablesClient.AppendUnique(tableMangle, chainForward, rule...); err != nil {
log.Errorf("failed to add mangle FORWARD guard rule: %v", err)
}
}
return nil
}
func (r *family) cleanUpDefaultForwardRules() error {
var merr *multierror.Error
// cleanJumpRules removes the OUTPUT jump to NETBIRD-NAT-OUTPUT among
// the others, so the chain below deletes cleanly instead of failing
// with "device or resource busy".
if err := r.cleanJumpRules(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("clean jump rules: %w", err))
}
for _, chainInfo := range []struct {
chain string
table string
}{
{chainRTFwdIn, tableFilter},
{chainRTFwdOut, tableFilter},
{chainRTPre, tableMangle},
{chainRTNAT, tableNat},
{chainRTRdr, tableNat},
{chainNATOutput, tableNat},
{chainRTMSSClamp, tableMangle},
} {
ok, err := r.iptablesClient.ChainExists(chainInfo.table, chainInfo.chain)
if err != nil {
merr = multierror.Append(merr, fmt.Errorf("check chain %s in table %s: %w", chainInfo.chain, chainInfo.table, err))
continue
}
if ok {
if err := r.iptablesClient.ClearAndDeleteChain(chainInfo.table, chainInfo.chain); err != nil {
merr = multierror.Append(merr, fmt.Errorf("clear and delete chain %s in table %s: %w", chainInfo.chain, chainInfo.table, err))
}
}
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) cleanJumpRules() error {
// locations maps each jump rule to the built-in table and chain it
// was inserted into, plus the netbird chain it targets.
locations := map[firewall.RuleID]struct{ table, chain, target string }{
jumpNATPost: {tableNat, chainPostrouting, chainRTNAT},
jumpManglePre: {tableMangle, chainPrerouting, chainRTPre},
jumpNATPre: {tableNat, chainPrerouting, chainRTRdr},
jumpMSSClamp: {tableMangle, chainForward, chainRTMSSClamp},
jumpNATOutput: {tableNat, chainOutput, chainNATOutput},
}
var merr *multierror.Error
for ruleID, loc := range locations {
rule, exists := r.rules[ruleID]
if !exists {
// Untracked (e.g. fresh start after an unclean shutdown with no
// restored state): if the target chain survived, remove the stale
// jump to it so the chain can be deleted.
ok, err := r.iptablesClient.ChainExists(loc.table, loc.target)
if err != nil {
merr = multierror.Append(merr, fmt.Errorf("check chain %s in table %s: %w", loc.target, loc.table, err))
continue
}
if !ok {
continue
}
rule = jumpRuleSpec(loc.target)
}
if err := r.iptablesClient.DeleteIfExists(loc.table, loc.chain, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete rule from chain %s in table %s: %w", loc.chain, loc.table, err))
continue
}
delete(r.rules, ruleID)
}
return nberrors.FormatErrorOrNil(merr)
}
// jumpRuleSpec builds the iptables rule spec that jumps to target. Create
// and cleanup sites share it so the installed and deleted specs cannot drift.
func jumpRuleSpec(target string) []string {
return []string{"-j", target}
}
func (r *family) cleanAclChains() error {
var merr *multierror.Error
if err := r.cleanInputAclChain(); err != nil {
merr = multierror.Append(merr, err)
}
for _, rule := range r.entries[mangleForwardKey] {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainForward, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete mangle %s guard rule %v: %w", chainForward, rule, err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) cleanInputAclChain() error {
ok, err := r.iptablesClient.ChainExists(tableFilter, chainACLInput)
if err != nil {
return fmt.Errorf("check chain %s: %w", chainACLInput, err)
}
if !ok {
return nil
}
var merr *multierror.Error
for _, rule := range r.entries[chainInput] {
if err := r.iptablesClient.DeleteIfExists(tableFilter, chainInput, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete %s rule %v: %w", chainInput, rule, err))
}
}
for _, rule := range r.entries[chainForward] {
if err := r.iptablesClient.DeleteIfExists(tableFilter, chainForward, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete %s rule %v: %w", chainForward, rule, err))
}
}
if err := r.iptablesClient.ClearAndDeleteChain(tableFilter, chainACLInput); err != nil {
merr = multierror.Append(merr, fmt.Errorf("clear and delete %s chain: %w", chainACLInput, err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) cleanupDataPlaneMark() error {
var merr *multierror.Error
if preRule, exists := r.rules[markManglePre]; exists {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainPrerouting, preRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove mangle prerouting rule: %w", err))
} else {
delete(r.rules, markManglePre)
}
}
if postRule, exists := r.rules[markManglePost]; exists {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainPostrouting, postRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove mangle postrouting rule: %w", err))
} else {
delete(r.rules, markManglePost)
}
}
return nberrors.FormatErrorOrNil(merr)
}

View File

@@ -0,0 +1,285 @@
//go:build !android
package iptables
import (
"fmt"
"net/netip"
"strconv"
"strings"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
)
func (r *family) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error) {
ruleID := rule.ID()
if _, exists := r.rules[ruleID+dnatSuffix]; exists {
return rule, nil
}
toDestination := rule.TranslatedAddress.String()
switch {
case len(rule.TranslatedPort.Values) == 0:
// no translated port, use original port
case len(rule.TranslatedPort.Values) == 1:
toDestination += fmt.Sprintf(":%d", rule.TranslatedPort.Values[0])
case rule.TranslatedPort.IsRange && len(rule.TranslatedPort.Values) == 2:
// need the "/originalport" suffix to avoid dnat port randomization
toDestination += fmt.Sprintf(":%d-%d/%d", rule.TranslatedPort.Values[0], rule.TranslatedPort.Values[1], rule.DestinationPort.Values[0])
default:
return nil, fmt.Errorf("invalid translated port: %v", rule.TranslatedPort)
}
proto := strings.ToLower(string(rule.Protocol))
rules := make(map[firewall.RuleID]ruleInfo, 3)
// DNAT rule
dnatRule := []string{
"!", "-i", r.wgIface.Name(),
"-p", proto,
"-j", "DNAT",
"--to-destination", toDestination,
}
dnatRule = append(dnatRule, applyPort("--dport", &rule.DestinationPort)...)
rules[ruleID+dnatSuffix] = ruleInfo{
table: tableNat,
chain: chainRTRdr,
rule: dnatRule,
}
// SNAT rule
snatRule := []string{
"-o", r.wgIface.Name(),
"-p", proto,
"-d", rule.TranslatedAddress.String(),
"-j", "MASQUERADE",
}
snatRule = append(snatRule, applyPort("--dport", &rule.TranslatedPort)...)
rules[ruleID+snatSuffix] = ruleInfo{
table: tableNat,
chain: chainRTNAT,
rule: snatRule,
}
// Forward filtering rule, if fwd policy is DROP
forwardRule := []string{
"-o", r.wgIface.Name(),
"-p", proto,
"-d", rule.TranslatedAddress.String(),
"-j", "ACCEPT",
}
forwardRule = append(forwardRule, applyPort("--dport", &rule.TranslatedPort)...)
rules[ruleID+fwdSuffix] = ruleInfo{
table: tableFilter,
chain: chainRTFwdOut,
rule: forwardRule,
}
// Request forwarding once the rule is about to be installed, releasing
// it if installation fails so the refcount tracks the real rules.
if err := r.ipFwdState.RequestForwarding(); err != nil {
return nil, err
}
for key, ruleInfo := range rules {
if err := r.iptablesClient.Append(ruleInfo.table, ruleInfo.chain, ruleInfo.rule...); err != nil {
if rollbackErr := r.rollbackRules(rules); rollbackErr != nil {
log.Errorf("rollback failed: %v", rollbackErr)
}
r.releaseForwarding()
return nil, fmt.Errorf("add rule %s: %w", key, err)
}
r.rules[key] = ruleInfo.rule
}
r.updateState()
return rule, nil
}
func (r *family) rollbackRules(rules map[firewall.RuleID]ruleInfo) error {
var merr *multierror.Error
for key, ruleInfo := range rules {
if err := r.iptablesClient.DeleteIfExists(ruleInfo.table, ruleInfo.chain, ruleInfo.rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("rollback rule %s: %w", key, err))
// On rollback error, add to rules map for next cleanup
r.rules[key] = ruleInfo.rule
}
}
if merr != nil {
r.updateState()
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) DeleteDNATRule(rule firewall.Rule) error {
ruleID := rule.ID()
var merr *multierror.Error
var found bool
if dnatRule, exists := r.rules[ruleID+dnatSuffix]; exists {
found = true
if err := r.iptablesClient.Delete(tableNat, chainRTRdr, dnatRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete DNAT rule: %w", err))
}
delete(r.rules, ruleID+dnatSuffix)
}
if snatRule, exists := r.rules[ruleID+snatSuffix]; exists {
found = true
if err := r.iptablesClient.Delete(tableNat, chainRTNAT, snatRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete SNAT rule: %w", err))
}
delete(r.rules, ruleID+snatSuffix)
}
if fwdRule, exists := r.rules[ruleID+fwdSuffix]; exists {
found = true
if err := r.iptablesClient.Delete(tableFilter, chainRTFwdOut, fwdRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete forward rule: %w", err))
}
delete(r.rules, ruleID+fwdSuffix)
}
r.updateState()
// Release once, only if the rule was present and removed.
if merr == nil && found {
r.releaseForwarding()
}
return nberrors.FormatErrorOrNil(merr)
}
// releaseForwarding drops one IP forwarding reference, logging any error.
func (r *family) releaseForwarding() {
if err := r.ipFwdState.ReleaseForwarding(); err != nil {
log.Errorf("release IP forwarding: %v", err)
}
}
func (r *family) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("inbound-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if _, exists := r.rules[ruleID]; exists {
return nil
}
dnatRule := []string{
"-i", r.wgIface.Name(),
"-p", strings.ToLower(protoForFamily(protocol, r.v6)),
"--dport", strconv.Itoa(int(originalPort)),
"-d", localAddr.String(),
"-m", "addrtype", "--dst-type", "LOCAL",
"-j", "DNAT",
"--to-destination", ":" + strconv.Itoa(int(translatedPort)),
}
info := ruleInfo{
table: tableNat,
chain: chainRTRdr,
rule: dnatRule,
}
if err := r.iptablesClient.Append(info.table, info.chain, info.rule...); err != nil {
return fmt.Errorf("add inbound DNAT rule: %w", err)
}
r.rules[ruleID] = info.rule
r.updateState()
return nil
}
// RemoveInboundDNAT removes an inbound DNAT rule.
func (r *family) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("inbound-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if dnatRule, exists := r.rules[ruleID]; exists {
if err := r.iptablesClient.Delete(tableNat, chainRTRdr, dnatRule...); err != nil {
return fmt.Errorf("delete inbound DNAT rule: %w", err)
}
delete(r.rules, ruleID)
}
r.updateState()
return nil
}
// ensureNATOutputChain lazily creates the OUTPUT NAT chain and jump rule on first use.
func (r *family) ensureNATOutputChain() error {
if _, exists := r.rules[jumpNATOutput]; exists {
return nil
}
chainExists, err := r.iptablesClient.ChainExists(tableNat, chainNATOutput)
if err != nil {
return fmt.Errorf("check chain %s: %w", chainNATOutput, err)
}
if !chainExists {
if err := r.iptablesClient.NewChain(tableNat, chainNATOutput); err != nil {
return fmt.Errorf("create chain %s: %w", chainNATOutput, err)
}
}
jumpRule := jumpRuleSpec(chainNATOutput)
if err := r.iptablesClient.Insert(tableNat, chainOutput, 1, jumpRule...); err != nil {
if !chainExists {
if delErr := r.iptablesClient.ClearAndDeleteChain(tableNat, chainNATOutput); delErr != nil {
log.Warnf("failed to rollback chain %s: %v", chainNATOutput, delErr)
}
}
return fmt.Errorf("add OUTPUT jump rule: %w", err)
}
r.rules[jumpNATOutput] = jumpRule
r.updateState()
return nil
}
// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
func (r *family) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if _, exists := r.rules[ruleID]; exists {
return nil
}
if err := r.ensureNATOutputChain(); err != nil {
return err
}
dnatRule := []string{
"-p", strings.ToLower(protoForFamily(protocol, localAddr.Is6())),
"--dport", strconv.Itoa(int(originalPort)),
"-d", localAddr.String(),
"-j", "DNAT",
"--to-destination", ":" + strconv.Itoa(int(translatedPort)),
}
if err := r.iptablesClient.Append(tableNat, chainNATOutput, dnatRule...); err != nil {
return fmt.Errorf("add output DNAT rule: %w", err)
}
r.rules[ruleID] = dnatRule
r.updateState()
return nil
}
// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
func (r *family) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if dnatRule, exists := r.rules[ruleID]; exists {
if err := r.iptablesClient.Delete(tableNat, chainNATOutput, dnatRule...); err != nil {
return fmt.Errorf("delete output DNAT rule: %w", err)
}
delete(r.rules, ruleID)
}
r.updateState()
return nil
}

View File

@@ -0,0 +1,252 @@
//go:build !android
package iptables
import (
"fmt"
"maps"
"net/netip"
"github.com/coreos/go-iptables/iptables"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbid "github.com/netbirdio/netbird/client/internal/acl/id"
"github.com/netbirdio/netbird/client/internal/routemanager/ipfwdstate"
"github.com/netbirdio/netbird/client/internal/routemanager/refcounter"
"github.com/netbirdio/netbird/client/internal/statemanager"
)
// constants needed to manage and create iptable rules
const (
tableFilter = "filter"
tableNat = "nat"
tableMangle = "mangle"
// chainACLInput is the peer ACL chain that holds installed
// peer-filtering rules.
chainACLInput = "NETBIRD-ACL-INPUT"
// mangleForwardKey is the entries map key for mangle FORWARD guard
// rules that prevent external DNAT from bypassing ACL rules.
mangleForwardKey chainKey = "MANGLE-FORWARD"
chainInput = "INPUT"
chainPostrouting = "POSTROUTING"
chainPrerouting = "PREROUTING"
chainForward = "FORWARD"
chainRTNAT = "NETBIRD-RT-NAT"
chainRTFwdIn = "NETBIRD-RT-FWD-IN"
chainRTFwdOut = "NETBIRD-RT-FWD-OUT"
chainRTPre = "NETBIRD-RT-PRE"
chainRTRdr = "NETBIRD-RT-RDR"
chainNATOutput = "NETBIRD-NAT-OUTPUT"
chainRTMSSClamp = "NETBIRD-RT-MSSCLAMP"
jumpManglePre = "jump-mangle-pre"
jumpNATPre = "jump-nat-pre"
jumpNATPost = "jump-nat-post"
jumpNATOutput = "jump-nat-output"
jumpMSSClamp = "jump-mss-clamp"
markManglePre = "mark-mangle-pre"
markManglePost = "mark-mangle-post"
matchSet = "--match-set"
dnatSuffix firewall.RuleID = "_dnat"
snatSuffix firewall.RuleID = "_snat"
fwdSuffix firewall.RuleID = "_fwd"
// ipv4TCPHeaderSize is the minimum IPv4 (20) + TCP (20) header size for MSS calculation.
ipv4TCPHeaderSize = 40
// ipv6TCPHeaderSize is the minimum IPv6 (40) + TCP (20) header size for MSS calculation.
ipv6TCPHeaderSize = 60
)
type ruleInfo struct {
chain string
table string
rule []string
}
type routeRules map[firewall.RuleID][]string
// ruleSpec is a single iptables rule expressed as its argument list
// (e.g. {"-i", "wg0", "-j", "DROP"}).
type ruleSpec []string
// chainKey identifies the chain a seeded entry belongs to. It holds
// built-in chain names ("INPUT", "FORWARD", "PREROUTING") plus the
// synthetic mangleForwardKey bucket for the mangle FORWARD guard rules.
type chainKey string
// aclEntries maps a chain to the rules seeded into it to jump into or
// guard the netbird ACL chains.
type aclEntries map[chainKey][]ruleSpec
type entry struct {
spec ruleSpec
position int
}
// ipsetCounter is the shared hash:net refcounter used by peer and
// route ACLs alike. The ipset library does not support comments, so
// the key is just the set name (string).
type ipsetCounter = refcounter.Counter[string, []netip.Prefix, struct{}]
// family holds the per-address-family iptables state. One instance
// handles route ACLs, peer ACLs, NAT, DNAT, and MSS clamping for a
// single family; the top-level Manager owns one for v4 and another
// for v6.
type family struct {
iptablesClient *iptables.IPTables
wgIface iFaceMapper
v6 bool
// Peer ACL chain bookkeeping.
entries aclEntries
optionalEntries map[chainKey][]entry
// filters holds peer + route filter rules keyed by content hash.
// AddFilterRule writes here; DeleteFilterRule looks up by id.
filters map[nbid.RuleID]*Rule
ipsetCounter *ipsetCounter
// rules holds NAT, jump, and MSS-clamping rules (auxiliary
// plumbing that isn't a filter rule).
rules routeRules
// Routing / NAT.
legacyManagement bool
mtu uint16
ipFwdState *ipfwdstate.IPForwardingState
stateManager *statemanager.Manager
}
func newFamily(iptablesClient *iptables.IPTables, wgIface iFaceMapper, mtu uint16) (*family, error) {
r := &family{
iptablesClient: iptablesClient,
wgIface: wgIface,
v6: iptablesClient.Proto() == iptables.ProtocolIPv6,
entries: make(aclEntries),
optionalEntries: make(map[chainKey][]entry),
filters: make(map[nbid.RuleID]*Rule),
rules: make(routeRules),
mtu: mtu,
ipFwdState: ipfwdstate.NewIPForwardingState(),
}
r.ipsetCounter = refcounter.New(
func(name string, sources []netip.Prefix) (struct{}, error) {
return struct{}{}, r.createIpSet(name, sources)
},
func(name string, _ struct{}) error {
return r.deleteIpSet(name)
},
)
return r, nil
}
// init wires the family to the state manager and installs both the
// route ACL containers and the peer ACL chain skeleton.
func (r *family) init(stateManager *statemanager.Manager) error {
r.stateManager = stateManager
if err := r.cleanUpDefaultForwardRules(); err != nil {
log.Errorf("failed to clean up rules from FORWARD chain: %s", err)
}
if err := r.createContainers(); err != nil {
return fmt.Errorf("create containers: %w", err)
}
if err := r.setupDataPlaneMark(); err != nil {
log.Errorf("failed to set up data plane mark: %v", err)
}
r.seedInitialEntries()
r.seedInitialOptionalEntries()
if err := r.cleanAclChains(); err != nil {
return fmt.Errorf("clean acl chains: %w", err)
}
if err := r.createDefaultChains(); err != nil {
return fmt.Errorf("create default chains: %w", err)
}
r.updateState()
return nil
}
// Reset tears down all firewall state owned by this family. ACL
// chain cleanup runs before route-chain cleanup because the route
// chains are still referenced by FORWARD jumps installed during
// seedInitialEntries; deleting them first would trip EBUSY.
func (r *family) Reset() error {
var merr *multierror.Error
if err := r.cleanAclChains(); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.cleanUpDefaultForwardRules(); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.ipsetCounter.Flush(); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.cleanupDataPlaneMark(); err != nil {
merr = multierror.Append(merr, err)
}
clear(r.rules)
clear(r.filters)
r.updateState()
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) updateState() {
if r.stateManager == nil {
return
}
var currentState *ShutdownState
if existing := r.stateManager.GetState(currentState); existing != nil {
if existingState, ok := existing.(*ShutdownState); ok {
currentState = existingState
}
}
if currentState == nil {
currentState = &ShutdownState{}
}
currentState.Lock()
defer currentState.Unlock()
// Clone the rule maps so the persisted state holds a private snapshot.
// The live maps keep being mutated by subsequent rule operations while
// the state manager marshals the state from its periodic-save goroutine.
// Sharing the maps by reference races the two and aborts the process with
// a concurrent map iteration and write. The ipset counter guards itself
// during marshaling, so it can be shared directly.
if r.v6 {
currentState.RouteRules6 = maps.Clone(r.rules)
currentState.RouteIPsetCounter6 = r.ipsetCounter
currentState.ACLEntries6 = maps.Clone(r.entries)
} else {
currentState.RouteRules = maps.Clone(r.rules)
currentState.RouteIPsetCounter = r.ipsetCounter
currentState.ACLEntries = maps.Clone(r.entries)
}
if err := r.stateManager.UpdateState(currentState); err != nil {
log.Errorf("failed to update state: %v", err)
}
}

View File

@@ -0,0 +1,346 @@
//go:build !android
package iptables
import (
"fmt"
"net/netip"
"slices"
"strconv"
"strings"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbid "github.com/netbirdio/netbird/client/internal/acl/id"
nbnet "github.com/netbirdio/netbird/client/net"
)
// AddFilterRule installs a packet-filtering rule. With destination
// empty, the rule goes to the peer ACL input chain plus a paired
// mangle PREROUTING rule for the redirect mark. With destination set
// (prefix or named set), it goes to the route ACL forward chain.
// Multi-source rules collapse to one iptables rule via the shared
// hash:net ipset.
func (r *family) AddFilterRule(
id []byte,
sources []netip.Prefix,
destination firewall.Network,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
) (firewall.Rule, error) {
ruleID := nbid.GenerateRuleID(sources, destination, proto, sPort, dPort, action)
if existing, ok := r.filters[ruleID]; ok {
return existing, nil
}
srcMatch, err := r.applySourceMatch(sourceNetwork(sources), sources)
if err != nil {
return nil, fmt.Errorf("apply source match: %w", err)
}
rule, err := r.installFilterRule(ruleID, srcMatch, destination, proto, sPort, dPort, action)
if err != nil {
r.dropSourceMatch(srcMatch)
return nil, err
}
r.filters[ruleID] = rule
r.updateState()
return rule, nil
}
func (r *family) hasRule(id nbid.RuleID) bool {
_, ok := r.filters[id]
return ok
}
// hasDNATRule reports whether this family owns the DNAT rule set for
// the given user id. DNAT rules live in r.rules under the well-known
// "<id>_dnat" key; the lookup here is used by Manager.DeleteDNATRule
// to pick the right family.
func (r *family) hasDNATRule(id firewall.RuleID) bool {
_, ok := r.rules[id+dnatSuffix]
return ok
}
// DeleteFilterRule removes a previously installed filter rule. The
// rule's stored chain/table identify where to delete from; source set
// references are recovered from the spec via findSets and dropped
// from the shared ipset counter.
func (r *family) DeleteFilterRule(rule firewall.Rule) error {
ruleID := rule.ID()
pr, ok := r.filters[ruleID]
if !ok {
log.Debugf("filter rule %s not found", ruleID)
return nil
}
// DeleteIfExists keeps both deletes idempotent so a retry after a
// partial failure does not error on the half that was already removed.
var merr *multierror.Error
if err := r.iptablesClient.DeleteIfExists(tableFilter, pr.chain, pr.specs...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete rule from %s: %w", pr.chain, err))
}
if pr.mangleSpecs != nil {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainRTPre, pr.mangleSpecs...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete mangle rule: %w", err))
}
}
if merr != nil {
// Leave the rule tracked so the caller retries the remaining half.
return nberrors.FormatErrorOrNil(merr)
}
// The rule is gone from iptables, so untrack it regardless of how the
// refcount decrement goes, but surface decrement failures so callers
// see the ipset desync.
delete(r.filters, ruleID)
r.updateState()
if err := r.decrementSetCounter(pr.specs); err != nil {
return fmt.Errorf("drop source set references: %w", err)
}
return nil
}
// findSets scans an iptables rule spec for "-m set --match-set <name>
// <dir>" fragments and returns the named sets in occurrence order.
// Used at delete time to drop ipsetCounter references.
func findSets(rule []string) []string {
var sets []string
for i, arg := range rule {
if arg == "-m" && i+3 < len(rule) && rule[i+1] == "set" && rule[i+2] == matchSet {
sets = append(sets, rule[i+3])
}
}
return sets
}
// sourceNetwork classifies a source-prefix list into the firewall.Network
// shape the rest of the spec-builder consumes: empty for match-any, a
// single prefix inline, or an ipset for multiple sources.
func sourceNetwork(sources []netip.Prefix) firewall.Network {
switch {
case len(sources) == 0:
return firewall.Network{}
case len(sources) == 1 && sources[0].Bits() == 0:
return firewall.Network{}
case len(sources) == 1:
return firewall.Network{Prefix: sources[0]}
default:
return firewall.Network{Set: firewall.NewPrefixSet(sources)}
}
}
// applySourceMatch returns the iptables match fragment for the rule's
// source. For a Set it increments the shared ipset's refcount; for a
// Prefix it emits a direct -s match; for the wildcard it returns nil.
func (r *family) applySourceMatch(network firewall.Network, prefixes []netip.Prefix) ([]string, error) {
switch {
case network.IsSet():
if r.ipsetCounter == nil {
return nil, fmt.Errorf("multi-source peer rule requires shared ipset counter")
}
name := r.ipsetName(network.Set.HashedName())
if _, err := r.ipsetCounter.Increment(name, prefixes); err != nil {
return nil, fmt.Errorf("ipset increment %s: %w", name, err)
}
return []string{"-m", "set", matchSet, name, "src"}, nil
case network.IsPrefix():
return []string{"-s", network.Prefix.String()}, nil
default:
return nil, nil
}
}
// dropSourceMatch undoes whatever applySourceMatch reserved when
// installing a rule fails. Safe to call when the spec is empty or holds
// only inline matchers. Decrement errors are logged but not returned:
// the install error is what the caller needs to see.
func (r *family) dropSourceMatch(srcMatch []string) {
if r.ipsetCounter == nil {
return
}
for _, name := range findSets(srcMatch) {
if _, err := r.ipsetCounter.Decrement(name); err != nil {
log.Errorf("rollback ipset decrement %s: %v", name, err)
}
}
}
// decrementSetCounter drops ipset references owned by a raw rule spec
// stored in r.rules (NAT / legacy route entries). It returns an error
// aggregate so the caller surfaces decrement failures.
func (r *family) decrementSetCounter(rule []string) error {
if r.ipsetCounter == nil {
return nil
}
var merr *multierror.Error
for _, name := range findSets(rule) {
if _, err := r.ipsetCounter.Decrement(name); err != nil {
merr = multierror.Append(merr, fmt.Errorf("decrement counter: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
// installFilterRule assembles and writes one iptables filter-chain
// rule. With destination empty the rule lands in the peer ACL input
// chain and a paired mangle PREROUTING rule is added for the redirect
// mark. With destination set the rule lands in the route ACL forward
// chain and there is no mangle pairing.
func (r *family) installFilterRule(
ruleID nbid.RuleID,
srcMatch []string,
destination firewall.Network,
protocol firewall.Protocol,
sPort, dPort *firewall.Port,
action firewall.Action,
) (*Rule, error) {
isRoute := !destination.IsZero()
proto := protoForFamily(protocol, r.v6)
specs := slices.Clone(srcMatch)
var destExp []string
if isRoute {
var err error
destExp, err = r.applyNetwork("-d", destination, nil)
if err != nil {
return nil, fmt.Errorf("apply network -d: %w", err)
}
specs = append(specs, destExp...)
}
specs = append(specs, filterMatchSpecs(proto, sPort, dPort)...)
var mangleSpecs []string
if !isRoute {
mangleSpecs = slices.Clone(specs)
mangleSpecs = append(mangleSpecs,
"-i", r.wgIface.Name(),
"-m", "addrtype", "--dst-type", "LOCAL",
"-j", "MARK", "--set-xmark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkRedirected),
)
}
specs = append(specs, "-j", actionToStr(action))
chain := chainACLInput
if isRoute {
chain = chainRTFwdIn
}
// Peer ACL drops are inserted at position 1 so they precede the
// chain's catch-all; route ACL drops are inserted at position 2
// to sit immediately after the established/related accept rule.
var err error
if action == firewall.ActionDrop {
pos := 1
if isRoute {
pos = 2
}
err = r.iptablesClient.Insert(tableFilter, chain, pos, specs...)
} else {
err = r.iptablesClient.Append(tableFilter, chain, specs...)
}
if err != nil {
r.dropSourceMatch(destExp)
return nil, fmt.Errorf("install filter rule on %s: %w", chain, err)
}
// The mangle redirect-mark rule is best effort: the filter rule itself
// is what enforces the ACL, so a mangle failure must not undo it. Drop
// the spec so teardown does not try to remove a rule that was not added.
if mangleSpecs != nil {
if err := r.iptablesClient.Append(tableMangle, chainRTPre, mangleSpecs...); err != nil {
log.Errorf("add mangle rule: %v", err)
mangleSpecs = nil
}
}
return &Rule{
id: ruleID,
specs: specs,
mangleSpecs: mangleSpecs,
chain: chain,
v6: r.v6,
}, nil
}
// applyNetwork resolves a firewall.Network into the iptables match
// fragment for the given direction flag (-s or -d). Set networks
// increment the shared ipset refcount; prefixes emit a direct match;
// an empty network returns no spec ("match any").
func (r *family) applyNetwork(flag string, network firewall.Network, prefixes []netip.Prefix) ([]string, error) {
direction := "src"
if flag == "-d" {
direction = "dst"
}
if network.IsSet() {
name := r.ipsetName(network.Set.HashedName())
if _, err := r.ipsetCounter.Increment(name, prefixes); err != nil {
return nil, fmt.Errorf("create or get ipset: %w", err)
}
return []string{"-m", "set", matchSet, name, direction}, nil
}
if network.IsPrefix() {
return []string{flag, network.Prefix.String()}, nil
}
// nolint:nilnil
return nil, nil
}
// protoForFamily translates ICMP to ICMPv6 for ip6tables.
// ip6tables requires "ipv6-icmp" (or "icmpv6") instead of "icmp".
func protoForFamily(protocol firewall.Protocol, v6 bool) string {
if v6 && protocol == firewall.ProtocolICMP {
return "ipv6-icmp"
}
return string(protocol)
}
// filterMatchSpecs returns the proto/port match fragment for a
// filtering rule. The source match (-s or -m set) is built by the
// caller and prepended.
func filterMatchSpecs(protocol string, sPort, dPort *firewall.Port) (specs []string) {
if protocol != "all" {
specs = append(specs, "-p", protocol)
}
specs = append(specs, applyPort("--sport", sPort)...)
specs = append(specs, applyPort("--dport", dPort)...)
return specs
}
func actionToStr(action firewall.Action) string {
if action == firewall.ActionAccept {
return "ACCEPT"
}
return "DROP"
}
func applyPort(flag string, port *firewall.Port) []string {
if port == nil {
return nil
}
if port.IsRange && len(port.Values) == 2 {
return []string{flag, fmt.Sprintf("%d:%d", port.Values[0], port.Values[1])}
}
if len(port.Values) > 1 {
portList := make([]string, len(port.Values))
for i, p := range port.Values {
portList[i] = strconv.Itoa(int(p))
}
return []string{"-m", "multiport", flag, strings.Join(portList, ",")}
}
return []string{flag, strconv.Itoa(int(port.Values[0]))}
}

View File

@@ -0,0 +1,93 @@
package iptables
import (
"fmt"
"github.com/coreos/go-iptables/iptables"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
)
// InterfaceAllower opens the NetBird interface on the iptables filter INPUT
// chain so the host firewall doesn't drop traffic the userspace firewall
// handles. It is the fallback used when nftables is unavailable (an
// iptables-legacy host).
//
// It opens INPUT only: the userspace router never forwards in the kernel.
// firewalld trust is handled by the uspfilter manager, not here.
type InterfaceAllower struct {
ifaceName string
ipt4 *iptables.IPTables
// ipt6 is nil when the interface has no IPv6 overlay address.
ipt6 *iptables.IPTables
}
// NewInterfaceAllower builds an iptables allower for the interface. It returns
// an error when iptables is unavailable, so the caller can fall back to
// firewalld trust.
func NewInterfaceAllower(wgIface iFaceMapper) (*InterfaceAllower, error) {
ipt4, err := iptables.NewWithProtocol(iptables.ProtocolIPv4)
if err != nil {
return nil, fmt.Errorf("iptables not available: %w", err)
}
if _, err := ipt4.ListChains(tableFilter); err != nil {
return nil, fmt.Errorf("iptables filter table not available: %w", err)
}
a := &InterfaceAllower{ifaceName: wgIface.Name(), ipt4: ipt4}
// Missing v6 must not break the v4 path: open v4 only and continue.
if wgIface.Address().HasIPv6() {
ipt6, err := iptables.NewWithProtocol(iptables.ProtocolIPv6)
if err != nil {
log.Warnf("ip6tables not available, opening interface on v4 only: %v", err)
} else if _, err := ipt6.ListChains(tableFilter); err != nil {
log.Warnf("ip6tables filter table not available, opening interface on v4 only: %v", err)
} else {
a.ipt6 = ipt6
}
}
return a, nil
}
// Apply inserts the interface accept rule on the filter INPUT chain. It removes
// any stale rule first so an unclean exit (e.g. SIGKILL, where Close never ran)
// is recovered deterministically rather than accumulating duplicates.
func (a *InterfaceAllower) Apply() error {
var merr *multierror.Error
for _, ipt := range a.clients() {
if err := ipt.DeleteIfExists(tableFilter, chainInput, a.inputRule()...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("clean stale interface accept rule: %w", err))
}
if err := ipt.Insert(tableFilter, chainInput, 1, a.inputRule()...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add interface accept rule: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
// Close removes the interface accept rule.
func (a *InterfaceAllower) Close() error {
var merr *multierror.Error
for _, ipt := range a.clients() {
if err := ipt.DeleteIfExists(tableFilter, chainInput, a.inputRule()...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove interface accept rule: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
func (a *InterfaceAllower) inputRule() []string {
return []string{"-i", a.ifaceName, "-j", "ACCEPT"}
}
func (a *InterfaceAllower) clients() []*iptables.IPTables {
clients := []*iptables.IPTables{a.ipt4}
if a.ipt6 != nil {
clients = append(clients, a.ipt6)
}
return clients
}

View File

@@ -0,0 +1,104 @@
//go:build !android
package iptables
import (
"fmt"
"net/netip"
"github.com/hashicorp/go-multierror"
"github.com/lrh3321/ipset-go"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
)
func (r *family) createIpSet(setName string, sources []netip.Prefix) error {
if err := r.createIPSet(setName); err != nil {
return fmt.Errorf("create set %s: %w", setName, err)
}
for _, prefix := range sources {
if err := r.addPrefixToIPSet(setName, prefix); err != nil {
// The refcounter records nothing when this callback errors,
// so destroy the set or it leaks in the kernel. A partial
// source set would also fail-open for deny rules, so the
// rule must fail rather than install with a missing source.
if derr := r.destroyIPSet(setName); derr != nil {
log.Warnf("rollback ipset %s after add failure: %v", setName, derr)
}
return fmt.Errorf("add element to set %s: %w", setName, err)
}
}
return nil
}
func (r *family) deleteIpSet(setName string) error {
if err := r.destroyIPSet(setName); err != nil {
return fmt.Errorf("destroy set %s: %w", setName, err)
}
log.Debugf("deleted unused ipset %s", setName)
return nil
}
func (r *family) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
name := r.ipsetName(set.HashedName())
var merr *multierror.Error
for _, prefix := range prefixes {
if err := r.addPrefixToIPSet(name, prefix); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add prefix to ipset: %w", err))
}
}
if merr == nil {
log.Debugf("updated set %s with prefixes %v", name, prefixes)
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) ipsetName(name string) string {
if r.v6 {
return name + "-v6"
}
return name
}
func (r *family) createIPSet(name string) error {
opts := ipset.CreateOptions{
Replace: true,
}
if r.v6 {
opts.Family = ipset.FamilyIPV6
}
if err := ipset.Create(name, ipset.TypeHashNet, opts); err != nil {
return fmt.Errorf("create ipset %s: %w", name, err)
}
log.Debugf("created ipset %s with type hash:net", name)
return nil
}
func (r *family) addPrefixToIPSet(name string, prefix netip.Prefix) error {
addr := prefix.Addr()
ip := addr.AsSlice()
entry := &ipset.Entry{
IP: ip,
CIDR: uint8(prefix.Bits()),
Replace: true,
}
if err := ipset.Add(name, entry); err != nil {
return fmt.Errorf("add prefix to ipset %s: %w", name, err)
}
return nil
}
func (r *family) destroyIPSet(name string) error {
return ipset.Destroy(name)
}

View File

@@ -3,7 +3,6 @@ package iptables
import (
"context"
"fmt"
"net"
"net/netip"
"sync"
@@ -18,25 +17,21 @@ import (
"github.com/netbirdio/netbird/client/internal/statemanager"
)
type resetter interface {
Reset() error
}
// Manager of iptables firewall
// Manager of iptables firewall. Per-family state (peer ACLs, route
// ACLs, NAT, DNAT, MSS clamping) lives on family; Manager dispatches
// by family and provides the public firewall.Manager surface.
type Manager struct {
mutex sync.Mutex
wgIface iFaceMapper
ipv4Client *iptables.IPTables
aclMgr *aclManager
router *router
family4 *family
rawSupported bool
// IPv6 counterparts, nil when no v6 overlay
ipv6Client *iptables.IPTables
aclMgr6 *aclManager
router6 *router
family6 *family
}
// iFaceMapper defines subset methods of interface required for manager
@@ -57,14 +52,9 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
ipv4Client: iptablesClient,
}
m.router, err = newRouter(iptablesClient, wgIface, mtu)
m.family4, err = newFamily(iptablesClient, wgIface, mtu)
if err != nil {
return nil, fmt.Errorf("create router: %w", err)
}
m.aclMgr, err = newAclManager(iptablesClient, wgIface)
if err != nil {
return nil, fmt.Errorf("create acl manager: %w", err)
return nil, fmt.Errorf("create family: %w", err)
}
if wgIface.Address().HasIPv6() {
@@ -81,21 +71,18 @@ func (m *Manager) createIPv6Components(wgIface iFaceMapper, mtu uint16) error {
if err != nil {
return fmt.Errorf("init ip6tables: %w", err)
}
m.ipv6Client = ip6Client
m.router6, err = newRouter(ip6Client, wgIface, mtu)
family6, err := newFamily(ip6Client, wgIface, mtu)
if err != nil {
return fmt.Errorf("create v6 router: %w", err)
return fmt.Errorf("create v6 family: %w", err)
}
// Share the same IP forwarding state with the v4 router, since
// Share the same IP forwarding state with the v4 family, since
// EnableIPForwarding controls both v4 and v6 sysctls.
m.router6.ipFwdState = m.router.ipFwdState
family6.ipFwdState = m.family4.ipFwdState
m.aclMgr6, err = newAclManager(ip6Client, wgIface)
if err != nil {
return fmt.Errorf("create v6 acl manager: %w", err)
}
m.ipv6Client = ip6Client
m.family6 = family6
return nil
}
@@ -109,7 +96,7 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
InterfaceState: &InterfaceState{
NameStr: m.wgIface.Name(),
WGAddress: m.wgIface.Address(),
MTU: m.router.mtu,
MTU: m.family4.mtu,
},
}
stateManager.RegisterState(state)
@@ -141,31 +128,24 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
return nil
}
// initChains initializes router and ACL chains for both address families,
// rolling back on failure.
// initChains initializes the per-family firewall state for both
// address families, rolling back on failure.
func (m *Manager) initChains(stateManager *statemanager.Manager) error {
type initStep struct {
name string
init func(*statemanager.Manager) error
mgr resetter
r *family
}
steps := []initStep{
{"router", m.router.init, m.router},
{"acl manager", m.aclMgr.init, m.aclMgr},
}
steps := []initStep{{"v4", m.family4}}
if m.hasIPv6() {
steps = append(steps,
initStep{"v6 router", m.router6.init, m.router6},
initStep{"v6 acl manager", m.aclMgr6.init, m.aclMgr6},
)
steps = append(steps, initStep{"v6", m.family6})
}
var initialized []initStep
for _, s := range steps {
if err := s.init(stateManager); err != nil {
if err := s.r.init(stateManager); err != nil {
for i := len(initialized) - 1; i >= 0; i-- {
if rerr := initialized[i].mgr.Reset(); rerr != nil {
if rerr := initialized[i].r.Reset(); rerr != nil {
log.Warnf("rollback %s: %v", initialized[i].name, rerr)
}
}
@@ -176,84 +156,50 @@ func (m *Manager) initChains(stateManager *statemanager.Manager) error {
return nil
}
// AddPeerFiltering adds a rule to the firewall
//
// Comment will be ignored because some system this feature is not supported
func (m *Manager) AddPeerFiltering(
id []byte,
ip net.IP,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
ipsetName string,
) ([]firewall.Rule, error) {
m.mutex.Lock()
defer m.mutex.Unlock()
if ip.To4() != nil {
return m.aclMgr.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
}
if !m.hasIPv6() {
return nil, fmt.Errorf("add peer filtering for %s: %w", ip, firewall.ErrIPv6NotInitialized)
}
return m.aclMgr6.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
}
func (m *Manager) AddRouteFiltering(
// AddFilterRule installs a packet-filtering rule. See firewall.Manager
// docs for destination semantics. Sources are a single address family;
// the rule is dispatched to the matching v4 / v6 backend.
func (m *Manager) AddFilterRule(
id []byte,
sources []netip.Prefix,
destination firewall.Network,
proto firewall.Protocol,
sPort, dPort *firewall.Port,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
) (firewall.Rule, error) {
if len(sources) == 0 {
return nil, firewall.ErrNoSources
}
m.mutex.Lock()
defer m.mutex.Unlock()
if isIPv6RouteRule(sources, destination) {
fam := m.family4
if isIPv6Rule(sources, destination) {
if !m.hasIPv6() {
return nil, fmt.Errorf("add route filtering: %w", firewall.ErrIPv6NotInitialized)
return nil, fmt.Errorf("add filtering: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
fam = m.family6
}
return m.router.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
return fam.AddFilterRule(id, sources, destination, proto, sPort, dPort, action)
}
func isIPv6RouteRule(sources []netip.Prefix, destination firewall.Network) bool {
if destination.IsPrefix() {
return destination.Prefix.Addr().Is6()
}
return len(sources) > 0 && sources[0].Addr().Is6()
}
// DeletePeerRule from the firewall by rule definition
func (m *Manager) DeletePeerRule(rule firewall.Rule) error {
// DeleteFilterRule removes a rule previously added via AddFilterRule.
// The rule is looked up by id in each family's filter cache.
func (m *Manager) DeleteFilterRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
if m.hasIPv6() && isIPv6IptRule(rule) {
return m.aclMgr6.DeletePeerRule(rule)
id := rule.ID()
if m.family4.hasRule(id) {
return m.family4.DeleteFilterRule(rule)
}
return m.aclMgr.DeletePeerRule(rule)
}
func isIPv6IptRule(rule firewall.Rule) bool {
r, ok := rule.(*Rule)
return ok && r.v6
}
// DeleteRouteRule deletes a routing rule.
// Route rules are keyed by content hash. Check v4 first, try v6 if not found.
func (m *Manager) DeleteRouteRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
if m.hasIPv6() && !m.router.hasRule(rule.ID()) {
return m.router6.DeleteRouteRule(rule)
if m.hasIPv6() && m.family6.hasRule(id) {
return m.family6.DeleteFilterRule(rule)
}
return m.router.DeleteRouteRule(rule)
log.Debugf("filter rule %s not found in any family", id)
return nil
}
func (m *Manager) IsServerRouteSupported() bool {
@@ -272,10 +218,10 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
if !m.hasIPv6() {
return fmt.Errorf("add NAT rule: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddNatRule(pair)
return m.family6.AddNatRule(pair)
}
if err := m.router.AddNatRule(pair); err != nil {
if err := m.family4.AddNatRule(pair); err != nil {
return err
}
@@ -284,7 +230,7 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
// wildcard 0.0.0.0/0 destination where the client resolves DNS.
if m.hasIPv6() && pair.Dynamic {
v6Pair := firewall.ToV6NatPair(pair)
if err := m.router6.AddNatRule(v6Pair); err != nil {
if err := m.family6.AddNatRule(v6Pair); err != nil {
return fmt.Errorf("add v6 NAT rule: %w", err)
}
}
@@ -300,18 +246,18 @@ func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
if !m.hasIPv6() {
return nil
}
return m.router6.RemoveNatRule(pair)
return m.family6.RemoveNatRule(pair)
}
var merr *multierror.Error
if err := m.router.RemoveNatRule(pair); err != nil {
if err := m.family4.RemoveNatRule(pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove v4 NAT rule: %w", err))
}
if m.hasIPv6() && pair.Dynamic {
v6Pair := firewall.ToV6NatPair(pair)
if err := m.router6.RemoveNatRule(v6Pair); err != nil {
if err := m.family6.RemoveNatRule(v6Pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove v6 NAT rule: %w", err))
}
}
@@ -320,11 +266,14 @@ func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
}
func (m *Manager) SetLegacyManagement(isLegacy bool) error {
if err := firewall.SetLegacyManagement(m.router, isLegacy); err != nil {
m.mutex.Lock()
defer m.mutex.Unlock()
if err := firewall.SetLegacyManagement(m.family4, isLegacy); err != nil {
return err
}
if m.hasIPv6() {
return firewall.SetLegacyManagement(m.router6, isLegacy)
return firewall.SetLegacyManagement(m.family6, isLegacy)
}
return nil
}
@@ -341,19 +290,13 @@ func (m *Manager) Close(stateManager *statemanager.Manager) error {
}
if m.hasIPv6() {
if err := m.aclMgr6.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset v6 acl manager: %w", err))
}
if err := m.router6.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset v6 router: %w", err))
if err := m.family6.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset v6 family: %w", err))
}
}
if err := m.aclMgr.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset acl manager: %w", err))
}
if err := m.router.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset router: %w", err))
if err := m.family4.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset family: %w", err))
}
// Appending to merr intentionally blocks DeleteState below so ShutdownState
@@ -372,27 +315,6 @@ func (m *Manager) Close(stateManager *statemanager.Manager) error {
return nberrors.FormatErrorOrNil(merr)
}
// AllowNetbird allows netbird interface traffic.
// This is called when USPFilter wraps the native firewall, adding blanket accept
// rules so that packet filtering is handled in userspace instead of by netfilter.
func (m *Manager) AllowNetbird() error {
var merr *multierror.Error
if _, err := m.AddPeerFiltering(nil, net.IP{0, 0, 0, 0}, firewall.ProtocolALL, nil, nil, firewall.ActionAccept, ""); err != nil {
merr = multierror.Append(merr, fmt.Errorf("allow netbird v4 interface traffic: %w", err))
}
if m.hasIPv6() {
if _, err := m.AddPeerFiltering(nil, net.IPv6zero, firewall.ProtocolALL, nil, nil, firewall.ActionAccept, ""); err != nil {
merr = multierror.Append(merr, fmt.Errorf("allow netbird v6 interface traffic: %w", err))
}
}
if err := firewalld.TrustInterface(m.wgIface.Name()); err != nil {
log.Warnf("failed to trust interface in firewalld: %v", err)
}
return nberrors.FormatErrorOrNil(merr)
}
// Flush doesn't need to be implemented for this manager
func (m *Manager) Flush() error { return nil }
@@ -402,14 +324,14 @@ func (m *Manager) SetLogLevel(log.Level) {
}
func (m *Manager) EnableRouting() error {
if err := m.router.ipFwdState.RequestForwarding(); err != nil {
if err := m.family4.ipFwdState.RequestForwarding(); err != nil {
return fmt.Errorf("enable IP forwarding: %w", err)
}
return nil
}
func (m *Manager) DisableRouting() error {
if err := m.router.ipFwdState.ReleaseForwarding(); err != nil {
if err := m.family4.ipFwdState.ReleaseForwarding(); err != nil {
return fmt.Errorf("disable IP forwarding: %w", err)
}
return nil
@@ -424,9 +346,9 @@ func (m *Manager) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error)
if !m.hasIPv6() {
return nil, fmt.Errorf("add DNAT rule: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddDNATRule(rule)
return m.family6.AddDNATRule(rule)
}
return m.router.AddDNATRule(rule)
return m.family4.AddDNATRule(rule)
}
// DeleteDNATRule deletes a DNAT rule
@@ -434,10 +356,10 @@ func (m *Manager) DeleteDNATRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
if m.hasIPv6() && !m.router.hasRule(rule.ID()+dnatSuffix) {
return m.router6.DeleteDNATRule(rule)
if m.hasIPv6() && !m.family4.hasDNATRule(rule.ID()) {
return m.family6.DeleteDNATRule(rule)
}
return m.router.DeleteDNATRule(rule)
return m.family4.DeleteDNATRule(rule)
}
// UpdateSet updates the set with the given prefixes
@@ -454,12 +376,12 @@ func (m *Manager) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
}
}
if err := m.router.UpdateSet(set, v4Prefixes); err != nil {
if err := m.family4.UpdateSet(set, v4Prefixes); err != nil {
return err
}
if m.hasIPv6() && len(v6Prefixes) > 0 {
if err := m.router6.UpdateSet(set, v6Prefixes); err != nil {
if err := m.family6.UpdateSet(set, v6Prefixes); err != nil {
return fmt.Errorf("update v6 set: %w", err)
}
}
@@ -476,9 +398,9 @@ func (m *Manager) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protoco
if !m.hasIPv6() {
return fmt.Errorf("add inbound DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
// RemoveInboundDNAT removes an inbound DNAT rule.
@@ -490,9 +412,9 @@ func (m *Manager) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Prot
if !m.hasIPv6() {
return fmt.Errorf("remove inbound DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
@@ -504,9 +426,9 @@ func (m *Manager) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol
if !m.hasIPv6() {
return fmt.Errorf("add output DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
@@ -518,14 +440,14 @@ func (m *Manager) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Proto
if !m.hasIPv6() {
return fmt.Errorf("remove output DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
const (
chainNameRaw = "NETBIRD-RAW"
chainOUTPUT = "OUTPUT"
chainOutput = "OUTPUT"
tableRaw = "raw"
)
@@ -600,15 +522,15 @@ func (m *Manager) initNoTrackChain() error {
jumpRule := []string{"-j", chainNameRaw}
if err := m.ipv4Client.InsertUnique(tableRaw, chainOUTPUT, 1, jumpRule...); err != nil {
if err := m.ipv4Client.InsertUnique(tableRaw, chainOutput, 1, jumpRule...); err != nil {
if delErr := m.ipv4Client.DeleteChain(tableRaw, chainNameRaw); delErr != nil {
log.Debugf("delete orphan chain: %v", delErr)
}
return fmt.Errorf("add output jump rule: %w", err)
}
if err := m.ipv4Client.InsertUnique(tableRaw, chainPREROUTING, 1, jumpRule...); err != nil {
if delErr := m.ipv4Client.DeleteIfExists(tableRaw, chainOUTPUT, jumpRule...); delErr != nil {
if err := m.ipv4Client.InsertUnique(tableRaw, chainPrerouting, 1, jumpRule...); err != nil {
if delErr := m.ipv4Client.DeleteIfExists(tableRaw, chainOutput, jumpRule...); delErr != nil {
log.Debugf("delete output jump rule: %v", delErr)
}
if delErr := m.ipv4Client.DeleteChain(tableRaw, chainNameRaw); delErr != nil {
@@ -635,11 +557,11 @@ func (m *Manager) cleanupNoTrackChain() error {
jumpRule := []string{"-j", chainNameRaw}
if err := m.ipv4Client.DeleteIfExists(tableRaw, chainOUTPUT, jumpRule...); err != nil {
if err := m.ipv4Client.DeleteIfExists(tableRaw, chainOutput, jumpRule...); err != nil {
return fmt.Errorf("remove output jump rule: %w", err)
}
if err := m.ipv4Client.DeleteIfExists(tableRaw, chainPREROUTING, jumpRule...); err != nil {
if err := m.ipv4Client.DeleteIfExists(tableRaw, chainPrerouting, jumpRule...); err != nil {
return fmt.Errorf("remove prerouting jump rule: %w", err)
}
@@ -654,3 +576,13 @@ func (m *Manager) cleanupNoTrackChain() error {
func getConntrackEstablished() []string {
return []string{"-m", "conntrack", "--ctstate", "RELATED,ESTABLISHED", "-j", "ACCEPT"}
}
// isIPv6Rule reports whether the rule belongs to the IPv6 family, from
// the destination prefix when set, otherwise from the (single-family)
// sources.
func isIPv6Rule(sources []netip.Prefix, destination firewall.Network) bool {
if destination.IsPrefix() {
return destination.Prefix.Addr().Is6()
}
return len(sources) > 0 && sources[0].Addr().Is6()
}

View File

@@ -67,46 +67,39 @@ func TestIptablesManager(t *testing.T) {
time.Sleep(time.Second)
}()
var rule2 []fw.Rule
var rule2 fw.Rule
t.Run("add second rule", func(t *testing.T) {
ip := netip.MustParseAddr("10.20.0.3")
port := &fw.Port{
IsRange: true,
Values: []uint16{8043, 8046},
}
rule2, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", port, nil, fw.ActionAccept, "")
rule2, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", port, nil, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
for _, r := range rule2 {
rr := r.(*Rule)
checkRuleSpecs(t, ipv4Client, rr.chain, true, rr.specs...)
}
rr := rule2.(*Rule)
checkRuleSpecs(t, ipv4Client, rr.chain, true, rr.specs...)
})
t.Run("delete second rule", func(t *testing.T) {
for _, r := range rule2 {
err := manager.DeletePeerRule(r)
require.NoError(t, err, "failed to delete rule")
}
require.Empty(t, manager.aclMgr.ipsetStore.ipsets, "rulesets index after removed second rule must be empty")
require.NoError(t, manager.DeleteFilterRule(rule2), "failed to delete rule")
})
t.Run("reset check", func(t *testing.T) {
// add second rule
ip := netip.MustParseAddr("10.20.0.3")
port := &fw.Port{Values: []uint16{5353}}
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "udp", nil, port, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "udp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
err = manager.Close(nil)
require.NoError(t, err, "failed to reset")
ok, err := ipv4Client.ChainExists("filter", chainNameInputRules)
ok, err := ipv4Client.ChainExists("filter", chainACLInput)
require.NoError(t, err, "failed check chain exists")
if ok {
require.NoErrorf(t, err, "chain '%v' still exists after Close", chainNameInputRules)
require.NoErrorf(t, err, "chain '%v' still exists after Close", chainACLInput)
}
})
}
@@ -128,15 +121,13 @@ func TestIptablesManagerDenyRules(t *testing.T) {
ip := netip.MustParseAddr("10.20.0.3")
port := &fw.Port{Values: []uint16{22}}
rule, err := manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", nil, port, fw.ActionDrop, "deny-ssh")
rule, err := manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", nil, port, fw.ActionDrop)
require.NoError(t, err, "failed to add deny rule")
require.NotEmpty(t, rule, "deny rule should not be empty")
require.NotNil(t, rule, "deny rule should not be nil")
// Verify the rule was added by checking iptables
for _, r := range rule {
rr := r.(*Rule)
checkRuleSpecs(t, ipv4Client, rr.chain, true, rr.specs...)
}
rr := rule.(*Rule)
checkRuleSpecs(t, ipv4Client, rr.chain, true, rr.specs...)
})
t.Run("deny rule precedence test", func(t *testing.T) {
@@ -144,36 +135,40 @@ func TestIptablesManagerDenyRules(t *testing.T) {
port := &fw.Port{Values: []uint16{80}}
// Add accept rule first
_, err := manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", nil, port, fw.ActionAccept, "accept-http")
_, err := manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add accept rule")
// Add deny rule second for same IP/port - this should take precedence
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", nil, port, fw.ActionDrop, "deny-http")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", nil, port, fw.ActionDrop)
require.NoError(t, err, "failed to add deny rule")
// Inspect the actual iptables rules to verify deny rule comes before accept rule
rules, err := ipv4Client.List("filter", chainNameInputRules)
rules, err := ipv4Client.List("filter", chainACLInput)
require.NoError(t, err, "failed to list iptables rules")
// Debug: print all rules
t.Logf("All iptables rules in chain %s:", chainNameInputRules)
t.Logf("All iptables rules in chain %s:", chainACLInput)
for i, rule := range rules {
t.Logf(" [%d] %s", i, rule)
}
// Single-source rules emit a direct `-s <ip>/32 ... --dport 80`
// match. Match on that shape instead of the legacy
// per-(action,port) ipset names ("deny-http"/"accept-http")
// that this test predates.
srcMatch := fmt.Sprintf("-s %s/32", ip)
var denyRuleIndex, acceptRuleIndex = -1, -1
for i, rule := range rules {
if strings.Contains(rule, "DROP") {
t.Logf("Found DROP rule at index %d: %s", i, rule)
if strings.Contains(rule, "deny-http") && strings.Contains(rule, "80") {
denyRuleIndex = i
}
if !strings.Contains(rule, srcMatch) || !strings.Contains(rule, "--dport 80") {
continue
}
if strings.Contains(rule, "ACCEPT") {
if strings.Contains(rule, "-j DROP") {
t.Logf("Found DROP rule at index %d: %s", i, rule)
denyRuleIndex = i
}
if strings.Contains(rule, "-j ACCEPT") {
t.Logf("Found ACCEPT rule at index %d: %s", i, rule)
if strings.Contains(rule, "accept-http") && strings.Contains(rule, "80") {
acceptRuleIndex = i
}
acceptRuleIndex = i
}
}
@@ -198,7 +193,6 @@ func TestIptablesManagerIPSet(t *testing.T) {
},
}
// just check on the local interface
manager, err := Create(mock, iface.DefaultMTU)
require.NoError(t, err)
require.NoError(t, manager.Init(nil))
@@ -212,27 +206,39 @@ func TestIptablesManagerIPSet(t *testing.T) {
time.Sleep(time.Second)
}()
var rule2 []fw.Rule
t.Run("add second rule", func(t *testing.T) {
var rule2 fw.Rule
t.Run("single source uses direct -s match (no ipset)", func(t *testing.T) {
ip := netip.MustParseAddr("10.20.0.3")
port := &fw.Port{
Values: []uint16{443},
}
rule2, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", port, nil, fw.ActionAccept, "default")
for _, r := range rule2 {
require.NoError(t, err, "failed to add rule")
require.Equal(t, r.(*Rule).ipsetName, "default-sport", "ipset name must be set")
require.Equal(t, r.(*Rule).ip, "10.20.0.3", "ipset IP must be set")
}
rule2, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", port, nil, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
require.NotNil(t, rule2)
require.Contains(t, rule2.(*Rule).specs, "-s",
"single-source rule should use direct -s match, not an ipset")
require.Empty(t, findSets(rule2.(*Rule).specs),
"single-source rule should not allocate a shared ipset")
})
t.Run("delete second rule", func(t *testing.T) {
for _, r := range rule2 {
err := manager.DeletePeerRule(r)
require.NoError(t, err, "failed to delete rule")
t.Run("delete single-source rule", func(t *testing.T) {
require.NoError(t, manager.DeleteFilterRule(rule2), "failed to delete rule")
})
require.Empty(t, manager.aclMgr.ipsetStore.ipsets, "rulesets index after removed second rule must be empty")
t.Run("multi-source uses shared ipset", func(t *testing.T) {
sources := []netip.Prefix{
netip.PrefixFrom(netip.MustParseAddr("10.20.0.3"), 32),
netip.PrefixFrom(netip.MustParseAddr("10.20.0.4"), 32),
netip.PrefixFrom(netip.MustParseAddr("10.20.0.5"), 32),
}
port := &fw.Port{Values: []uint16{8080}}
multi, err := manager.AddFilterRule(nil, sources, fw.Network{}, "tcp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add multi-source rule")
require.NotNil(t, multi, "multi-source rule must produce one iptables rule")
sets := findSets(multi.(*Rule).specs)
require.Len(t, sets, 1, "multi-source rule must reference exactly one ipset")
require.NoError(t, manager.DeleteFilterRule(multi))
})
t.Run("reset check", func(t *testing.T) {
@@ -283,7 +289,7 @@ func TestIptablesCreatePerformance(t *testing.T) {
start := time.Now()
for i := 0; i < testMax; i++ {
port := &fw.Port{Values: []uint16{uint16(1000 + i)}}
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", nil, port, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
}

File diff suppressed because it is too large Load Diff

View File

@@ -31,7 +31,7 @@ func TestIptablesManager_RestoreOrCreateContainers(t *testing.T) {
iptablesClient, err := iptables.NewWithProtocol(iptables.ProtocolIPv4)
require.NoError(t, err, "failed to init iptables client")
manager, err := newRouter(iptablesClient, ifaceMock, iface.DefaultMTU)
manager, err := newFamily(iptablesClient, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "should return a valid iptables manager")
require.NoError(t, manager.init(nil))
@@ -52,12 +52,12 @@ func TestIptablesManager_RestoreOrCreateContainers(t *testing.T) {
// 11. MSS clamping rule for outbound traffic
require.Len(t, manager.rules, 11, "should have created rules map")
exists, err := manager.iptablesClient.Exists(tableNat, chainPOSTROUTING, "-j", chainRTNAT)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableNat, chainPOSTROUTING)
exists, err := manager.iptablesClient.Exists(tableNat, chainPostrouting, "-j", chainRTNAT)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableNat, chainPostrouting)
require.True(t, exists, "postrouting jump rule should exist")
exists, err = manager.iptablesClient.Exists(tableMangle, chainPREROUTING, "-j", chainRTPRE)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainPREROUTING)
exists, err = manager.iptablesClient.Exists(tableMangle, chainPrerouting, "-j", chainRTPre)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainPrerouting)
require.True(t, exists, "prerouting jump rule should exist")
pair := firewall.RouterPair{
@@ -84,7 +84,7 @@ func TestIptablesManager_AddNatRule(t *testing.T) {
iptablesClient, err := iptables.NewWithProtocol(iptables.ProtocolIPv4)
require.NoError(t, err, "failed to init iptables client")
manager, err := newRouter(iptablesClient, ifaceMock, iface.DefaultMTU)
manager, err := newFamily(iptablesClient, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "shouldn't return error")
require.NoError(t, manager.init(nil))
@@ -95,7 +95,7 @@ func TestIptablesManager_AddNatRule(t *testing.T) {
err = manager.AddNatRule(testCase.InputPair)
require.NoError(t, err, "marking rule should be inserted")
natRuleKey := firewall.GenKey(firewall.NatFormat, testCase.InputPair)
natRuleKey := testCase.InputPair.GenKey(firewall.NatFormat)
markingRule := []string{
"-i", ifaceMock.Name(),
"-m", "conntrack",
@@ -106,8 +106,8 @@ func TestIptablesManager_AddNatRule(t *testing.T) {
fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasquerade),
}
exists, err := iptablesClient.Exists(tableMangle, chainRTPRE, markingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPRE)
exists, err := iptablesClient.Exists(tableMangle, chainRTPre, markingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPre)
if testCase.InputPair.Masquerade {
require.True(t, exists, "marking rule should be created")
foundRule, found := manager.rules[natRuleKey]
@@ -121,7 +121,7 @@ func TestIptablesManager_AddNatRule(t *testing.T) {
// Check inverse rule
inversePair := firewall.GetInversePair(testCase.InputPair)
inverseRuleKey := firewall.GenKey(firewall.NatFormat, inversePair)
inverseRuleKey := inversePair.GenKey(firewall.NatFormat)
inverseMarkingRule := []string{
"!", "-i", ifaceMock.Name(),
"-m", "conntrack",
@@ -132,8 +132,8 @@ func TestIptablesManager_AddNatRule(t *testing.T) {
fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasqueradeReturn),
}
exists, err = iptablesClient.Exists(tableMangle, chainRTPRE, inverseMarkingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPRE)
exists, err = iptablesClient.Exists(tableMangle, chainRTPre, inverseMarkingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPre)
if testCase.InputPair.Masquerade {
require.True(t, exists, "inverse marking rule should be created")
foundRule, found := manager.rules[inverseRuleKey]
@@ -157,7 +157,7 @@ func TestIptablesManager_RemoveNatRule(t *testing.T) {
t.Run(testCase.Name, func(t *testing.T) {
iptablesClient, _ := iptables.NewWithProtocol(iptables.ProtocolIPv4)
manager, err := newRouter(iptablesClient, ifaceMock, iface.DefaultMTU)
manager, err := newFamily(iptablesClient, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "shouldn't return error")
require.NoError(t, manager.init(nil))
defer func() {
@@ -170,7 +170,7 @@ func TestIptablesManager_RemoveNatRule(t *testing.T) {
err = manager.RemoveNatRule(testCase.InputPair)
require.NoError(t, err, "shouldn't return error")
natRuleKey := firewall.GenKey(firewall.NatFormat, testCase.InputPair)
natRuleKey := testCase.InputPair.GenKey(firewall.NatFormat)
markingRule := []string{
"-i", ifaceMock.Name(),
"-m", "conntrack",
@@ -181,8 +181,8 @@ func TestIptablesManager_RemoveNatRule(t *testing.T) {
fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasquerade),
}
exists, err := iptablesClient.Exists(tableMangle, chainRTPRE, markingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPRE)
exists, err := iptablesClient.Exists(tableMangle, chainRTPre, markingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPre)
require.False(t, exists, "marking rule should not exist")
_, found := manager.rules[natRuleKey]
@@ -190,7 +190,7 @@ func TestIptablesManager_RemoveNatRule(t *testing.T) {
// Check inverse rule removal
inversePair := firewall.GetInversePair(testCase.InputPair)
inverseRuleKey := firewall.GenKey(firewall.NatFormat, inversePair)
inverseRuleKey := inversePair.GenKey(firewall.NatFormat)
inverseMarkingRule := []string{
"!", "-i", ifaceMock.Name(),
"-m", "conntrack",
@@ -201,8 +201,8 @@ func TestIptablesManager_RemoveNatRule(t *testing.T) {
fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasqueradeReturn),
}
exists, err = iptablesClient.Exists(tableMangle, chainRTPRE, inverseMarkingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPRE)
exists, err = iptablesClient.Exists(tableMangle, chainRTPre, inverseMarkingRule...)
require.NoError(t, err, "should be able to query the iptables %s table and %s chain", tableMangle, chainRTPre)
require.False(t, exists, "inverse marking rule should not exist")
_, found = manager.rules[inverseRuleKey]
@@ -219,13 +219,13 @@ func TestRouter_AddRouteFiltering(t *testing.T) {
iptablesClient, err := iptables.NewWithProtocol(iptables.ProtocolIPv4)
require.NoError(t, err, "Failed to create iptables client")
r, err := newRouter(iptablesClient, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "Failed to create router manager")
r, err := newFamily(iptablesClient, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "Failed to create family manager")
require.NoError(t, r.init(nil))
defer func() {
err := r.Reset()
require.NoError(t, err, "Failed to reset router")
require.NoError(t, err, "Failed to reset family")
}()
tests := []struct {
@@ -334,62 +334,30 @@ func TestRouter_AddRouteFiltering(t *testing.T) {
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
ruleKey, err := r.AddRouteFiltering(nil, tt.sources, firewall.Network{Prefix: tt.destination}, tt.proto, tt.sPort, tt.dPort, tt.action)
require.NoError(t, err, "AddRouteFiltering failed")
ruleKey, err := r.AddFilterRule(nil, tt.sources, firewall.Network{Prefix: tt.destination}, tt.proto, tt.sPort, tt.dPort, tt.action)
require.NoError(t, err, "AddFilterRule failed")
// Check if the rule is in the internal map
rule, ok := r.rules[ruleKey.ID()]
assert.True(t, ok, "Rule not found in internal map")
stored, ok := r.filters[ruleKey.ID()]
require.True(t, ok, "rule not stored in filters")
t.Logf("Internal rule: %v", stored.specs)
// Log the internal rule
t.Logf("Internal rule: %v", rule)
// Check if the rule exists in iptables
exists, err := iptablesClient.Exists(tableFilter, chainRTFWDIN, rule...)
exists, err := iptablesClient.Exists(tableFilter, chainRTFwdIn, stored.specs...)
assert.NoError(t, err, "Failed to check rule existence")
assert.True(t, exists, "Rule not found in iptables")
var source firewall.Network
if len(tt.sources) > 1 {
source.Set = firewall.NewPrefixSet(tt.sources)
} else if len(tt.sources) > 0 {
source.Prefix = tt.sources[0]
}
// Verify rule content
params := routeFilteringRuleParams{
Source: source,
Destination: firewall.Network{Prefix: tt.destination},
Proto: tt.proto,
SPort: tt.sPort,
DPort: tt.dPort,
Action: tt.action,
}
expectedRule, err := r.genRouteRuleSpec(params, nil)
require.NoError(t, err, "Failed to generate expected rule spec")
if tt.expectSet {
setName := firewall.NewPrefixSet(tt.sources).HashedName()
expectedRule, err = r.genRouteRuleSpec(params, nil)
require.NoError(t, err, "Failed to generate expected rule spec with set")
// Check if the set was created
_, exists := r.ipsetCounter.Get(setName)
assert.True(t, exists, "IPSet not created")
assert.NotEmpty(t, findSets(stored.specs), "Rule should reference an ipset")
}
assert.Equal(t, expectedRule, rule, "Rule content mismatch")
// Clean up
err = r.DeleteRouteRule(ruleKey)
require.NoError(t, err, "Failed to delete rule")
require.NoError(t, r.DeleteFilterRule(ruleKey), "Failed to delete rule")
})
}
}
func TestFindSetNameInRule(t *testing.T) {
r := &router{}
testCases := []struct {
name string
rule []string
@@ -430,7 +398,7 @@ func TestFindSetNameInRule(t *testing.T) {
for _, tc := range testCases {
t.Run(tc.name, func(t *testing.T) {
result := r.findSets(tc.rule)
result := findSets(tc.rule)
if len(result) != len(tc.expected) {
t.Errorf("Expected %d sets, got %d. Sets found: %v", len(tc.expected), len(result), result)

View File

@@ -0,0 +1,263 @@
//go:build !android
package iptables
import (
"fmt"
"strings"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnet "github.com/netbirdio/netbird/client/net"
)
func (r *family) AddNatRule(pair firewall.RouterPair) error {
if r.legacyManagement {
log.Warnf("This peer is connected to a NetBird Management service with an older version. Allowing all traffic for %s", pair.Destination)
if err := r.addLegacyRouteRule(pair); err != nil {
return fmt.Errorf("add legacy routing rule: %w", err)
}
}
if pair.Masquerade {
if err := r.addNatRule(pair); err != nil {
return fmt.Errorf("add nat rule: %w", err)
}
if err := r.addNatRule(firewall.GetInversePair(pair)); err != nil {
return fmt.Errorf("add inverse nat rule: %w", err)
}
}
r.updateState()
return nil
}
// RemoveNatRule removes an iptables rule pair from forwarding and nat chains
func (r *family) RemoveNatRule(pair firewall.RouterPair) error {
if pair.Masquerade {
if err := r.removeNatRule(pair); err != nil {
return fmt.Errorf("remove nat rule: %w", err)
}
if err := r.removeNatRule(firewall.GetInversePair(pair)); err != nil {
return fmt.Errorf("remove inverse nat rule: %w", err)
}
}
if err := r.removeLegacyRouteRule(pair); err != nil {
return fmt.Errorf("remove legacy routing rule: %w", err)
}
r.updateState()
return nil
}
// addLegacyRouteRule adds a legacy routing rule for mgmt servers pre route acls
func (r *family) addLegacyRouteRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.ForwardingFormat)
if err := r.removeLegacyRouteRule(pair); err != nil {
return err
}
rule := []string{"-s", pair.Source.String(), "-d", pair.Destination.String(), "-j", "ACCEPT"}
if err := r.iptablesClient.Append(tableFilter, chainRTFwdIn, rule...); err != nil {
return fmt.Errorf("add legacy forwarding rule %s -> %s: %w", pair.Source, pair.Destination, err)
}
r.rules[ruleID] = rule
return nil
}
func (r *family) removeLegacyRouteRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.ForwardingFormat)
if rule, exists := r.rules[ruleID]; exists {
if err := r.iptablesClient.DeleteIfExists(tableFilter, chainRTFwdIn, rule...); err != nil {
return fmt.Errorf("remove legacy forwarding rule %s -> %s: %w", pair.Source, pair.Destination, err)
}
delete(r.rules, ruleID)
if err := r.decrementSetCounter(rule); err != nil {
return fmt.Errorf("decrement ipset counter: %w", err)
}
}
return nil
}
// GetLegacyManagement returns the current legacy management mode
func (r *family) GetLegacyManagement() bool {
return r.legacyManagement
}
// SetLegacyManagement sets the route manager to use legacy management mode
func (r *family) SetLegacyManagement(isLegacy bool) {
r.legacyManagement = isLegacy
}
// RemoveAllLegacyRouteRules removes all legacy routing rules for mgmt servers pre route acls
func (r *family) RemoveAllLegacyRouteRules() error {
var merr *multierror.Error
for k, rule := range r.rules {
if !strings.HasPrefix(string(k), firewall.ForwardingFormatPrefix) {
continue
}
if err := r.iptablesClient.DeleteIfExists(tableFilter, chainRTFwdIn, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove legacy forwarding rule: %w", err))
} else {
delete(r.rules, k)
}
}
r.updateState()
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) addPostroutingRules() error {
// First rule for outbound masquerade
rule1 := []string{
"-m", "mark", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasquerade),
"!", "-o", "lo",
"-j", "MASQUERADE",
}
if err := r.iptablesClient.Append(tableNat, chainRTNAT, rule1...); err != nil {
return fmt.Errorf("add outbound masquerade rule: %w", err)
}
r.rules["static-nat-outbound"] = rule1
// Second rule for return traffic masquerade
rule2 := []string{
"-m", "mark", "--mark", fmt.Sprintf("%#x", nbnet.PreroutingFwmarkMasqueradeReturn),
"-o", r.wgIface.Name(),
"-j", "MASQUERADE",
}
if err := r.iptablesClient.Append(tableNat, chainRTNAT, rule2...); err != nil {
return fmt.Errorf("add return masquerade rule: %w", err)
}
r.rules["static-nat-return"] = rule2
return nil
}
// addMSSClampingRules adds MSS clamping rules to prevent fragmentation for forwarded traffic.
func (r *family) addMSSClampingRules() error {
overhead := uint16(ipv4TCPHeaderSize)
if r.v6 {
overhead = ipv6TCPHeaderSize
}
mss := r.mtu - overhead
// Add jump rule from FORWARD chain in mangle table to our custom chain
jumpRule := jumpRuleSpec(chainRTMSSClamp)
if err := r.iptablesClient.Insert(tableMangle, chainForward, 1, jumpRule...); err != nil {
return fmt.Errorf("add jump to MSS clamp chain: %w", err)
}
r.rules[jumpMSSClamp] = jumpRule
ruleOut := []string{
"-o", r.wgIface.Name(),
"-p", "tcp",
"--tcp-flags", "SYN,RST", "SYN",
"-j", "TCPMSS",
"--set-mss", fmt.Sprintf("%d", mss),
}
if err := r.iptablesClient.Append(tableMangle, chainRTMSSClamp, ruleOut...); err != nil {
return fmt.Errorf("add outbound MSS clamp rule: %w", err)
}
r.rules["mss-clamp-out"] = ruleOut
return nil
}
func (r *family) insertEstablishedRule(chain string) error {
establishedRule := getConntrackEstablished()
err := r.iptablesClient.Insert(tableFilter, chain, 1, establishedRule...)
if err != nil {
return fmt.Errorf("insert established rule: %w", err)
}
ruleID := firewall.RuleID("established-" + chain)
r.rules[ruleID] = establishedRule
return nil
}
func (r *family) addNatRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.NatFormat)
if rule, exists := r.rules[ruleID]; exists {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainRTPre, rule...); err != nil {
return fmt.Errorf("remove existing marking rule for %s: %w", pair.Destination, err)
}
delete(r.rules, ruleID)
}
markValue := nbnet.PreroutingFwmarkMasquerade
if pair.Inverse {
markValue = nbnet.PreroutingFwmarkMasqueradeReturn
}
rule := []string{"-i", r.wgIface.Name()}
if pair.Inverse {
rule = []string{"!", "-i", r.wgIface.Name()}
}
rule = append(rule,
"-m", "conntrack",
"--ctstate", "NEW",
)
sourceExp, err := r.applyNetwork("-s", pair.Source, nil)
if err != nil {
return fmt.Errorf("apply network -s: %w", err)
}
destExp, err := r.applyNetwork("-d", pair.Destination, nil)
if err != nil {
return fmt.Errorf("apply network -d: %w", err)
}
rule = append(rule, sourceExp...)
rule = append(rule, destExp...)
rule = append(rule,
"-j", "MARK", "--set-mark", fmt.Sprintf("%#x", markValue),
)
// Ensure nat rules come first, so the mark can be overwritten.
// Currently overwritten by the dst-type LOCAL rules for redirected traffic.
if err := r.iptablesClient.Insert(tableMangle, chainRTPre, 1, rule...); err != nil {
r.dropSourceMatch(rule)
return fmt.Errorf("add marking rule for %s: %w", pair.Destination, err)
}
r.rules[ruleID] = rule
return nil
}
func (r *family) removeNatRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.NatFormat)
if rule, exists := r.rules[ruleID]; exists {
if err := r.iptablesClient.DeleteIfExists(tableMangle, chainRTPre, rule...); err != nil {
return fmt.Errorf("remove marking rule for %s: %w", pair.Destination, err)
}
delete(r.rules, ruleID)
if err := r.decrementSetCounter(rule); err != nil {
return fmt.Errorf("decrement ipset counter: %w", err)
}
} else {
log.Debugf("marking rule %s not found", ruleID)
}
return nil
}

View File

@@ -1,18 +1,20 @@
package iptables
// Rule to handle management of rules
type Rule struct {
ruleID string
ipsetName string
import "github.com/netbirdio/netbird/client/firewall/manager"
// Rule to handle management of rules. Source set membership (when the
// rule was built against a shared hash:net ipset) is encoded in specs;
// DeleteFilterRule recovers it via findSets so the refcounter can drop
// the right reference.
type Rule struct {
id manager.RuleID
specs []string
mangleSpecs []string
ip string
chain string
v6 bool
}
// GetRuleID returns the rule id
func (r *Rule) ID() string {
return r.ruleID
// ID returns the rule id
func (r *Rule) ID() manager.RuleID {
return r.id
}

View File

@@ -1,127 +0,0 @@
package iptables
import (
"encoding/json"
"maps"
)
type ipList struct {
ips map[string]struct{}
}
func newIpList(ip string) *ipList {
ips := make(map[string]struct{})
ips[ip] = struct{}{}
return &ipList{
ips: ips,
}
}
func (s *ipList) addIP(ip string) {
s.ips[ip] = struct{}{}
}
// clone returns a deep copy of the ipList with its own ips map.
func (s *ipList) clone() *ipList {
if s == nil {
return nil
}
return &ipList{ips: maps.Clone(s.ips)}
}
// MarshalJSON implements json.Marshaler
func (s *ipList) MarshalJSON() ([]byte, error) {
return json.Marshal(struct {
IPs map[string]struct{} `json:"ips"`
}{
IPs: s.ips,
})
}
// UnmarshalJSON implements json.Unmarshaler
func (s *ipList) UnmarshalJSON(data []byte) error {
temp := struct {
IPs map[string]struct{} `json:"ips"`
}{}
if err := json.Unmarshal(data, &temp); err != nil {
return err
}
s.ips = temp.IPs
if temp.IPs == nil {
temp.IPs = make(map[string]struct{})
}
return nil
}
type ipsetStore struct {
ipsets map[string]*ipList
}
func newIpsetStore() *ipsetStore {
return &ipsetStore{
ipsets: make(map[string]*ipList),
}
}
// clone returns a deep copy of the ipsetStore with its own ipsets map and
// independent ipList entries.
func (s *ipsetStore) clone() *ipsetStore {
if s == nil {
return nil
}
cloned := &ipsetStore{ipsets: make(map[string]*ipList, len(s.ipsets))}
for name, list := range s.ipsets {
cloned.ipsets[name] = list.clone()
}
return cloned
}
func (s *ipsetStore) ipset(ipsetName string) (*ipList, bool) {
r, ok := s.ipsets[ipsetName]
return r, ok
}
func (s *ipsetStore) addIpList(ipsetName string, list *ipList) {
s.ipsets[ipsetName] = list
}
func (s *ipsetStore) deleteIpset(ipsetName string) {
delete(s.ipsets, ipsetName)
}
func (s *ipsetStore) ipsetNames() []string {
names := make([]string, 0, len(s.ipsets))
for name := range s.ipsets {
names = append(names, name)
}
return names
}
// MarshalJSON implements json.Marshaler
func (s *ipsetStore) MarshalJSON() ([]byte, error) {
return json.Marshal(struct {
IPSets map[string]*ipList `json:"ipsets"`
}{
IPSets: s.ipsets,
})
}
// UnmarshalJSON implements json.Unmarshaler
func (s *ipsetStore) UnmarshalJSON(data []byte) error {
temp := struct {
IPSets map[string]*ipList `json:"ipsets"`
}{}
if err := json.Unmarshal(data, &temp); err != nil {
return err
}
s.ipsets = temp.IPSets
if temp.IPSets == nil {
temp.IPSets = make(map[string]*ipList)
}
return nil
}

View File

@@ -29,17 +29,13 @@ type ShutdownState struct {
InterfaceState *InterfaceState `json:"interface_state,omitempty"`
RouteRules routeRules `json:"route_rules,omitempty"`
RouteIPsetCounter *ipsetCounter `json:"route_ipset_counter,omitempty"`
ACLEntries aclEntries `json:"acl_entries,omitempty"`
ACLIPsetStore *ipsetStore `json:"acl_ipset_store,omitempty"`
// IPv6 counterparts
RouteRules routeRules `json:"route_rules,omitempty"`
RouteRules6 routeRules `json:"route_rules_v6,omitempty"`
RouteIPsetCounter *ipsetCounter `json:"route_ipset_counter,omitempty"`
RouteIPsetCounter6 *ipsetCounter `json:"route_ipset_counter_v6,omitempty"`
ACLEntries6 aclEntries `json:"acl_entries_v6,omitempty"`
ACLIPsetStore6 *ipsetStore `json:"acl_ipset_store_v6,omitempty"`
ACLEntries aclEntries `json:"acl_entries,omitempty"`
ACLEntries6 aclEntries `json:"acl_entries_v6,omitempty"`
}
func (s *ShutdownState) Name() string {
@@ -57,17 +53,14 @@ func (s *ShutdownState) Cleanup() error {
}
if s.RouteRules != nil {
ipt.router.rules = s.RouteRules
ipt.family4.rules = s.RouteRules
}
if s.RouteIPsetCounter != nil {
ipt.router.ipsetCounter.LoadData(s.RouteIPsetCounter)
ipt.family4.ipsetCounter.LoadData(s.RouteIPsetCounter)
}
if s.ACLEntries != nil {
ipt.aclMgr.entries = s.ACLEntries
}
if s.ACLIPsetStore != nil {
ipt.aclMgr.ipsetStore = s.ACLIPsetStore
ipt.family4.entries = s.ACLEntries
}
// Clean up v6 state even if the current run has no IPv6.
@@ -79,16 +72,13 @@ func (s *ShutdownState) Cleanup() error {
}
if ipt.hasIPv6() {
if s.RouteRules6 != nil {
ipt.router6.rules = s.RouteRules6
ipt.family6.rules = s.RouteRules6
}
if s.RouteIPsetCounter6 != nil {
ipt.router6.ipsetCounter.LoadData(s.RouteIPsetCounter6)
ipt.family6.ipsetCounter.LoadData(s.RouteIPsetCounter6)
}
if s.ACLEntries6 != nil {
ipt.aclMgr6.entries = s.ACLEntries6
}
if s.ACLIPsetStore6 != nil {
ipt.aclMgr6.ipsetStore = s.ACLIPsetStore6
ipt.family6.entries = s.ACLEntries6
}
}

View File

@@ -0,0 +1,27 @@
//go:build privileged
package iptables
import (
"fmt"
"net"
"net/netip"
)
func pfx(ip net.IP) []netip.Prefix {
if ip == nil {
return []netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)}
}
if ip.IsUnspecified() {
if ip.To4() != nil {
return []netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)}
}
return []netip.Prefix{netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
}
a, ok := netip.AddrFromSlice(ip)
if !ok {
panic(fmt.Sprintf("invalid IP length: %d", len(ip)))
}
a = a.Unmap()
return []netip.Prefix{netip.PrefixFrom(a, a.BitLen())}
}

View File

@@ -3,7 +3,6 @@ package manager
import (
"errors"
"fmt"
"net"
"net/netip"
"sort"
@@ -16,6 +15,12 @@ import (
// method but the IPv6 firewall components were not initialized.
var ErrIPv6NotInitialized = errors.New("IPv6 firewall not initialized")
// ErrNoSources is returned when AddFilterRule is called with an empty
// source list. "Match any source" must be expressed explicitly with a
// /0 prefix; an empty list is a caller error and is rejected rather
// than silently widening the rule to every source.
var ErrNoSources = errors.New("rule has no sources")
const (
ForwardingFormatPrefix = "netbird-fwd-"
ForwardingFormat = "netbird-fwd-%s-%t"
@@ -23,13 +28,18 @@ const (
NatFormat = "netbird-nat-%s-%t"
)
// RuleID identifies a firewall rule. It is a typed string so the
// compiler catches accidental mixing with arbitrary string keys. It is
// only an identifier and does not implement Rule.
type RuleID string
// Rule abstraction should be implemented by each firewall manager
//
// Each firewall type for different OS can use different type
// of the properties to hold data of the created rule
type Rule interface {
// ID returns the rule id
ID() string
ID() RuleID
}
// RuleDirection is the traffic direction which a rule is applied
@@ -91,6 +101,13 @@ func (d Network) IsPrefix() bool {
return d.Prefix.IsValid()
}
// IsZero returns true if the network designates no destination, i.e. it
// is the zero value. A zero Network is the peer-rule sentinel; a non-zero
// one carries a prefix or set destination.
func (d Network) IsZero() bool {
return !d.IsPrefix() && !d.IsSet()
}
// Manager is the high level abstraction of a firewall manager
//
// It declares methods which handle actions required by the
@@ -98,46 +115,42 @@ func (d Network) IsPrefix() bool {
type Manager interface {
Init(stateManager *statemanager.Manager) error
// AllowNetbird allows netbird interface traffic
AllowNetbird() error
// AddPeerFiltering adds a rule to the firewall
// AddFilterRule adds a packet-filtering rule to the firewall.
//
// If comment argument is empty firewall manager should set
// rule ID as comment for the rule
// If destination is the zero Network, the rule applies to traffic
// inbound to this node, i.e. peer ACL semantics, installed in
// the kernel's input chain. If destination is set (prefix or
// set), the rule applies to forwarded traffic with that
// destination, route ACL semantics, installed in the forward
// chain.
//
// Note: Callers should call Flush() after adding rules to ensure
// they are applied to the kernel and rule handles are refreshed.
AddPeerFiltering(
// sources must be a single address family; the caller splits mixed
// families and calls once per family. "Match any source" must be
// expressed with an explicit /0 prefix; an empty sources list is
// rejected with ErrNoSources so a zeroed list can never widen a
// rule to every source.
//
// Note: callers should call Flush() after adding rules.
AddFilterRule(
id []byte,
ip net.IP,
sources []netip.Prefix,
destination Network,
proto Protocol,
sPort *Port,
dPort *Port,
action Action,
ipsetName string,
) ([]Rule, error)
) (Rule, error)
// DeletePeerRule from the firewall by rule definition
DeletePeerRule(rule Rule) error
// DeleteFilterRule removes a filtering rule previously added via
// AddFilterRule. The rule's own type identifies whether it lives
// in the peer (input) or route (forward) path.
DeleteFilterRule(rule Rule) error
// IsServerRouteSupported returns true if the firewall supports server side routing operations
IsServerRouteSupported() bool
IsStateful() bool
AddRouteFiltering(
id []byte,
sources []netip.Prefix,
destination Network,
proto Protocol,
sPort, dPort *Port,
action Action,
) (Rule, error)
// DeleteRouteRule deletes a routing rule
DeleteRouteRule(rule Rule) error
// AddNatRule inserts a routing NAT rule
AddNatRule(pair RouterPair) error
@@ -185,8 +198,9 @@ type Manager interface {
SetupEBPFProxyNoTrack(proxyPort, wgPort uint16) error
}
func GenKey(format string, pair RouterPair) string {
return fmt.Sprintf(format, pair.ID, pair.Inverse)
// GenKey builds the rule id for this pair from the given format.
func (p RouterPair) GenKey(format string) RuleID {
return RuleID(fmt.Sprintf(format, p.ID, p.Inverse))
}
// LegacyManager defines the interface for legacy management operations
@@ -242,6 +256,20 @@ func MergeIPRanges(prefixes []netip.Prefix) []netip.Prefix {
return merged
}
// UnmapPrefix normalizes a v4-mapped v6 prefix (::ffff:a.b.c.d) to its
// plain v4 form, shifting the prefix length out of the 96-bit mapped
// range. Other prefixes are returned unchanged. Keeping prefixes
// unmapped ensures v4 rules match consistently and the match builders
// read the correct address length.
func UnmapPrefix(p netip.Prefix) netip.Prefix {
addr := p.Addr()
if !addr.Is4In6() {
return p
}
bits := max(p.Bits()-96, 0)
return netip.PrefixFrom(addr.Unmap(), bits)
}
// SortPrefixes sorts the given slice of netip.Prefix in place.
// It sorts first by IP address, then by prefix length (most specific to least specific).
func SortPrefixes(prefixes []netip.Prefix) {

View File

@@ -13,13 +13,13 @@ type ForwardRule struct {
TranslatedPort Port
}
func (r ForwardRule) ID() string {
func (r ForwardRule) ID() RuleID {
id := fmt.Sprintf("%s;%s;%s;%s",
r.Protocol,
r.DestinationPort.String(),
r.TranslatedAddress.String(),
r.TranslatedPort.String())
return id
return RuleID(id)
}
func (r ForwardRule) String() string {

View File

@@ -40,7 +40,7 @@ func (h Set) Comment() string {
// NewPrefixSet generates a unique name for an ipset based on the given prefixes.
func NewPrefixSet(prefixes []netip.Prefix) Set {
// sort for consistent naming
prefixes = slices.Clone(prefixes)
SortPrefixes(prefixes)
hash := sha256.New()

View File

@@ -1,713 +0,0 @@
package nftables
import (
"bytes"
"fmt"
"net"
"slices"
"strconv"
"strings"
"time"
"github.com/google/nftables"
"github.com/google/nftables/binaryutil"
"github.com/google/nftables/expr"
log "github.com/sirupsen/logrus"
"golang.org/x/sys/unix"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnet "github.com/netbirdio/netbird/client/net"
)
const (
// rules chains contains the effective ACL rules
chainNameInputRules = "netbird-acl-input-rules"
// filter chains contains the rules that jump to the rules chains
chainNameInputFilter = "netbird-acl-input-filter"
chainNameForwardFilter = "netbird-acl-forward-filter"
chainNameManglePrerouting = "netbird-mangle-prerouting"
chainNameManglePostrouting = "netbird-mangle-postrouting"
)
const flushError = "flush: %w"
type AclManager struct {
rConn *nftables.Conn
sConn *nftables.Conn
wgIface iFaceMapper
routingFwChainName string
af addrFamily
workTable *nftables.Table
chainInputRules *nftables.Chain
chainPrerouting *nftables.Chain
ipsetStore *ipsetStore
rules map[string]*Rule
}
func newAclManager(table *nftables.Table, wgIface iFaceMapper, routingFwChainName string) (*AclManager, error) {
// sConn is used for creating sets and adding/removing elements from them
// it's differ then rConn (which does create new conn for each flush operation)
// and is permanent. Using same connection for both type of operations
// overloads netlink with high amount of rules ( > 10000)
sConn, err := nftables.New(nftables.AsLasting())
if err != nil {
return nil, fmt.Errorf("create nf conn: %w", err)
}
return &AclManager{
rConn: &nftables.Conn{},
sConn: sConn,
wgIface: wgIface,
workTable: table,
routingFwChainName: routingFwChainName,
af: familyForAddr(table.Family == nftables.TableFamilyIPv4),
ipsetStore: newIpsetStore(),
rules: make(map[string]*Rule),
}, nil
}
func (m *AclManager) init(workTable *nftables.Table) error {
m.workTable = workTable
return m.createDefaultChains()
}
// AddPeerFiltering rule to the firewall
//
// If comment argument is empty firewall manager should set
// rule ID as comment for the rule
func (m *AclManager) AddPeerFiltering(
id []byte,
ip net.IP,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
ipsetName string,
) ([]firewall.Rule, error) {
var ipset *nftables.Set
if ipsetName != "" {
var err error
ipset, err = m.addIpToSet(ipsetName, ip)
if err != nil {
return nil, err
}
}
newRules := make([]firewall.Rule, 0, 2)
ioRule, err := m.addIOFiltering(ip, proto, sPort, dPort, action, ipset)
if err != nil {
return nil, err
}
newRules = append(newRules, ioRule)
return newRules, nil
}
// DeletePeerRule from the firewall by rule definition
func (m *AclManager) DeletePeerRule(rule firewall.Rule) error {
r, ok := rule.(*Rule)
if !ok {
return fmt.Errorf("invalid rule type")
}
if r.nftSet == nil {
if err := m.rConn.DelRule(r.nftRule); err != nil {
log.Errorf("failed to delete rule: %v", err)
}
if r.mangleRule != nil {
if err := m.rConn.DelRule(r.mangleRule); err != nil {
log.Errorf("failed to delete mangle rule: %v", err)
}
}
delete(m.rules, r.ID())
return m.rConn.Flush()
}
ips, ok := m.ipsetStore.ips(r.nftSet.Name)
if !ok {
if err := m.rConn.DelRule(r.nftRule); err != nil {
log.Errorf("failed to delete rule: %v", err)
}
if r.mangleRule != nil {
if err := m.rConn.DelRule(r.mangleRule); err != nil {
log.Errorf("failed to delete mangle rule: %v", err)
}
}
delete(m.rules, r.ID())
return m.rConn.Flush()
}
if _, ok := ips[r.ip.String()]; ok {
err := m.sConn.SetDeleteElements(r.nftSet, []nftables.SetElement{{Key: ipToBytes(r.ip, m.af)}})
if err != nil {
log.Errorf("delete elements for set %q: %v", r.nftSet.Name, err)
}
if err := m.sConn.Flush(); err != nil {
log.Debugf("flush error of set delete element, %s", r.nftSet.Name)
return err
}
m.ipsetStore.DeleteIpFromSet(r.nftSet.Name, r.ip)
}
// if after delete, set still contains other IPs,
// no need to delete firewall rule and we should exit here
if len(ips) > 0 {
return nil
}
if err := m.rConn.DelRule(r.nftRule); err != nil {
log.Errorf("failed to delete rule: %v", err)
}
if r.mangleRule != nil {
if err := m.rConn.DelRule(r.mangleRule); err != nil {
log.Errorf("failed to delete mangle rule: %v", err)
}
}
if err := m.rConn.Flush(); err != nil {
return err
}
delete(m.rules, r.ID())
m.ipsetStore.DeleteReferenceFromIpSet(r.nftSet.Name)
if m.ipsetStore.HasReferenceToSet(r.nftSet.Name) {
return nil
}
// we delete last IP from the set, that means we need to delete
// set itself and associated firewall rule too
m.rConn.FlushSet(r.nftSet)
m.rConn.DelSet(r.nftSet)
m.ipsetStore.deleteIpset(r.nftSet.Name)
return nil
}
// createDefaultAllowRules creates default allow rules for the input and output chains
func (m *AclManager) createDefaultAllowRules() error {
expIn := []expr.Any{
&expr.Verdict{
Kind: expr.VerdictAccept,
},
}
_ = m.rConn.InsertRule(&nftables.Rule{
Table: m.workTable,
Chain: m.chainInputRules,
Position: 0,
Exprs: expIn,
})
if err := m.rConn.Flush(); err != nil {
return fmt.Errorf(flushError, err)
}
return nil
}
// Flush rule/chain/set operations from the buffer
//
// Method also get all rules after flush and refreshes handle values in the rulesets
func (m *AclManager) Flush() error {
if err := m.flushWithBackoff(); err != nil {
return err
}
if err := m.refreshRuleHandles(m.chainInputRules, false); err != nil {
log.Errorf("failed to refresh rule handles ipv4 input chain: %v", err)
}
if err := m.refreshRuleHandles(m.chainPrerouting, true); err != nil {
log.Errorf("failed to refresh rule handles prerouting chain: %v", err)
}
return nil
}
func (m *AclManager) addIOFiltering(
ip net.IP,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
ipset *nftables.Set,
) (*Rule, error) {
ruleId := generatePeerRuleId(ip, proto, sPort, dPort, action, ipset)
if r, ok := m.rules[ruleId]; ok {
return &Rule{
nftRule: r.nftRule,
mangleRule: r.mangleRule,
nftSet: r.nftSet,
ruleID: r.ruleID,
ip: ip,
}, nil
}
var expressions []expr.Any
if proto != firewall.ProtocolALL {
expressions = append(expressions, &expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: m.af.protoOffset,
Len: uint32(1),
})
protoData, err := m.af.protoNum(proto)
if err != nil {
return nil, fmt.Errorf("convert protocol to number: %v", err)
}
expressions = append(expressions, &expr.Cmp{
Register: 1,
Op: expr.CmpOpEq,
Data: []byte{protoData},
})
}
rawIP := ipToBytes(ip, m.af)
// check if rawIP contains zeroed IPv4 0.0.0.0 value
// in that case not add IP match expression into the rule definition
if slices.ContainsFunc(rawIP, func(v byte) bool { return v != 0 }) {
expressions = append(expressions,
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: m.af.srcAddrOffset,
Len: m.af.addrLen,
},
)
// add individual IP for match if no ipset defined
if ipset == nil {
expressions = append(expressions,
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: rawIP,
},
)
} else {
expressions = append(expressions,
&expr.Lookup{
SourceRegister: 1,
SetName: ipset.Name,
SetID: ipset.ID,
},
)
}
}
expressions = append(expressions, applyPort(sPort, true)...)
expressions = append(expressions, applyPort(dPort, false)...)
mainExpressions := slices.Clone(expressions)
switch action {
case firewall.ActionAccept:
mainExpressions = append(mainExpressions, &expr.Verdict{Kind: expr.VerdictAccept})
case firewall.ActionDrop:
mainExpressions = append(mainExpressions, &expr.Verdict{Kind: expr.VerdictDrop})
}
userData := []byte(ruleId)
chain := m.chainInputRules
rule := &nftables.Rule{
Table: m.workTable,
Chain: chain,
Exprs: mainExpressions,
UserData: userData,
}
// Insert DROP rules at the beginning, append ACCEPT rules at the end
var nftRule *nftables.Rule
if action == firewall.ActionDrop {
nftRule = m.rConn.InsertRule(rule)
} else {
nftRule = m.rConn.AddRule(rule)
}
if err := m.rConn.Flush(); err != nil {
return nil, fmt.Errorf("flush input rule %s: %v", ruleId, err)
}
ruleStruct := &Rule{
nftRule: nftRule,
// best effort mangle rule
mangleRule: m.createPreroutingRule(expressions, userData),
nftSet: ipset,
ruleID: ruleId,
ip: ip,
}
m.rules[ruleId] = ruleStruct
if ipset != nil {
m.ipsetStore.AddReferenceToIpset(ipset.Name)
}
return ruleStruct, nil
}
func (m *AclManager) createPreroutingRule(expressions []expr.Any, userData []byte) *nftables.Rule {
if m.chainPrerouting == nil {
log.Warn("prerouting chain is not created")
return nil
}
preroutingExprs := slices.Clone(expressions)
// interface
preroutingExprs = append([]expr.Any{
&expr.Meta{
Key: expr.MetaKeyIIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(m.wgIface.Name()),
},
}, preroutingExprs...)
// local destination and mark
preroutingExprs = append(preroutingExprs,
&expr.Fib{
Register: 1,
ResultADDRTYPE: true,
FlagDADDR: true,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(unix.RTN_LOCAL),
},
&expr.Immediate{
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkRedirected),
},
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
SourceRegister: true,
},
)
nfRule := m.rConn.AddRule(&nftables.Rule{
Table: m.workTable,
Chain: m.chainPrerouting,
Exprs: preroutingExprs,
UserData: userData,
})
if err := m.rConn.Flush(); err != nil {
log.Errorf("failed to flush mangle rule %s: %v", string(userData), err)
return nil
}
return nfRule
}
func (m *AclManager) createDefaultChains() (err error) {
// chainNameInputRules
chain := m.createChain(chainNameInputRules)
err = m.rConn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chain.Name, err)
return fmt.Errorf(flushError, err)
}
m.chainInputRules = chain
// netbird-acl-input-filter
// type filter hook input priority filter; policy accept;
chain = m.createFilterChainWithHook(chainNameInputFilter, nftables.ChainHookInput)
m.addJumpRule(chain, m.chainInputRules.Name, expr.MetaKeyIIFNAME) // to netbird-acl-input-rules
m.addDropExpressions(chain, expr.MetaKeyIIFNAME)
err = m.rConn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chain.Name, err)
return err
}
// netbird-acl-forward-filter
chainFwFilter := m.createFilterChainWithHook(chainNameForwardFilter, nftables.ChainHookForward)
m.addJumpRulesToRtForward(chainFwFilter) // to netbird-rt-fwd
m.addDropExpressions(chainFwFilter, expr.MetaKeyIIFNAME)
err = m.rConn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chainNameForwardFilter, err)
return fmt.Errorf(flushError, err)
}
if err := m.allowRedirectedTraffic(chainFwFilter); err != nil {
log.Errorf("failed to allow redirected traffic: %s", err)
}
return nil
}
// Makes redirected traffic originally destined for the host itself (now subject to the forward filter)
// go through the input filter as well. This will enable e.g. Docker services to keep working by accessing the
// netbird peer IP.
func (m *AclManager) allowRedirectedTraffic(chainFwFilter *nftables.Chain) error {
// Chain is created by route manager
// TODO: move creation to a common place
m.chainPrerouting = &nftables.Chain{
Name: chainNameManglePrerouting,
Table: m.workTable,
Type: nftables.ChainTypeFilter,
Hooknum: nftables.ChainHookPrerouting,
Priority: nftables.ChainPriorityMangle,
}
m.addFwmarkToForward(chainFwFilter)
if err := m.rConn.Flush(); err != nil {
return fmt.Errorf(flushError, err)
}
return nil
}
func (m *AclManager) addFwmarkToForward(chainFwFilter *nftables.Chain) {
m.rConn.InsertRule(&nftables.Rule{
Table: m.workTable,
Chain: chainFwFilter,
Exprs: []expr.Any{
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkRedirected),
},
&expr.Verdict{
Kind: expr.VerdictAccept,
},
},
})
}
func (m *AclManager) addJumpRulesToRtForward(chainFwFilter *nftables.Chain) {
expressions := []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(m.wgIface.Name()),
},
&expr.Verdict{
Kind: expr.VerdictJump,
Chain: m.routingFwChainName,
},
}
_ = m.rConn.AddRule(&nftables.Rule{
Table: m.workTable,
Chain: chainFwFilter,
Exprs: expressions,
})
}
func (m *AclManager) createChain(name string) *nftables.Chain {
chain := &nftables.Chain{
Name: name,
Table: m.workTable,
}
chain = m.rConn.AddChain(chain)
insertReturnTrafficRule(m.rConn, m.workTable, chain)
return chain
}
func (m *AclManager) createFilterChainWithHook(name string, hookNum *nftables.ChainHook) *nftables.Chain {
polAccept := nftables.ChainPolicyAccept
chain := &nftables.Chain{
Name: name,
Table: m.workTable,
Hooknum: hookNum,
Priority: nftables.ChainPriorityFilter,
Type: nftables.ChainTypeFilter,
Policy: &polAccept,
}
return m.rConn.AddChain(chain)
}
func (m *AclManager) addDropExpressions(chain *nftables.Chain, ifaceKey expr.MetaKey) []expr.Any {
expressions := []expr.Any{
&expr.Meta{Key: ifaceKey, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(m.wgIface.Name()),
},
&expr.Verdict{Kind: expr.VerdictDrop},
}
_ = m.rConn.AddRule(&nftables.Rule{
Table: m.workTable,
Chain: chain,
Exprs: expressions,
})
return nil
}
func (m *AclManager) addJumpRule(chain *nftables.Chain, to string, ifaceKey expr.MetaKey) {
expressions := []expr.Any{
&expr.Meta{Key: ifaceKey, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(m.wgIface.Name()),
},
&expr.Verdict{
Kind: expr.VerdictJump,
Chain: to,
},
}
_ = m.rConn.AddRule(&nftables.Rule{
Table: chain.Table,
Chain: chain,
Exprs: expressions,
})
}
func (m *AclManager) addIpToSet(ipsetName string, ip net.IP) (*nftables.Set, error) {
ipset, err := m.rConn.GetSetByName(m.workTable, ipsetName)
rawIP := ipToBytes(ip, m.af)
if err != nil {
if ipset, err = m.createSet(m.workTable, ipsetName); err != nil {
return nil, fmt.Errorf("get set name: %v", err)
}
m.ipsetStore.newIpset(ipset.Name)
}
if m.ipsetStore.IsIpInSet(ipset.Name, ip) {
return ipset, nil
}
if err := m.sConn.SetAddElements(ipset, []nftables.SetElement{{Key: rawIP}}); err != nil {
return nil, fmt.Errorf("add set element for the first time: %v", err)
}
m.ipsetStore.AddIpToSet(ipset.Name, ip)
if err := m.sConn.Flush(); err != nil {
return nil, fmt.Errorf("flush add elements: %v", err)
}
return ipset, nil
}
// createSet in given table by name
func (m *AclManager) createSet(table *nftables.Table, name string) (*nftables.Set, error) {
ipset := &nftables.Set{
Name: name,
Table: table,
Dynamic: true,
KeyType: m.af.setKeyType,
}
if err := m.rConn.AddSet(ipset, nil); err != nil {
return nil, fmt.Errorf("create set: %v", err)
}
if err := m.rConn.Flush(); err != nil {
return nil, fmt.Errorf("flush created set: %v", err)
}
return ipset, nil
}
func (m *AclManager) flushWithBackoff() (err error) {
backoff := 4
backoffTime := 1000 * time.Millisecond
for i := 0; ; i++ {
err = m.rConn.Flush()
if err != nil {
log.Debugf("failed to flush nftables: %v", err)
if !strings.Contains(err.Error(), "busy") {
return
}
log.Error("failed to flush nftables, retrying...")
if i == backoff-1 {
return err
}
time.Sleep(backoffTime)
backoffTime *= 2
continue
}
break
}
return
}
func (m *AclManager) refreshRuleHandles(chain *nftables.Chain, mangle bool) error {
if m.workTable == nil || chain == nil {
return nil
}
list, err := m.rConn.GetRules(m.workTable, chain)
if err != nil {
return err
}
for _, rule := range list {
if len(rule.UserData) == 0 {
continue
}
split := bytes.Split(rule.UserData, []byte(" "))
r, ok := m.rules[string(split[0])]
if ok {
if mangle {
*r.mangleRule = *rule
} else {
*r.nftRule = *rule
}
}
}
return nil
}
func generatePeerRuleId(ip net.IP, proto firewall.Protocol, sPort *firewall.Port, dPort *firewall.Port, action firewall.Action, ipset *nftables.Set) string {
rulesetID := ":" + string(proto) + ":"
if sPort != nil {
rulesetID += sPort.String()
}
rulesetID += ":"
if dPort != nil {
rulesetID += dPort.String()
}
rulesetID += ":"
rulesetID += strconv.Itoa(int(action))
if ipset == nil {
return "ip:" + ip.String() + rulesetID
}
return "set:" + ipset.Name + rulesetID
}
func ifname(n string) []byte {
b := make([]byte, 16)
copy(b, n+"\x00")
return b
}
// ipToBytes converts net.IP to the correct byte length for the address family.
func ipToBytes(ip net.IP, af addrFamily) []byte {
if af.addrLen == 4 {
return ip.To4()
}
return ip.To16()
}

View File

@@ -0,0 +1,880 @@
//go:build !android
package nftables
import (
"bytes"
"errors"
"fmt"
"slices"
"strings"
"time"
"github.com/coreos/go-iptables/iptables"
"github.com/google/nftables"
"github.com/google/nftables/binaryutil"
"github.com/google/nftables/expr"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
"golang.org/x/sys/unix"
nberrors "github.com/netbirdio/netbird/client/errors"
"github.com/netbirdio/netbird/client/firewall/firewalld"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnet "github.com/netbirdio/netbird/client/net"
)
func (r *family) createContainers() error {
r.chains[chainNameRoutingFw] = r.conn.AddChain(&nftables.Chain{
Name: chainNameRoutingFw,
Table: r.workTable,
})
prio := *nftables.ChainPriorityNATSource - 1
r.chains[chainNameRoutingNat] = r.conn.AddChain(&nftables.Chain{
Name: chainNameRoutingNat,
Table: r.workTable,
Hooknum: nftables.ChainHookPostrouting,
Priority: &prio,
Type: nftables.ChainTypeNAT,
})
r.chains[chainNameRoutingRdr] = r.conn.AddChain(&nftables.Chain{
Name: chainNameRoutingRdr,
Table: r.workTable,
Hooknum: nftables.ChainHookPrerouting,
Priority: nftables.ChainPriorityNATDest,
Type: nftables.ChainTypeNAT,
})
r.chains[chainNameManglePostrouting] = r.conn.AddChain(&nftables.Chain{
Name: chainNameManglePostrouting,
Table: r.workTable,
Hooknum: nftables.ChainHookPostrouting,
Priority: nftables.ChainPriorityMangle,
Type: nftables.ChainTypeFilter,
})
r.chains[chainNameManglePrerouting] = r.conn.AddChain(&nftables.Chain{
Name: chainNameManglePrerouting,
Table: r.workTable,
Hooknum: nftables.ChainHookPrerouting,
Priority: nftables.ChainPriorityMangle,
Type: nftables.ChainTypeFilter,
})
r.chains[chainNameMangleForward] = r.conn.AddChain(&nftables.Chain{
Name: chainNameMangleForward,
Table: r.workTable,
Hooknum: nftables.ChainHookForward,
Priority: nftables.ChainPriorityMangle,
Type: nftables.ChainTypeFilter,
})
insertReturnTrafficRule(r.conn, r.workTable, r.chains[chainNameRoutingFw])
r.addPostroutingRules()
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("initialize tables: %v", err)
}
if err := r.addMSSClampingRules(); err != nil {
log.Errorf("failed to add MSS clamping rules: %s", err)
}
// Kernel routing opens both INPUT and FORWARD.
if err := r.openInterface(true); err != nil {
log.Errorf("failed to open interface in foreign chains: %s", err)
}
if err := firewalld.TrustInterface(r.wgIface.Name()); err != nil {
log.Warnf("failed to trust interface in firewalld: %v", err)
}
if err := r.refreshRulesMap(); err != nil {
log.Errorf("failed to refresh rules: %s", err)
}
return nil
}
// setupDataPlaneMark configures the fwmark for the data plane
func (r *family) setupDataPlaneMark() error {
if r.chains[chainNameManglePrerouting] == nil || r.chains[chainNameManglePostrouting] == nil {
return errors.New("no mangle chains found")
}
ctNew := getCtNewExprs()
preExprs := []expr.Any{
&expr.Meta{
Key: expr.MetaKeyIIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
}
preExprs = append(preExprs, ctNew...)
preExprs = append(preExprs,
&expr.Immediate{
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.DataPlaneMarkIn),
},
&expr.Ct{
Key: expr.CtKeyMARK,
Register: 1,
SourceRegister: true,
},
)
preNftRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameManglePrerouting],
Exprs: preExprs,
}
r.conn.AddRule(preNftRule)
postExprs := []expr.Any{
&expr.Meta{
Key: expr.MetaKeyOIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
}
postExprs = append(postExprs, ctNew...)
postExprs = append(postExprs,
&expr.Immediate{
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.DataPlaneMarkOut),
},
&expr.Ct{
Key: expr.CtKeyMARK,
Register: 1,
SourceRegister: true,
},
)
postNftRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameManglePostrouting],
Exprs: postExprs,
}
r.conn.AddRule(postNftRule)
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush: %w", err)
}
return nil
}
// openInterface adds passthrough accept rules for the NetBird interface to the
// kernel's filter table and external chains so they don't drop our traffic.
// includeForward also opens the FORWARD chains (kernel routing); when false only
// INPUT is opened, which is all the userspace router needs since it never
// forwards in the kernel.
func (r *family) openInterface(includeForward bool) error {
var merr *multierror.Error
if err := r.acceptFilterTableRules(includeForward); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.acceptExternalChainsRules(includeForward); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add accept rules to external chains: %w", err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) acceptFilterTableRules(includeForward bool) error {
if r.filterTable == nil {
return nil
}
fw := "iptables"
defer func() {
log.Debugf("Used %s to add accept input/forward rules", fw)
}()
// Try iptables first and fallback to nftables if iptables is not available.
// Use the correct protocol (iptables vs ip6tables) for the address family.
ipt, err := iptables.NewWithProtocol(r.iptablesProto())
if err != nil {
log.Warnf("Will use nftables to manipulate the filter table because iptables is not available: %v", err)
fw = "nftables"
return r.acceptFilterRulesNftables(r.filterTable, includeForward)
}
if err := r.acceptFilterRulesIptables(ipt, includeForward); err != nil {
log.Warnf("iptables failed (table may be incompatible), falling back to nftables: %v", err)
fw = "nftables"
return r.acceptFilterRulesNftables(r.filterTable, includeForward)
}
return nil
}
func (r *family) acceptFilterRulesIptables(ipt *iptables.IPTables, includeForward bool) error {
var merr *multierror.Error
if includeForward {
for _, rule := range r.getAcceptForwardRules() {
if err := ipt.Insert("filter", chainNameForward, 1, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add iptables forward rule: %v", err))
} else {
log.Debugf("added iptables forward rule: %v", rule)
}
}
}
inputRule := r.getAcceptInputRule()
if err := ipt.Insert("filter", chainNameInput, 1, inputRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("add iptables input rule: %v", err))
} else {
log.Debugf("added iptables input rule: %v", inputRule)
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) getAcceptForwardRules() [][]string {
intf := r.wgIface.Name()
return [][]string{
{"-i", intf, "-j", "ACCEPT"},
{"-o", intf, "-m", "conntrack", "--ctstate", "RELATED,ESTABLISHED", "-j", "ACCEPT"},
}
}
func (r *family) getAcceptInputRule() []string {
return []string{"-i", r.wgIface.Name(), "-j", "ACCEPT"}
}
// acceptFilterRulesNftables adds accept rules to the ip filter table using nftables.
// This is used when iptables is not available.
func (r *family) acceptFilterRulesNftables(table *nftables.Table, includeForward bool) error {
intf := ifname(r.wgIface.Name())
if includeForward {
forwardChain := &nftables.Chain{
Name: chainNameForward,
Table: table,
Type: nftables.ChainTypeFilter,
Hooknum: nftables.ChainHookForward,
Priority: nftables.ChainPriorityFilter,
}
r.insertForwardAcceptRules(forwardChain, intf)
}
inputChain := &nftables.Chain{
Name: chainNameInput,
Table: table,
Type: nftables.ChainTypeFilter,
Hooknum: nftables.ChainHookInput,
Priority: nftables.ChainPriorityFilter,
}
r.insertInputAcceptRule(inputChain, intf)
return r.conn.Flush()
}
// acceptExternalChainsRules adds accept rules to external chains (non-netbird, non-iptables tables).
// It dynamically finds chains at call time to handle chains that may have been created after startup.
func (r *family) acceptExternalChainsRules(includeForward bool) error {
chains := r.findExternalChains()
if len(chains) == 0 {
return nil
}
intf := ifname(r.wgIface.Name())
for _, chain := range chains {
r.applyExternalChainAccept(chain, intf, includeForward)
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush external chain rules: %w", err)
}
return nil
}
func (r *family) applyExternalChainAccept(chain *nftables.Chain, intf []byte, includeForward bool) {
if chain.Hooknum == nil {
log.Debugf("skipping external chain %s/%s: hooknum is nil", chain.Table.Name, chain.Name)
return
}
log.Debugf("adding accept rules to external %s chain: %s %s/%s",
hookName(chain.Hooknum), familyName(chain.Table.Family), chain.Table.Name, chain.Name)
switch *chain.Hooknum {
case *nftables.ChainHookForward:
if includeForward {
r.insertForwardAcceptRules(chain, intf)
}
case *nftables.ChainHookInput:
r.insertInputAcceptRule(chain, intf)
}
}
func (r *family) insertForwardAcceptRules(chain *nftables.Chain, intf []byte) {
existing, err := r.existingNetbirdRulesInChain(chain)
if err != nil {
log.Warnf("skip forward accept rules in %s/%s: %v", chain.Table.Name, chain.Name, err)
return
}
r.insertForwardIifRule(chain, intf, existing)
r.insertForwardOifEstablishedRule(chain, intf, existing)
}
func (r *family) insertForwardIifRule(chain *nftables.Chain, intf []byte, existing map[string]bool) {
if existing[userDataAcceptForwardRuleIif] {
return
}
r.conn.InsertRule(&nftables.Rule{
Table: chain.Table,
Chain: chain,
Exprs: []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{Op: expr.CmpOpEq, Register: 1, Data: intf},
&expr.Counter{},
&expr.Verdict{Kind: expr.VerdictAccept},
},
UserData: []byte(userDataAcceptForwardRuleIif),
})
}
func (r *family) insertForwardOifEstablishedRule(chain *nftables.Chain, intf []byte, existing map[string]bool) {
if existing[userDataAcceptForwardRuleOif] {
return
}
exprs := []expr.Any{
&expr.Meta{Key: expr.MetaKeyOIFNAME, Register: 1},
&expr.Cmp{Op: expr.CmpOpEq, Register: 1, Data: intf},
}
r.conn.InsertRule(&nftables.Rule{
Table: chain.Table,
Chain: chain,
Exprs: append(exprs, getEstablishedExprs(2)...),
UserData: []byte(userDataAcceptForwardRuleOif),
})
}
func (r *family) insertInputAcceptRule(chain *nftables.Chain, intf []byte) {
existing, err := r.existingNetbirdRulesInChain(chain)
if err != nil {
log.Warnf("skip input accept rule in %s/%s: %v", chain.Table.Name, chain.Name, err)
return
}
if existing[userDataAcceptInputRule] {
return
}
r.conn.InsertRule(&nftables.Rule{
Table: chain.Table,
Chain: chain,
Exprs: []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{Op: expr.CmpOpEq, Register: 1, Data: intf},
&expr.Counter{},
&expr.Verdict{Kind: expr.VerdictAccept},
},
UserData: []byte(userDataAcceptInputRule),
})
}
// existingNetbirdRulesInChain returns the set of netbird-owned UserData tags present in a chain; callers must bail on error since InsertRule is additive.
func (r *family) existingNetbirdRulesInChain(chain *nftables.Chain) (map[string]bool, error) {
rules, err := r.conn.GetRules(chain.Table, chain)
if err != nil {
return nil, fmt.Errorf("list rules: %w", err)
}
present := map[string]bool{}
for _, rule := range rules {
if !isNetbirdAcceptRuleTag(rule.UserData) {
continue
}
present[string(rule.UserData)] = true
}
return present, nil
}
func isNetbirdAcceptRuleTag(userData []byte) bool {
switch string(userData) {
case userDataAcceptForwardRuleIif,
userDataAcceptForwardRuleOif,
userDataAcceptInputRule:
return true
}
return false
}
func (r *family) removeAcceptFilterRules() error {
var merr *multierror.Error
if err := r.removeFilterTableRules(); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.removeExternalChainsRules(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove external chain rules: %w", err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) removeFilterTableRules() error {
if r.filterTable == nil {
return nil
}
ipt, err := iptables.NewWithProtocol(r.iptablesProto())
if err != nil {
log.Debugf("iptables not available, using nftables to remove filter rules: %v", err)
return r.removeAcceptRulesFromTable(r.filterTable)
}
if err := r.removeAcceptFilterRulesIptables(ipt); err != nil {
log.Debugf("iptables removal failed (table may be incompatible), falling back to nftables: %v", err)
return r.removeAcceptRulesFromTable(r.filterTable)
}
return nil
}
func (r *family) removeAcceptRulesFromTable(table *nftables.Table) error {
chains, err := r.conn.ListChainsOfTableFamily(table.Family)
if err != nil {
return fmt.Errorf("list chains: %v", err)
}
for _, chain := range chains {
if chain.Table.Name != table.Name {
continue
}
if chain.Name != chainNameForward && chain.Name != chainNameInput {
continue
}
if err := r.removeAcceptRulesFromChain(table, chain); err != nil {
return err
}
}
return r.conn.Flush()
}
func (r *family) removeAcceptRulesFromChain(table *nftables.Table, chain *nftables.Chain) error {
rules, err := r.conn.GetRules(table, chain)
if err != nil {
return fmt.Errorf("get rules from %s/%s: %v", table.Name, chain.Name, err)
}
for _, rule := range rules {
if bytes.Equal(rule.UserData, []byte(userDataAcceptForwardRuleIif)) ||
bytes.Equal(rule.UserData, []byte(userDataAcceptForwardRuleOif)) ||
bytes.Equal(rule.UserData, []byte(userDataAcceptInputRule)) {
if err := r.conn.DelRule(rule); err != nil {
return fmt.Errorf("delete rule from %s/%s: %v", table.Name, chain.Name, err)
}
}
}
return nil
}
// removeExternalChainsRules removes our accept rules from all external chains.
// This is deterministic - it scans for chains at removal time rather than relying on saved state,
// ensuring cleanup works even after a crash or if chains changed.
func (r *family) removeExternalChainsRules() error {
chains := r.findExternalChains()
if len(chains) == 0 {
return nil
}
var merr *multierror.Error
for _, chain := range chains {
if err := r.removeAcceptRulesFromChain(chain.Table, chain); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove rules from external chain %s/%s: %w", chain.Table.Name, chain.Name, err))
continue
}
if err := r.conn.Flush(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("flush external chain %s/%s: %w", chain.Table.Name, chain.Name, err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
// findExternalChains scans for chains from non-netbird tables that have FORWARD or INPUT hooks.
// This is used both at startup (to know where to add rules) and at cleanup (to ensure deterministic removal).
func (r *family) findExternalChains() []*nftables.Chain {
var chains []*nftables.Chain
families := []nftables.TableFamily{r.af.tableFamily, nftables.TableFamilyINet}
for _, family := range families {
allChains, err := r.conn.ListChainsOfTableFamily(family)
if err != nil {
log.Debugf("list chains for family %d: %v", family, err)
continue
}
for _, chain := range allChains {
if r.isExternalChain(chain) {
chains = append(chains, chain)
}
}
}
return chains
}
func (r *family) isExternalChain(chain *nftables.Chain) bool {
if r.workTable != nil && chain.Table.Name == r.workTable.Name {
return false
}
// Skip firewalld-owned chains. Firewalld creates its chains with the
// NFT_CHAIN_OWNER flag, so inserting rules into them returns EPERM.
// We delegate acceptance to firewalld by trusting the interface instead.
if chain.Table.Name == firewalldTableName {
return false
}
// Skip iptables/ip6tables-managed tables (adding nft-native rules breaks iptables-save compat)
if (chain.Table.Family == nftables.TableFamilyIPv4 || chain.Table.Family == nftables.TableFamilyIPv6) && isIptablesTable(chain.Table.Name) {
return false
}
if chain.Type != nftables.ChainTypeFilter {
return false
}
if chain.Hooknum == nil {
return false
}
return *chain.Hooknum == *nftables.ChainHookForward || *chain.Hooknum == *nftables.ChainHookInput
}
func isIptablesTable(name string) bool {
switch name {
case tableNameFilter, tableNat, tableMangle, tableRaw, tableSecurity:
return true
}
return false
}
func (r *family) removeAcceptFilterRulesIptables(ipt *iptables.IPTables) error {
var merr *multierror.Error
for _, rule := range r.getAcceptForwardRules() {
if err := ipt.DeleteIfExists("filter", chainNameForward, rule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove iptables forward rule: %v", err))
}
}
inputRule := r.getAcceptInputRule()
if err := ipt.DeleteIfExists("filter", chainNameInput, inputRule...); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove iptables input rule: %v", err))
}
return nberrors.FormatErrorOrNil(merr)
}
// Flush rule/chain/set operations from the buffer
//
// Method also get all rules after flush and refreshes handle values in the rulesets
func (r *family) Flush() error {
if err := r.flushWithBackoff(); err != nil {
return err
}
if err := r.refreshRuleHandles(r.chainInputRules, false); err != nil {
log.Errorf("failed to refresh rule handles ipv4 input chain: %v", err)
}
if err := r.refreshRuleHandles(r.chainPrerouting, true); err != nil {
log.Errorf("failed to refresh rule handles prerouting chain: %v", err)
}
return nil
}
// queuePreroutingRule builds the prerouting mangle rule that marks
// redirected traffic and queues it on the connection without flushing,
// so the caller can commit it in the same transaction as the rule it
// pairs with. Returns nil when the prerouting chain is absent, in which
// case nothing is queued.
func (r *family) queuePreroutingRule(expressions []expr.Any, userData []byte) *nftables.Rule {
if r.chainPrerouting == nil {
log.Warn("prerouting chain is not created")
return nil
}
preroutingExprs := slices.Clone(expressions)
// interface
preroutingExprs = append([]expr.Any{
&expr.Meta{
Key: expr.MetaKeyIIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
}, preroutingExprs...)
// local destination and mark
preroutingExprs = append(preroutingExprs,
&expr.Fib{
Register: 1,
ResultADDRTYPE: true,
FlagDADDR: true,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(unix.RTN_LOCAL),
},
&expr.Immediate{
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkRedirected),
},
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
SourceRegister: true,
},
)
return r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chainPrerouting,
Exprs: preroutingExprs,
UserData: userData,
})
}
func (r *family) createDefaultChains() (err error) {
// chainNameInputRules
chain := r.createChain(chainNameInputRules)
err = r.conn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chain.Name, err)
return fmt.Errorf(flushError, err)
}
r.chainInputRules = chain
// netbird-acl-input-filter
// type filter hook input priority filter; policy accept;
chain = r.createFilterChainWithHook(chainNameInputFilter, nftables.ChainHookInput)
r.addJumpRule(chain, r.chainInputRules.Name, expr.MetaKeyIIFNAME) // to netbird-acl-input-rules
r.addDropExpressions(chain, expr.MetaKeyIIFNAME)
err = r.conn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chain.Name, err)
return err
}
// netbird-acl-forward-filter
chainFwFilter := r.createFilterChainWithHook(chainNameForwardFilter, nftables.ChainHookForward)
r.addJumpRulesToRtForward(chainFwFilter) // to netbird-rt-fwd
r.addDropExpressions(chainFwFilter, expr.MetaKeyIIFNAME)
err = r.conn.Flush()
if err != nil {
log.Debugf("failed to create chain (%s): %s", chainNameForwardFilter, err)
return fmt.Errorf(flushError, err)
}
if err := r.allowRedirectedTraffic(chainFwFilter); err != nil {
log.Errorf("failed to allow redirected traffic: %s", err)
}
return nil
}
// Makes redirected traffic originally destined for the host itself (now subject to the forward filter)
// go through the input filter as well. This will enable e.g. Docker services to keep working by accessing the
// netbird peer IP.
func (r *family) allowRedirectedTraffic(chainFwFilter *nftables.Chain) error {
r.chainPrerouting = r.chains[chainNameManglePrerouting]
r.addFwmarkToForward(chainFwFilter)
if err := r.conn.Flush(); err != nil {
return fmt.Errorf(flushError, err)
}
return nil
}
func (r *family) addFwmarkToForward(chainFwFilter *nftables.Chain) {
r.conn.InsertRule(&nftables.Rule{
Table: r.workTable,
Chain: chainFwFilter,
Exprs: []expr.Any{
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkRedirected),
},
&expr.Verdict{
Kind: expr.VerdictAccept,
},
},
})
}
func (r *family) addJumpRulesToRtForward(chainFwFilter *nftables.Chain) {
expressions := []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Verdict{
Kind: expr.VerdictJump,
Chain: r.routingFwChainName,
},
}
_ = r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: chainFwFilter,
Exprs: expressions,
})
}
func (r *family) createChain(name string) *nftables.Chain {
chain := &nftables.Chain{
Name: name,
Table: r.workTable,
}
chain = r.conn.AddChain(chain)
insertReturnTrafficRule(r.conn, r.workTable, chain)
return chain
}
func (r *family) createFilterChainWithHook(name string, hookNum *nftables.ChainHook) *nftables.Chain {
polAccept := nftables.ChainPolicyAccept
chain := &nftables.Chain{
Name: name,
Table: r.workTable,
Hooknum: hookNum,
Priority: nftables.ChainPriorityFilter,
Type: nftables.ChainTypeFilter,
Policy: &polAccept,
}
return r.conn.AddChain(chain)
}
func (r *family) addDropExpressions(chain *nftables.Chain, ifaceKey expr.MetaKey) []expr.Any {
expressions := []expr.Any{
&expr.Meta{Key: ifaceKey, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Verdict{Kind: expr.VerdictDrop},
}
_ = r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: chain,
Exprs: expressions,
})
return nil
}
func (r *family) addJumpRule(chain *nftables.Chain, to string, ifaceKey expr.MetaKey) {
expressions := []expr.Any{
&expr.Meta{Key: ifaceKey, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Verdict{
Kind: expr.VerdictJump,
Chain: to,
},
}
_ = r.conn.AddRule(&nftables.Rule{
Table: chain.Table,
Chain: chain,
Exprs: expressions,
})
}
func (r *family) flushWithBackoff() (err error) {
backoff := 4
backoffTime := 1000 * time.Millisecond
for i := 0; ; i++ {
err = r.conn.Flush()
if err != nil {
log.Debugf("failed to flush nftables: %v", err)
if !strings.Contains(err.Error(), "busy") {
return
}
log.Error("failed to flush nftables, retrying...")
if i == backoff-1 {
return err
}
time.Sleep(backoffTime)
backoffTime *= 2
continue
}
break
}
return
}
func (r *family) refreshRuleHandles(chain *nftables.Chain, mangle bool) error {
if r.workTable == nil || chain == nil {
return nil
}
list, err := r.conn.GetRules(r.workTable, chain)
if err != nil {
return err
}
for _, rule := range list {
if len(rule.UserData) == 0 {
continue
}
pr, ok := r.filters[firewall.RuleID(rule.UserData)]
if !ok {
continue
}
if mangle {
if pr.mangleRule != nil {
*pr.mangleRule = *rule
}
} else {
*pr.nftRule = *rule
}
}
return nil
}

View File

@@ -0,0 +1,565 @@
//go:build !android
package nftables
import (
"fmt"
"net/netip"
"github.com/google/nftables"
"github.com/google/nftables/binaryutil"
"github.com/google/nftables/expr"
"github.com/google/nftables/xt"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
)
func (r *family) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error) {
ruleID := rule.ID()
if _, exists := r.rules[ruleID+dnatSuffix]; exists {
return rule, nil
}
protoNum, err := r.af.protoNum(rule.Protocol)
if err != nil {
return nil, fmt.Errorf("convert protocol to number: %w", err)
}
// Request forwarding once the rule is about to be installed, releasing
// it if a later step fails so the refcount tracks the real rules.
if err := r.ipFwdState.RequestForwarding(); err != nil {
return nil, err
}
if err := r.addDnatRedirect(rule, protoNum, ruleID); err != nil {
r.releaseForwarding()
return nil, err
}
if err := r.addDnatMasq(rule, protoNum, ruleID); err != nil {
r.releaseForwarding()
delete(r.rules, ruleID+dnatSuffix)
return nil, err
}
// Unlike iptables, there's no point in adding "out" rules in the forward chain here as our policy is ACCEPT.
// To overcome DROP policies in other chains, we'd have to add rules to the chains there.
// We also cannot just add "oif <iface> accept" there and filter in our own table as we don't know what is supposed to be allowed.
// TODO: find chains with drop policies and add rules there
if err := r.conn.Flush(); err != nil {
r.releaseForwarding()
return nil, fmt.Errorf("flush rules: %w", err)
}
return &rule, nil
}
func (r *family) addDnatRedirect(rule firewall.ForwardRule, protoNum uint8, ruleID firewall.RuleID) error {
dnatExprs := []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{
Op: expr.CmpOpNeq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: []byte{protoNum},
},
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseTransportHeader,
Offset: 2,
Len: 2,
},
}
portExprs, err := r.applyPort(&rule.DestinationPort, false)
if err != nil {
return fmt.Errorf("apply destination port: %w", err)
}
dnatExprs = append(dnatExprs, portExprs...)
// shifted translated port is not supported in nftables, so we hand this over to xtables
if rule.TranslatedPort.IsRange && len(rule.TranslatedPort.Values) == 2 {
if rule.TranslatedPort.Values[0] != rule.DestinationPort.Values[0] ||
rule.TranslatedPort.Values[1] != rule.DestinationPort.Values[1] {
return r.addXTablesRedirect(dnatExprs, ruleID, rule)
}
}
additionalExprs, regProtoMin, regProtoMax, err := r.handleTranslatedPort(rule)
if err != nil {
return err
}
dnatExprs = append(dnatExprs, additionalExprs...)
dnatExprs = append(dnatExprs,
&expr.NAT{
Type: expr.NATTypeDestNAT,
Family: uint32(r.af.tableFamily),
RegAddrMin: 1,
RegProtoMin: regProtoMin,
RegProtoMax: regProtoMax,
},
)
dnatRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingRdr],
Exprs: dnatExprs,
UserData: []byte(ruleID + dnatSuffix),
}
r.conn.AddRule(dnatRule)
r.rules[ruleID+dnatSuffix] = dnatRule
return nil
}
func (r *family) handleTranslatedPort(rule firewall.ForwardRule) ([]expr.Any, uint32, uint32, error) {
switch {
case rule.TranslatedPort.IsRange && len(rule.TranslatedPort.Values) == 2:
return r.handlePortRange(rule)
case len(rule.TranslatedPort.Values) == 0:
return r.handleAddressOnly(rule)
case len(rule.TranslatedPort.Values) == 1:
return r.handleSinglePort(rule)
default:
return nil, 0, 0, fmt.Errorf("invalid translated port: %v", rule.TranslatedPort)
}
}
func (r *family) handlePortRange(rule firewall.ForwardRule) ([]expr.Any, uint32, uint32, error) {
exprs := []expr.Any{
&expr.Immediate{
Register: 1,
Data: rule.TranslatedAddress.AsSlice(),
},
&expr.Immediate{
Register: 2,
Data: binaryutil.BigEndian.PutUint16(rule.TranslatedPort.Values[0]),
},
&expr.Immediate{
Register: 3,
Data: binaryutil.BigEndian.PutUint16(rule.TranslatedPort.Values[1]),
},
}
return exprs, 2, 3, nil
}
func (r *family) handleAddressOnly(rule firewall.ForwardRule) ([]expr.Any, uint32, uint32, error) {
exprs := []expr.Any{
&expr.Immediate{
Register: 1,
Data: rule.TranslatedAddress.AsSlice(),
},
}
return exprs, 0, 0, nil
}
func (r *family) handleSinglePort(rule firewall.ForwardRule) ([]expr.Any, uint32, uint32, error) {
exprs := []expr.Any{
&expr.Immediate{
Register: 1,
Data: rule.TranslatedAddress.AsSlice(),
},
&expr.Immediate{
Register: 2,
Data: binaryutil.BigEndian.PutUint16(rule.TranslatedPort.Values[0]),
},
}
return exprs, 2, 0, nil
}
func (r *family) addXTablesRedirect(dnatExprs []expr.Any, ruleID firewall.RuleID, rule firewall.ForwardRule) error {
dnatExprs = append(dnatExprs,
&expr.Counter{},
&expr.Target{
Name: "DNAT",
Rev: 2,
Info: &xt.NatRange2{
NatRange: xt.NatRange{
Flags: uint(xt.NatRangeMapIPs | xt.NatRangeProtoSpecified | xt.NatRangeProtoOffset),
MinIP: rule.TranslatedAddress.AsSlice(),
MaxIP: rule.TranslatedAddress.AsSlice(),
MinPort: rule.TranslatedPort.Values[0],
MaxPort: rule.TranslatedPort.Values[1],
},
BasePort: rule.DestinationPort.Values[0],
},
},
)
natTable := &nftables.Table{
Name: tableNat,
Family: r.af.tableFamily,
}
dnatRule := &nftables.Rule{
Table: natTable,
Chain: &nftables.Chain{
Name: chainNameNatPrerouting,
Table: natTable,
Type: nftables.ChainTypeNAT,
Hooknum: nftables.ChainHookPrerouting,
Priority: nftables.ChainPriorityNATDest,
},
Exprs: dnatExprs,
UserData: []byte(ruleID + dnatSuffix),
}
r.conn.AddRule(dnatRule)
r.rules[ruleID+dnatSuffix] = dnatRule
return nil
}
func (r *family) addDnatMasq(rule firewall.ForwardRule, protoNum uint8, ruleID firewall.RuleID) error {
portExprs, err := r.applyPort(&rule.TranslatedPort, false)
if err != nil {
return fmt.Errorf("apply translated port: %w", err)
}
masqExprs := []expr.Any{
&expr.Meta{Key: expr.MetaKeyOIFNAME, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: []byte{protoNum},
},
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: r.af.dstAddrOffset,
Len: r.af.addrLen,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: rule.TranslatedAddress.AsSlice(),
},
}
masqExprs = append(masqExprs, portExprs...)
masqExprs = append(masqExprs, &expr.Masq{})
masqRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingNat],
Exprs: masqExprs,
UserData: []byte(ruleID + snatSuffix),
}
r.conn.AddRule(masqRule)
r.rules[ruleID+snatSuffix] = masqRule
return nil
}
func (r *family) DeleteDNATRule(rule firewall.Rule) error {
ruleID := rule.ID()
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
var merr *multierror.Error
var needsFlush bool
var found bool
if dnatRule, exists := r.rules[ruleID+dnatSuffix]; exists {
found = true
if dnatRule.Handle == 0 {
log.Warnf("dnat rule %s has no handle, removing stale entry", ruleID+dnatSuffix)
delete(r.rules, ruleID+dnatSuffix)
} else if err := r.conn.DelRule(dnatRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete dnat rule: %w", err))
} else {
needsFlush = true
}
}
if masqRule, exists := r.rules[ruleID+snatSuffix]; exists {
found = true
if masqRule.Handle == 0 {
log.Warnf("snat rule %s has no handle, removing stale entry", ruleID+snatSuffix)
delete(r.rules, ruleID+snatSuffix)
} else if err := r.conn.DelRule(masqRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete snat rule: %w", err))
} else {
needsFlush = true
}
}
if needsFlush {
if err := r.conn.Flush(); err != nil {
merr = multierror.Append(merr, fmt.Errorf(flushError, err))
}
}
if merr != nil {
return nberrors.FormatErrorOrNil(merr)
}
delete(r.rules, ruleID+dnatSuffix)
delete(r.rules, ruleID+snatSuffix)
// Release once, only if the rule was present and removed.
if found {
r.releaseForwarding()
}
return nil
}
// releaseForwarding drops one IP forwarding reference, logging any error.
func (r *family) releaseForwarding() {
if err := r.ipFwdState.ReleaseForwarding(); err != nil {
log.Errorf("release IP forwarding: %v", err)
}
}
func (r *family) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("inbound-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if _, exists := r.rules[ruleID]; exists {
return nil
}
protoNum, err := r.af.protoNum(protocol)
if err != nil {
return fmt.Errorf("convert protocol to number: %w", err)
}
exprs := []expr.Any{
&expr.Meta{Key: expr.MetaKeyIIFNAME, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 2},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 2,
Data: []byte{protoNum},
},
&expr.Payload{
DestRegister: 3,
Base: expr.PayloadBaseTransportHeader,
Offset: 2,
Len: 2,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 3,
Data: binaryutil.BigEndian.PutUint16(originalPort),
},
}
bits := 32
if localAddr.Is6() {
bits = 128
}
exprs = append(exprs, prefixMatchExprs(r.af, netip.PrefixFrom(localAddr, bits), false)...)
exprs = append(exprs,
&expr.Immediate{
Register: 1,
Data: localAddr.AsSlice(),
},
&expr.Immediate{
Register: 2,
Data: binaryutil.BigEndian.PutUint16(translatedPort),
},
&expr.NAT{
Type: expr.NATTypeDestNAT,
Family: uint32(r.af.tableFamily),
RegAddrMin: 1,
RegProtoMin: 2,
RegProtoMax: 0,
},
)
dnatRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingRdr],
Exprs: exprs,
UserData: []byte(ruleID),
}
r.conn.AddRule(dnatRule)
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("add inbound DNAT rule: %w", err)
}
r.rules[ruleID] = dnatRule
return nil
}
// RemoveInboundDNAT removes an inbound DNAT rule.
func (r *family) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
ruleID := firewall.RuleID(fmt.Sprintf("inbound-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
rule, exists := r.rules[ruleID]
if !exists {
return nil
}
if rule.Handle == 0 {
log.Warnf("inbound DNAT rule %s has no handle, removing stale entry", ruleID)
delete(r.rules, ruleID)
return nil
}
if err := r.conn.DelRule(rule); err != nil {
return fmt.Errorf("delete inbound DNAT rule %s: %w", ruleID, err)
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush delete inbound DNAT rule: %w", err)
}
delete(r.rules, ruleID)
return nil
}
// ensureNATOutputChain lazily creates the OUTPUT NAT chain on first use.
func (r *family) ensureNATOutputChain() error {
if _, exists := r.chains[chainNameNATOutput]; exists {
return nil
}
r.chains[chainNameNATOutput] = r.conn.AddChain(&nftables.Chain{
Name: chainNameNATOutput,
Table: r.workTable,
Hooknum: nftables.ChainHookOutput,
Priority: nftables.ChainPriorityNATDest,
Type: nftables.ChainTypeNAT,
})
if err := r.conn.Flush(); err != nil {
delete(r.chains, chainNameNATOutput)
return fmt.Errorf("create NAT output chain: %w", err)
}
return nil
}
// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
func (r *family) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
ruleID := firewall.RuleID(fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
if _, exists := r.rules[ruleID]; exists {
return nil
}
if err := r.ensureNATOutputChain(); err != nil {
return err
}
protoNum, err := r.af.protoNum(protocol)
if err != nil {
return fmt.Errorf("convert protocol to number: %w", err)
}
exprs := []expr.Any{
&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 1},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: []byte{protoNum},
},
&expr.Payload{
DestRegister: 2,
Base: expr.PayloadBaseTransportHeader,
Offset: 2,
Len: 2,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 2,
Data: binaryutil.BigEndian.PutUint16(originalPort),
},
}
bits := 32
if localAddr.Is6() {
bits = 128
}
exprs = append(exprs, prefixMatchExprs(r.af, netip.PrefixFrom(localAddr, bits), false)...)
exprs = append(exprs,
&expr.Immediate{
Register: 1,
Data: localAddr.AsSlice(),
},
&expr.Immediate{
Register: 2,
Data: binaryutil.BigEndian.PutUint16(translatedPort),
},
&expr.NAT{
Type: expr.NATTypeDestNAT,
Family: uint32(r.af.tableFamily),
RegAddrMin: 1,
RegProtoMin: 2,
},
)
dnatRule := &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameNATOutput],
Exprs: exprs,
UserData: []byte(ruleID),
}
r.conn.AddRule(dnatRule)
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("add output DNAT rule: %w", err)
}
r.rules[ruleID] = dnatRule
return nil
}
// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
func (r *family) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, originalPort, translatedPort uint16) error {
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
ruleID := firewall.RuleID(fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, originalPort, translatedPort))
rule, exists := r.rules[ruleID]
if !exists {
return nil
}
if rule.Handle == 0 {
log.Warnf("output DNAT rule %s has no handle, removing stale entry", ruleID)
delete(r.rules, ruleID)
return nil
}
if err := r.conn.DelRule(rule); err != nil {
return fmt.Errorf("delete output DNAT rule %s: %w", ruleID, err)
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush delete output DNAT rule: %w", err)
}
delete(r.rules, ruleID)
return nil
}

View File

@@ -0,0 +1,249 @@
//go:build !android
package nftables
import (
"fmt"
"net/netip"
"github.com/coreos/go-iptables/iptables"
"github.com/google/nftables"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
"github.com/netbirdio/netbird/client/firewall/firewalld"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
"github.com/netbirdio/netbird/client/internal/routemanager/ipfwdstate"
"github.com/netbirdio/netbird/client/internal/routemanager/refcounter"
)
const (
tableNat = "nat"
tableMangle = "mangle"
tableRaw = "raw"
tableSecurity = "security"
chainNameNatPrerouting = "PREROUTING"
chainNameRoutingFw = "netbird-rt-fwd"
chainNameRoutingNat = "netbird-rt-postrouting"
chainNameRoutingRdr = "netbird-rt-redirect"
chainNameNATOutput = "netbird-nat-output"
chainNameForward = "FORWARD"
chainNameMangleForward = "netbird-mangle-forward"
// Peer ACL chain names.
chainNameInputRules = "netbird-acl-input-rules"
chainNameInputFilter = "netbird-acl-input-filter"
chainNameForwardFilter = "netbird-acl-forward-filter"
chainNameManglePrerouting = "netbird-mangle-prerouting"
chainNameManglePostrouting = "netbird-mangle-postrouting"
flushError = "flush: %w"
firewalldTableName = "firewalld"
userDataAcceptForwardRuleIif = "frwacceptiif"
userDataAcceptForwardRuleOif = "frwacceptoif"
userDataAcceptInputRule = "inputaccept"
dnatSuffix firewall.RuleID = "_dnat"
snatSuffix firewall.RuleID = "_snat"
// ipv4TCPHeaderSize is the minimum IPv4 (20) + TCP (20) header size for MSS calculation.
ipv4TCPHeaderSize = 40
// ipv6TCPHeaderSize is the minimum IPv6 (40) + TCP (20) header size for MSS calculation.
ipv6TCPHeaderSize = 60
// maxPrefixesSet 1638 prefixes start to fail, taking some margin
maxPrefixesSet = 1500
refreshRulesMapError = "refresh rules map: %w"
)
var (
errFilterTableNotFound = fmt.Errorf("'filter' table not found")
)
type setInput struct {
set firewall.Set
prefixes []netip.Prefix
}
// family holds the per-address-family nftables state. One instance
// handles route ACLs, peer ACLs, NAT, DNAT, and MSS clamping for a
// single family; the top-level Manager owns one for v4 and another
// for v6. The name predates the peer-ACL absorption; it's effectively
// the per-family backend now.
type family struct {
conn *nftables.Conn
workTable *nftables.Table
filterTable *nftables.Table
chains map[string]*nftables.Chain
// filters holds peer + route filter rules keyed by content hash.
// AddFilterRule writes here; DeleteFilterRule looks up by id.
filters map[firewall.RuleID]*Rule
// rules holds NAT, DNAT, and external accept rules (auxiliary
// plumbing that isn't a filter rule).
rules map[firewall.RuleID]*nftables.Rule
// Peer ACL chain handles.
chainInputRules *nftables.Chain
chainPrerouting *nftables.Chain
routingFwChainName string
ipsetCounter *refcounter.Counter[string, setInput, *nftables.Set]
af addrFamily
wgIface iFaceMapper
ipFwdState *ipfwdstate.IPForwardingState
legacyManagement bool
mtu uint16
}
func newFamily(workTable *nftables.Table, wgIface iFaceMapper, mtu uint16) *family {
r := &family{
conn: &nftables.Conn{},
workTable: workTable,
chains: make(map[string]*nftables.Chain),
filters: make(map[firewall.RuleID]*Rule),
rules: make(map[firewall.RuleID]*nftables.Rule),
routingFwChainName: chainNameRoutingFw,
af: familyForAddr(workTable.Family == nftables.TableFamilyIPv4),
wgIface: wgIface,
ipFwdState: ipfwdstate.NewIPForwardingState(),
mtu: mtu,
}
r.ipsetCounter = refcounter.New(
r.createIpSet,
r.deleteIpSet,
)
var err error
r.filterTable, err = r.loadFilterTable()
if err != nil {
log.Debugf("ip filter table not found: %v", err)
}
return r
}
func (r *family) init(workTable *nftables.Table) error {
r.workTable = workTable
if err := r.removeAcceptFilterRules(); err != nil {
log.Errorf("failed to clean up rules from filter table: %s", err)
}
if err := r.createContainers(); err != nil {
return fmt.Errorf("create containers: %w", err)
}
if err := r.setupDataPlaneMark(); err != nil {
log.Errorf("failed to set up data plane mark: %v", err)
}
if err := r.createDefaultChains(); err != nil {
return fmt.Errorf("create default acl chains: %w", err)
}
return nil
}
// Reset cleans existing nftables filter table rules from the system
func (r *family) Reset() error {
// clear without deleting the ipsets, the nf table will be deleted by the caller
r.ipsetCounter.Clear()
var merr *multierror.Error
if err := r.removeAcceptFilterRules(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove accept filter rules: %w", err))
}
if err := firewalld.UntrustInterface(r.wgIface.Name()); err != nil {
merr = multierror.Append(merr, err)
}
if err := r.removeNatPreroutingRules(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove filter prerouting rules: %w", err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) loadFilterTable() (*nftables.Table, error) {
tables, err := r.conn.ListTablesOfFamily(r.af.tableFamily)
if err != nil {
return nil, fmt.Errorf("list tables: %w", err)
}
for _, table := range tables {
if table.Name == "filter" {
return table, nil
}
}
return nil, errFilterTableNotFound
}
func hookName(hook *nftables.ChainHook) string {
if hook == nil {
return "unknown"
}
switch *hook {
case *nftables.ChainHookForward:
return chainNameForward
case *nftables.ChainHookInput:
return chainNameInput
default:
return fmt.Sprintf("hook(%d)", *hook)
}
}
func familyName(family nftables.TableFamily) string {
switch family {
case nftables.TableFamilyIPv4:
return "ip"
case nftables.TableFamilyIPv6:
return "ip6"
case nftables.TableFamilyINet:
return "inet"
default:
return fmt.Sprintf("family(%d)", family)
}
}
func (r *family) iptablesProto() iptables.Protocol {
if r.af.tableFamily == nftables.TableFamilyIPv6 {
return iptables.ProtocolIPv6
}
return iptables.ProtocolIPv4
}
func (r *family) refreshRulesMap() error {
var merr *multierror.Error
newRules := make(map[firewall.RuleID]*nftables.Rule)
for _, chain := range r.chains {
rules, err := r.conn.GetRules(chain.Table, chain)
if err != nil {
merr = multierror.Append(merr, fmt.Errorf("list rules for chain %s: %w", chain.Name, err))
// preserve existing entries for this chain since we can't verify their state
for k, v := range r.rules {
if v.Chain != nil && v.Chain.Name == chain.Name {
newRules[k] = v
}
}
continue
}
for _, rule := range rules {
if len(rule.UserData) > 0 {
newRules[firewall.RuleID(rule.UserData)] = rule
}
}
}
r.rules = newRules
return nberrors.FormatErrorOrNil(merr)
}

View File

@@ -0,0 +1,512 @@
//go:build !android
package nftables
import (
"fmt"
"net"
"net/netip"
"slices"
"github.com/google/nftables"
"github.com/google/nftables/binaryutil"
"github.com/google/nftables/expr"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbid "github.com/netbirdio/netbird/client/internal/acl/id"
)
// AddFilterRule installs one nftables packet-filter rule. With
// destination empty the rule goes to the peer ACL input chain plus a
// paired prerouting mangle rule for the redirect mark. With
// destination set (prefix or named set) it goes to the route ACL
// forward chain. Multi-source rules collapse to one nftables rule
// backed by the shared refcounted hash:net set.
func (r *family) AddFilterRule(
id []byte,
sources []netip.Prefix,
destination firewall.Network,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
) (firewall.Rule, error) {
isRoute := !destination.IsZero()
ruleID := nbid.GenerateRuleID(sources, destination, proto, sPort, dPort, action)
if existing, ok := r.filters[ruleID]; ok {
return existing, nil
}
srcExprs, err := r.applyNetwork(sourceNetwork(sources), sources, true)
if err != nil {
return nil, fmt.Errorf("apply source: %w", err)
}
var exprs []expr.Any
if isRoute {
exprs, err = r.buildRouteFilterExprs(srcExprs, destination, proto, sPort, dPort)
} else {
exprs, err = r.buildPeerFilterExprs(srcExprs, proto, sPort, dPort)
}
if err != nil {
r.dropNetworkMatch(srcExprs)
return nil, err
}
mainExprs := slices.Clone(exprs)
verdict := expr.VerdictAccept
if action == firewall.ActionDrop {
verdict = expr.VerdictDrop
}
mainExprs = append(mainExprs, &expr.Verdict{Kind: verdict})
chain := r.chainInputRules
if isRoute {
chain = r.chains[chainNameRoutingFw]
}
userData := []byte(ruleID)
// Build the paired prerouting mangle rule before flushing so both
// rules commit in one transaction. An anonymous port set binds to
// exactly one rule, so the mangle rule needs its own expression list
// with fresh sets, not a clone of the main rule's. Guard on the
// prerouting chain first: building the expressions queues the port
// set, so skipping the build when there is no chain to bind it to
// keeps an unbound set out of the connection batch.
var mangleRule *nftables.Rule
if !isRoute && r.chainPrerouting != nil {
mangleExprs, err := r.buildPeerFilterExprs(srcExprs, proto, sPort, dPort)
if err != nil {
r.dropNetworkMatch(exprs)
return nil, fmt.Errorf("build mangle rule: %w", err)
}
mangleRule = r.queuePreroutingRule(mangleExprs, userData)
}
nftRule := &nftables.Rule{
Table: r.workTable,
Chain: chain,
Exprs: mainExprs,
UserData: userData,
}
if action == firewall.ActionDrop {
nftRule = r.conn.InsertRule(nftRule)
} else {
nftRule = r.conn.AddRule(nftRule)
}
if err := r.conn.Flush(); err != nil {
r.dropNetworkMatch(exprs)
return nil, fmt.Errorf(flushError, err)
}
rule := &Rule{
nftRule: nftRule,
mangleRule: mangleRule,
sources: sources,
id: ruleID,
}
r.filters[ruleID] = rule
log.Debugf("added filter rule: sources=%v, destination=%v, proto=%v, sPort=%v, dPort=%v, action=%v",
sources, destination, proto, sPort, dPort, action)
return rule, nil
}
// buildPeerFilterExprs assembles the input-chain (peer ACL) match: the
// IP-header protocol byte read via Payload, then source, then ports
// (no counter), matching the historical peer shape so per-rule kernel
// state is identical to pre-unification.
func (r *family) buildPeerFilterExprs(
srcExprs []expr.Any,
proto firewall.Protocol,
sPort, dPort *firewall.Port,
) ([]expr.Any, error) {
var exprs []expr.Any
if proto != firewall.ProtocolALL {
protoNum, err := r.af.protoNum(proto)
if err != nil {
return nil, fmt.Errorf("convert protocol to number: %w", err)
}
exprs = append(exprs,
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: r.af.protoOffset,
Len: 1,
},
&expr.Cmp{Op: expr.CmpOpEq, Register: 1, Data: []byte{protoNum}},
)
}
exprs = append(exprs, srcExprs...)
portExprs, err := r.applyPorts(sPort, dPort)
if err != nil {
return nil, err
}
exprs = append(exprs, portExprs...)
return exprs, nil
}
// buildRouteFilterExprs assembles the forward-chain (route ACL) match:
// source, then destination, then optional proto/ports, then a counter.
func (r *family) buildRouteFilterExprs(
srcExprs []expr.Any,
destination firewall.Network,
proto firewall.Protocol,
sPort, dPort *firewall.Port,
) ([]expr.Any, error) {
exprs := append([]expr.Any{}, srcExprs...)
destExprs, err := r.applyNetwork(destination, nil, false)
if err != nil {
return nil, fmt.Errorf("apply destination: %w", err)
}
exprs = append(exprs, destExprs...)
if proto != firewall.ProtocolALL {
protoNum, err := r.af.protoNum(proto)
if err != nil {
r.dropNetworkMatch(destExprs)
return nil, fmt.Errorf("convert protocol to number: %w", err)
}
exprs = append(exprs,
&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 1},
&expr.Cmp{Op: expr.CmpOpEq, Register: 1, Data: []byte{protoNum}},
)
portExprs, err := r.applyPorts(sPort, dPort)
if err != nil {
r.dropNetworkMatch(destExprs)
return nil, err
}
exprs = append(exprs, portExprs...)
}
exprs = append(exprs, &expr.Counter{})
return exprs, nil
}
func (r *family) hasRule(id firewall.RuleID) bool {
_, ok := r.filters[id]
return ok
}
func (r *family) hasDNATRule(id firewall.RuleID) bool {
_, ok := r.rules[id+dnatSuffix]
return ok
}
// DeleteFilterRule removes a previously installed filter rule. Source
// set references are recovered from the stored rule's expressions via
// findSets and dropped from the shared refcounter.
func (r *family) DeleteFilterRule(rule firewall.Rule) error {
ruleID := rule.ID()
pr, ok := r.filters[ruleID]
if !ok {
log.Debugf("filter rule %s not found", ruleID)
return nil
}
// A freshly added rule carries no handle until it is read back from
// the kernel, and Flush only refreshes the peer chains. Pull live
// handles for this rule's chain before deciding it is stale so route
// rules (which Flush never refreshes) can actually be deleted. A
// refresh failure aborts the delete without touching tracking state,
// so the caller can retry while the rule may still exist in the kernel.
if pr.nftRule.Handle == 0 {
if err := r.refreshRuleHandles(pr.nftRule.Chain, false); err != nil {
return fmt.Errorf("refresh handles for chain %s: %w", pr.nftRule.Chain.Name, err)
}
}
// Refresh the mangle handle independently: the main rule's handle can
// be populated while the prerouting refresh during Flush failed, and
// gating the mangle refresh on the main handle would leak the mangle
// rule on delete.
if pr.mangleRule != nil && pr.mangleRule.Handle == 0 {
if err := r.refreshRuleHandles(r.chainPrerouting, true); err != nil {
return fmt.Errorf("refresh mangle handles: %w", err)
}
}
if pr.nftRule.Handle == 0 {
log.Warnf("filter rule %s has no handle, removing stale entry", ruleID)
r.dropNetworkMatch(pr.nftRule.Exprs)
delete(r.filters, ruleID)
return nil
}
if err := r.conn.DelRule(pr.nftRule); err != nil {
log.Errorf("queue rule delete: %v", err)
}
if pr.mangleRule != nil {
if err := r.conn.DelRule(pr.mangleRule); err != nil {
log.Errorf("queue mangle rule delete: %v", err)
}
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush delete %s: %w", ruleID, err)
}
r.dropNetworkMatch(pr.nftRule.Exprs)
delete(r.filters, ruleID)
return nil
}
func (r *family) decrementSetCounter(rule *nftables.Rule) error {
if r.ipsetCounter == nil {
return nil
}
sets := findSets(rule)
var merr *multierror.Error
for _, setName := range sets {
if _, err := r.ipsetCounter.Decrement(setName); err != nil {
merr = multierror.Append(merr, fmt.Errorf("decrement set counter: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
// dropNetworkMatch undoes whatever the source/destination match
// reserved. Safe to call when the spec is empty or holds only inline
// matchers.
func (r *family) dropNetworkMatch(exprs []expr.Any) {
if r.ipsetCounter == nil {
return
}
for _, e := range exprs {
lookup, ok := e.(*expr.Lookup)
if !ok {
continue
}
if _, err := r.ipsetCounter.Decrement(lookup.SetName); err != nil {
log.Errorf("rollback ipset decrement %s: %v", lookup.SetName, err)
}
}
}
func (r *family) applyNetwork(
network firewall.Network,
setPrefixes []netip.Prefix,
isSource bool,
) ([]expr.Any, error) {
if network.IsSet() {
exprs, err := r.getIpSet(network.Set, setPrefixes, isSource)
if err != nil {
side := "destination"
if isSource {
side = "source"
}
return nil, fmt.Errorf("%s set: %w", side, err)
}
return exprs, nil
}
if network.IsPrefix() {
return prefixMatchExprs(r.af, network.Prefix, isSource), nil
}
return nil, nil
}
// applyPort builds the transport-header port match. A single value
// compares directly, a range uses a range expression, and multiple
// values go through an anonymous constant set: consecutive cmp
// expressions AND together, so chained equality comparisons could
// never match more than one port. The set is queued on the
// connection and committed by the caller's flush together with the
// rule that binds it.
func (r *family) applyPort(port *firewall.Port, isSource bool) ([]expr.Any, error) {
if port == nil || len(port.Values) == 0 {
return nil, nil
}
// dst port
offset := uint32(2)
if isSource {
// src port
offset = 0
}
exprs := []expr.Any{
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseTransportHeader,
Offset: offset,
Len: 2,
},
}
switch {
case port.IsRange && len(port.Values) == 2:
exprs = append(exprs, &expr.Range{
Op: expr.CmpOpEq,
Register: 1,
FromData: binaryutil.BigEndian.PutUint16(port.Values[0]),
ToData: binaryutil.BigEndian.PutUint16(port.Values[1]),
})
case len(port.Values) == 1:
exprs = append(exprs, &expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.BigEndian.PutUint16(port.Values[0]),
})
default:
lookup, err := r.anonymousPortSet(port.Values)
if err != nil {
return nil, err
}
exprs = append(exprs, lookup)
}
return exprs, nil
}
// anonymousPortSet queues an anonymous constant set holding the given
// ports on the connection and returns a lookup against it. The set is
// committed by the caller's flush together with the rule that binds it.
func (r *family) anonymousPortSet(values []uint16) (*expr.Lookup, error) {
set := &nftables.Set{
Anonymous: true,
Constant: true,
Table: r.workTable,
KeyType: nftables.TypeInetService,
}
elements := make([]nftables.SetElement, 0, len(values))
for _, p := range values {
elements = append(elements, nftables.SetElement{Key: binaryutil.BigEndian.PutUint16(p)})
}
if err := r.conn.AddSet(set, elements); err != nil {
return nil, fmt.Errorf("add anonymous port set: %w", err)
}
return &expr.Lookup{
SourceRegister: 1,
SetID: set.ID,
SetName: set.Name,
}, nil
}
// applyPorts builds the source then destination port matches.
func (r *family) applyPorts(sPort, dPort *firewall.Port) ([]expr.Any, error) {
sPortExprs, err := r.applyPort(sPort, true)
if err != nil {
return nil, fmt.Errorf("apply source port: %w", err)
}
dPortExprs, err := r.applyPort(dPort, false)
if err != nil {
return nil, fmt.Errorf("apply destination port: %w", err)
}
return append(sPortExprs, dPortExprs...), nil
}
// prefixMatchExprs is the family-aware match sequence for a CIDR
// prefix. /0 returns nil; a host prefix (full bit length for the
// family) skips the bitwise step since the mask is all-ones. Shared
// between family and aclManager so both treat single prefixes
// identically.
func prefixMatchExprs(af addrFamily, prefix netip.Prefix, isSource bool) []expr.Any {
offset := af.dstAddrOffset
if isSource {
offset = af.srcAddrOffset
}
ones := prefix.Bits()
if ones == 0 {
return nil
}
payload := &expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: offset,
Len: af.addrLen,
}
cmp := &expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: prefix.Masked().Addr().AsSlice(),
}
if ones == af.totalBits {
return []expr.Any{payload, cmp}
}
mask := net.CIDRMask(ones, af.totalBits)
xor := make([]byte, af.addrLen)
return []expr.Any{
payload,
&expr.Bitwise{
DestRegister: 1,
SourceRegister: 1,
Len: af.addrLen,
Mask: mask,
Xor: xor,
},
cmp,
}
}
func getCtNewExprs() []expr.Any {
return []expr.Any{
&expr.Ct{
Key: expr.CtKeySTATE,
Register: 1,
},
&expr.Bitwise{
SourceRegister: 1,
DestRegister: 1,
Len: 4,
Mask: binaryutil.NativeEndian.PutUint32(expr.CtStateBitNEW),
Xor: binaryutil.NativeEndian.PutUint32(0),
},
&expr.Cmp{
Op: expr.CmpOpNeq,
Register: 1,
Data: []byte{0, 0, 0, 0},
},
}
}
// sourceNetwork classifies a source-prefix list into the firewall.Network
// shape the rest of the spec-builder consumes: empty for match-any, a
// single prefix inline, or an ipset for multiple sources.
func sourceNetwork(sources []netip.Prefix) firewall.Network {
switch {
case len(sources) == 0:
return firewall.Network{}
case len(sources) == 1 && sources[0].Bits() == 0:
return firewall.Network{}
case len(sources) == 1:
return firewall.Network{Prefix: sources[0]}
default:
return firewall.Network{Set: firewall.NewPrefixSet(sources)}
}
}
func ifname(n string) []byte {
b := make([]byte, 16)
copy(b, n+"\x00")
return b
}
// findSets scans an nftables rule's expressions for expr.Lookup and
// returns the named sets in occurrence order. Used at delete time to
// drop ipsetCounter references; peer and route ACLs go through it.
func findSets(rule *nftables.Rule) []string {
var sets []string
for _, e := range rule.Exprs {
if lookup, ok := e.(*expr.Lookup); ok {
sets = append(sets, lookup.SetName)
}
}
return sets
}

View File

@@ -0,0 +1,90 @@
//go:build privileged
package nftables
import (
"bytes"
"os"
"testing"
"github.com/google/nftables"
"github.com/stretchr/testify/require"
"github.com/netbirdio/netbird/client/iface"
)
// TestInterfaceAllowerInputOnly verifies the userspace-mode allower opens the
// interface on the INPUT hook of foreign chains only (not FORWARD, since the
// userspace router never forwards in the kernel), creates no netbird work
// table, and removes its rules on Close.
func TestInterfaceAllowerInputOnly(t *testing.T) {
if os.Geteuid() != 0 {
t.Skip("root required")
}
require.False(t, ipTableExists(t, getTableName()), "precondition: no stale netbird table")
conn := &nftables.Conn{}
extTable := conn.AddTable(&nftables.Table{Name: "nbtest_extchains", Family: nftables.TableFamilyINet})
inputChain := conn.AddChain(&nftables.Chain{
Name: "ext_input", Table: extTable,
Hooknum: nftables.ChainHookInput, Priority: nftables.ChainPriorityFilter, Type: nftables.ChainTypeFilter,
})
forwardChain := conn.AddChain(&nftables.Chain{
Name: "ext_forward", Table: extTable,
Hooknum: nftables.ChainHookForward, Priority: nftables.ChainPriorityFilter, Type: nftables.ChainTypeFilter,
})
require.NoError(t, conn.Flush(), "create external table and chains")
t.Cleanup(func() {
c := &nftables.Conn{}
c.DelTable(extTable)
_ = c.Flush()
})
allower, err := NewInterfaceAllower(ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "create allower")
require.NoError(t, allower.Apply(), "apply")
require.True(t, chainHasUserData(t, extTable, inputChain, userDataAcceptInputRule),
"external INPUT chain should get the accept rule")
require.Len(t, listRules(t, extTable, forwardChain), 0,
"external FORWARD chain must not be opened in userspace mode")
require.False(t, ipTableExists(t, getTableName()),
"allower must not create a netbird work table")
require.NoError(t, allower.Close(), "close")
require.False(t, chainHasUserData(t, extTable, inputChain, userDataAcceptInputRule),
"accept rule should be removed on close")
}
func listRules(t *testing.T, table *nftables.Table, chain *nftables.Chain) []*nftables.Rule {
t.Helper()
c := &nftables.Conn{}
rules, err := c.GetRules(table, chain)
require.NoError(t, err)
return rules
}
func chainHasUserData(t *testing.T, table *nftables.Table, chain *nftables.Chain, ud string) bool {
for _, r := range listRules(t, table, chain) {
if bytes.Equal(r.UserData, []byte(ud)) {
return true
}
}
return false
}
func ipTableExists(t *testing.T, name string) bool {
t.Helper()
c := &nftables.Conn{}
for _, fam := range []nftables.TableFamily{nftables.TableFamilyIPv4, nftables.TableFamilyIPv6} {
tbls, err := c.ListTablesOfFamily(fam)
require.NoError(t, err)
for _, tb := range tbls {
if tb.Name == name {
return true
}
}
}
return false
}

View File

@@ -0,0 +1,107 @@
package nftables
import (
"fmt"
"github.com/google/nftables"
"github.com/hashicorp/go-multierror"
nberrors "github.com/netbirdio/netbird/client/errors"
)
// InterfaceAllower opens the NetBird interface in the kernel's filter table and
// external chains and keeps them reconciled via a netlink monitor, so the host
// firewall doesn't drop traffic the NetBird firewall handles. It is used by the
// userspace firewall, where routing happens in the forwarder, so only INPUT is
// opened (the userspace router never forwards in the kernel).
//
// It owns its own families/connection and never creates a netbird work table.
// firewalld trust is handled by the caller, not here. Its operations are serial
// (Apply before the monitor starts; reconciles run on the single monitor
// goroutine; Close stops the monitor before removing), so it needs no locking.
//
// TODO: this opens nftables and the iptables-nft filter table (detected via
// nft), but not a legacy-iptables ruleset running in parallel with nftables.
// Such a host would keep its legacy filter chains closed for the interface.
type InterfaceAllower struct {
family4 *family
family6 *family
extMonitor *externalChainMonitor
}
// NewInterfaceAllower builds an allower for the given interface. It returns an
// error when nftables is unavailable (e.g. an iptables-legacy host), so the
// caller can fall back to firewalld trust.
func NewInterfaceAllower(wgIface iFaceMapper, mtu uint16) (*InterfaceAllower, error) {
tableName := getTableName()
family4 := newFamily(&nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv4}, wgIface, mtu)
// Probe nftables availability before committing to this backend.
if _, err := family4.conn.ListChainsOfTableFamily(nftables.TableFamilyINet); err != nil {
return nil, fmt.Errorf("nftables not available: %w", err)
}
a := &InterfaceAllower{family4: family4}
if wgIface.Address().HasIPv6() {
a.family6 = newFamily(&nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv6}, wgIface, mtu)
}
a.extMonitor = newExternalChainMonitor(a)
return a, nil
}
// Apply opens the interface (INPUT only) in the foreign filter chains and starts
// reconciling them on nftables changes.
func (a *InterfaceAllower) Apply() error {
var merr *multierror.Error
for _, f := range a.families() {
// Remove any stale accepts first so a prior unclean exit (e.g. SIGKILL,
// where Close never ran) is recovered deterministically rather than
// accumulating duplicate rules on the iptables filter table.
if err := f.removeAcceptFilterRules(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("clean stale accept rules: %w", err))
}
if err := f.openInterface(false); err != nil {
merr = multierror.Append(merr, err)
}
}
a.extMonitor.start()
return nberrors.FormatErrorOrNil(merr)
}
// families returns the configured address families (v4, and v6 when present).
func (a *InterfaceAllower) families() []*family {
families := []*family{a.family4}
if a.family6 != nil {
families = append(families, a.family6)
}
return families
}
// reconcileExternalChains re-applies the INPUT accepts to external chains. It
// implements externalChainReconciler for the monitor.
func (a *InterfaceAllower) reconcileExternalChains() error {
var merr *multierror.Error
for _, f := range a.families() {
if err := f.acceptExternalChainsRules(false); err != nil {
merr = multierror.Append(merr, err)
}
}
return nberrors.FormatErrorOrNil(merr)
}
// Close stops the monitor and removes the accept rules.
func (a *InterfaceAllower) Close() error {
a.extMonitor.stop()
var merr *multierror.Error
for _, f := range a.families() {
if err := f.removeAcceptFilterRules(); err != nil {
merr = multierror.Append(merr, err)
}
}
return nberrors.FormatErrorOrNil(merr)
}

View File

@@ -0,0 +1,210 @@
//go:build !android
package nftables
import (
"encoding/binary"
"fmt"
"net/netip"
"github.com/google/nftables"
"github.com/google/nftables/expr"
log "github.com/sirupsen/logrus"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
"github.com/netbirdio/netbird/client/internal/routemanager/refcounter"
)
func (r *family) getIpSet(set firewall.Set, prefixes []netip.Prefix, isSource bool) ([]expr.Any, error) {
ref, err := r.ipsetCounter.Increment(set.HashedName(), setInput{
set: set,
prefixes: prefixes,
})
if err != nil {
return nil, fmt.Errorf("create or get ipset: %w", err)
}
return r.getIpSetExprs(ref, isSource)
}
func (r *family) createIpSet(setName string, input setInput) (*nftables.Set, error) {
// overlapping prefixes will result in an error, so we need to merge them
prefixes := firewall.MergeIPRanges(input.prefixes)
nfset := &nftables.Set{
Name: setName,
Comment: input.set.Comment(),
Table: r.workTable,
// required for prefixes
Interval: true,
KeyType: r.af.setKeyType,
}
elements := r.convertPrefixesToSet(prefixes)
nElements := len(elements)
maxElements := maxPrefixesSet * 2
initialElements := elements[:min(maxElements, nElements)]
if err := r.conn.AddSet(nfset, initialElements); err != nil {
return nil, fmt.Errorf("error adding set %s: %w", setName, err)
}
if err := r.conn.Flush(); err != nil {
return nil, fmt.Errorf("flush error: %w", err)
}
log.Debugf("Created new ipset: %s with %d initial prefixes (total prefixes %d)", setName, len(initialElements)/2, len(prefixes))
// The set is committed now. If a later batch fails, destroy it: the
// refcounter records nothing on a create-callback error, so it would
// otherwise leak, and a partial source set fails-open for deny rules.
if err := r.addRemainingElements(nfset, elements, maxElements); err != nil {
if derr := r.deleteIpSet(setName, nfset); derr != nil {
log.Warnf("rollback ipset %s after add failure: %v", setName, derr)
}
return nil, err
}
log.Infof("Created new ipset: %s with %d prefixes", setName, len(prefixes))
return nfset, nil
}
// addRemainingElements adds element batches beyond the initial one in
// maxElements-sized chunks, flushing each. Called after the set has been
// created with its first batch.
func (r *family) addRemainingElements(nfset *nftables.Set, elements []nftables.SetElement, maxElements int) error {
nElements := len(elements)
for subStart := maxElements; subStart < nElements; subStart += maxElements {
subEnd := min(subStart+maxElements, nElements)
subElement := elements[subStart:subEnd]
nSubPrefixes := len(subElement) / 2
log.Tracef("Adding new prefixes (%d) in ipset: %s", nSubPrefixes, nfset.Name)
if err := r.conn.SetAddElements(nfset, subElement); err != nil {
return fmt.Errorf("error adding prefixes (%d) to set %s: %w", nSubPrefixes, nfset.Name, err)
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf("flush error: %w", err)
}
log.Debugf("Added new prefixes (%d) in ipset: %s", nSubPrefixes, nfset.Name)
}
return nil
}
func (r *family) convertPrefixesToSet(prefixes []netip.Prefix) []nftables.SetElement {
var elements []nftables.SetElement
for _, prefix := range prefixes {
// nftables needs half-open intervals [firstIP, lastIP) for prefixes
// e.g. 10.0.0.0/24 becomes [10.0.0.0, 10.0.1.0), 10.1.1.1/32 becomes [10.1.1.1, 10.1.1.2) etc
firstIP := prefix.Addr()
// For a /0 the last address is the broadcast and its Next() overflows
// to an invalid Addr with an empty key, so wrap to the zero address,
// which nftables reads as the open end of a full-range interval.
var lastKey []byte
if prefix.Bits() == 0 {
lastKey = make([]byte, r.af.addrLen)
} else {
lastKey = calculateLastIP(prefix).Next().AsSlice()
}
// the nft tool also adds a zero-address IntervalEnd element, see https://github.com/google/nftables/issues/247
// nftables.SetElement{Key: make([]byte, r.af.addrLen), IntervalEnd: true},
elements = append(elements,
nftables.SetElement{Key: firstIP.AsSlice()},
nftables.SetElement{Key: lastKey, IntervalEnd: true},
)
}
return elements
}
// calculateLastIP determines the last IP in a given prefix.
func calculateLastIP(prefix netip.Prefix) netip.Addr {
masked := prefix.Masked()
if masked.Addr().Is4() {
hostMask := ^uint32(0) >> masked.Bits()
lastIP := uint32FromNetipAddr(masked.Addr()) | hostMask
return netip.AddrFrom4(uint32ToBytes(lastIP))
}
// IPv6: set host bits to all 1s
b := masked.Addr().As16()
bits := masked.Bits()
for i := bits; i < 128; i++ {
b[i/8] |= 1 << (7 - i%8)
}
return netip.AddrFrom16(b)
}
// Utility function to convert netip.Addr to uint32.
func uint32FromNetipAddr(addr netip.Addr) uint32 {
b := addr.As4()
return binary.BigEndian.Uint32(b[:])
}
// Utility function to convert uint32 to a netip-compatible byte slice.
func uint32ToBytes(ip uint32) [4]byte {
var b [4]byte
binary.BigEndian.PutUint32(b[:], ip)
return b
}
func (r *family) deleteIpSet(setName string, nfset *nftables.Set) error {
r.conn.DelSet(nfset)
if err := r.conn.Flush(); err != nil {
return fmt.Errorf(flushError, err)
}
log.Debugf("Deleted unused ipset %s", setName)
return nil
}
func (r *family) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
nfset, err := r.conn.GetSetByName(r.workTable, set.HashedName())
if err != nil {
return fmt.Errorf("get set %s: %w", set.HashedName(), err)
}
// Overlapping prefixes (e.g. duplicate resolved addresses) make the
// interval set reject the batch, so merge them as createIpSet does.
prefixes = firewall.MergeIPRanges(prefixes)
elements := r.convertPrefixesToSet(prefixes)
// Add in batches sized like createIpSet so a large update does not
// exceed the netlink message size limit.
maxElements := maxPrefixesSet * 2
for start := 0; start < len(elements); start += maxElements {
end := min(start+maxElements, len(elements))
if err := r.conn.SetAddElements(nfset, elements[start:end]); err != nil {
return fmt.Errorf("add elements to set %s: %w", set.HashedName(), err)
}
if err := r.conn.Flush(); err != nil {
return fmt.Errorf(flushError, err)
}
}
log.Debugf("updated set %s with %d prefixes", set.HashedName(), len(prefixes))
return nil
}
func (r *family) getIpSetExprs(ref refcounter.Ref[*nftables.Set], isSource bool) ([]expr.Any, error) {
// dst offset by default
offset := r.af.dstAddrOffset
if isSource {
// src offset
offset = r.af.srcAddrOffset
}
return []expr.Any{
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseNetworkHeader,
Offset: offset,
Len: r.af.addrLen,
},
&expr.Lookup{
SourceRegister: 1,
SetName: ref.Out.Name,
SetID: ref.Out.ID,
},
}, nil
}

View File

@@ -0,0 +1,36 @@
package nftables
import (
"net/netip"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
// TestConvertPrefixesToSetWildcard verifies that a /0 prefix produces a
// usable interval. The last address of a /0 is the broadcast, whose Next()
// overflows to an invalid Addr with an empty key; the IntervalEnd must wrap
// to the zero address instead so nftables sees a full-range interval.
func TestConvertPrefixesToSetWildcard(t *testing.T) {
tests := []struct {
name string
af addrFamily
prefix string
}{
{"IPv4 /0", afIPv4, "0.0.0.0/0"},
{"IPv6 /0", afIPv6, "::/0"},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
r := &family{af: tt.af}
elements := r.convertPrefixesToSet([]netip.Prefix{netip.MustParsePrefix(tt.prefix)})
require.Len(t, elements, 2, "expected start and interval-end element")
assert.False(t, elements[0].IntervalEnd, "first element is the interval start")
assert.True(t, elements[1].IntervalEnd, "second element is the interval end")
assert.Len(t, elements[1].Key, int(tt.af.addrLen), "interval-end key must be a zero address, not empty")
})
}
}

View File

@@ -1,85 +0,0 @@
package nftables
import (
"net"
)
type ipsetStore struct {
ipsetReference map[string]int
ipsets map[string]map[string]struct{} // ipsetName -> list of ips
}
func newIpsetStore() *ipsetStore {
return &ipsetStore{
ipsetReference: make(map[string]int),
ipsets: make(map[string]map[string]struct{}),
}
}
func (s *ipsetStore) ips(ipsetName string) (map[string]struct{}, bool) {
r, ok := s.ipsets[ipsetName]
return r, ok
}
func (s *ipsetStore) newIpset(ipsetName string) map[string]struct{} {
s.ipsetReference[ipsetName] = 0
ipList := make(map[string]struct{})
s.ipsets[ipsetName] = ipList
return ipList
}
func (s *ipsetStore) deleteIpset(ipsetName string) {
delete(s.ipsetReference, ipsetName)
delete(s.ipsets, ipsetName)
}
func (s *ipsetStore) DeleteIpFromSet(ipsetName string, ip net.IP) {
ipList, ok := s.ipsets[ipsetName]
if !ok {
return
}
delete(ipList, ip.String())
}
func (s *ipsetStore) AddIpToSet(ipsetName string, ip net.IP) {
ipList, ok := s.ipsets[ipsetName]
if !ok {
return
}
ipList[ip.String()] = struct{}{}
}
func (s *ipsetStore) IsIpInSet(ipsetName string, ip net.IP) bool {
ipList, ok := s.ipsets[ipsetName]
if !ok {
return false
}
_, ok = ipList[ip.String()]
return ok
}
func (s *ipsetStore) AddReferenceToIpset(ipsetName string) {
s.ipsetReference[ipsetName]++
}
func (s *ipsetStore) DeleteReferenceFromIpSet(ipsetName string) {
r, ok := s.ipsetReference[ipsetName]
if !ok {
return
}
if r == 0 {
return
}
s.ipsetReference[ipsetName]--
}
func (s *ipsetStore) HasReferenceToSet(ipsetName string) bool {
if _, ok := s.ipsetReference[ipsetName]; !ok {
return false
}
if s.ipsetReference[ipsetName] == 0 {
return false
}
return true
}

View File

@@ -0,0 +1,60 @@
package nftables
import (
"testing"
"github.com/google/nftables/expr"
"github.com/stretchr/testify/require"
)
func TestBuildLegacyRouteRuleExpressions(t *testing.T) {
sourcePayload := &expr.Payload{}
sourceCmp := &expr.Cmp{}
destinationPayload := &expr.Payload{}
destinationCmp := &expr.Cmp{}
nilSourceDestination := &expr.Payload{}
nilDestinationSource := &expr.Cmp{}
tests := []struct {
name string
source []expr.Any
destination []expr.Any
matches []expr.Any
}{
{
name: "both non-empty",
source: []expr.Any{sourcePayload, sourceCmp},
destination: []expr.Any{destinationPayload, destinationCmp},
matches: []expr.Any{sourcePayload, sourceCmp, destinationPayload, destinationCmp},
},
{
name: "nil source",
destination: []expr.Any{nilSourceDestination},
matches: []expr.Any{nilSourceDestination},
},
{
name: "nil destination",
source: []expr.Any{nilDestinationSource},
matches: []expr.Any{nilDestinationSource},
},
{
name: "both nil",
},
}
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
got := buildLegacyRouteRuleExpressions(tt.source, tt.destination)
require.Len(t, got, len(tt.matches)+2)
for i, match := range tt.matches {
require.Same(t, match, got[i])
}
require.IsType(t, &expr.Counter{}, got[len(tt.matches)])
verdict, ok := got[len(tt.matches)+1].(*expr.Verdict)
require.True(t, ok)
require.Equal(t, expr.VerdictAccept, verdict.Kind)
})
}
}

View File

@@ -3,7 +3,6 @@ package nftables
import (
"context"
"fmt"
"net"
"net/netip"
"os"
"sync"
@@ -16,7 +15,6 @@ import (
"golang.org/x/sys/unix"
nberrors "github.com/netbirdio/netbird/client/errors"
"github.com/netbirdio/netbird/client/firewall/firewalld"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
"github.com/netbirdio/netbird/client/iface/wgaddr"
"github.com/netbirdio/netbird/client/internal/statemanager"
@@ -45,18 +43,17 @@ type iFaceMapper interface {
Address() wgaddr.Address
}
// Manager of iptables firewall
// Manager of nftables firewall. Per-family state (peer ACLs, route
// ACLs, NAT, DNAT, MSS clamping) lives on family; Manager dispatches
// by family and provides the public firewall.Manager surface.
type Manager struct {
mutex sync.Mutex
rConn *nftables.Conn
wgIface iFaceMapper
router *router
aclManager *AclManager
// IPv6 counterparts, nil when no v6 overlay
router6 *router
aclManager6 *AclManager
family4 *family
// IPv6 counterpart, nil when no v6 overlay.
family6 *family
notrackOutputChain *nftables.Chain
notrackPreroutingChain *nftables.Chain
@@ -74,21 +71,10 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
tableName := getTableName()
workTable := &nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv4}
var err error
m.router, err = newRouter(workTable, wgIface, mtu)
if err != nil {
return nil, fmt.Errorf("create router: %w", err)
}
m.aclManager, err = newAclManager(workTable, wgIface, chainNameRoutingFw)
if err != nil {
return nil, fmt.Errorf("create acl manager: %w", err)
}
m.family4 = newFamily(workTable, wgIface, mtu)
if wgIface.Address().HasIPv6() {
if err := m.createIPv6Components(tableName, wgIface, mtu); err != nil {
return nil, fmt.Errorf("create IPv6 firewall: %w", err)
}
m.createIPv6Components(tableName, wgIface, mtu)
}
m.extMonitor = newExternalChainMonitor(m)
@@ -96,30 +82,19 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
return m, nil
}
func (m *Manager) createIPv6Components(tableName string, wgIface iFaceMapper, mtu uint16) error {
func (m *Manager) createIPv6Components(tableName string, wgIface iFaceMapper, mtu uint16) {
workTable6 := &nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv6}
var err error
m.router6, err = newRouter(workTable6, wgIface, mtu)
if err != nil {
return fmt.Errorf("create v6 router: %w", err)
}
m.family6 = newFamily(workTable6, wgIface, mtu)
// Share the same IP forwarding state with the v4 router, since
// EnableIPForwarding controls both v4 and v6 sysctls.
m.router6.ipFwdState = m.router.ipFwdState
m.aclManager6, err = newAclManager(workTable6, wgIface, chainNameRoutingFw)
if err != nil {
return fmt.Errorf("create v6 acl manager: %w", err)
}
return nil
m.family6.ipFwdState = m.family4.ipFwdState
}
// hasIPv6 reports whether the manager has IPv6 components initialized.
func (m *Manager) hasIPv6() bool {
return m.router6 != nil
return m.family6 != nil
}
func (m *Manager) initIPv6() error {
@@ -128,12 +103,8 @@ func (m *Manager) initIPv6() error {
return fmt.Errorf("create v6 work table: %w", err)
}
if err := m.router6.init(workTable6); err != nil {
return fmt.Errorf("v6 router init: %w", err)
}
if err := m.aclManager6.init(workTable6); err != nil {
return fmt.Errorf("v6 acl manager init: %w", err)
if err := m.family6.init(workTable6); err != nil {
return fmt.Errorf("v6 family init: %w", err)
}
return nil
@@ -156,19 +127,20 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
// reconcileExternalChains re-applies passthrough accept rules to external
// filter chains for both IPv4 and IPv6 routers. Called by the monitor when
// tables or chains appear (e.g. after firewalld reloads).
// tables or chains appear (e.g. after firewalld reloads). Kernel routing opens
// both INPUT and FORWARD.
func (m *Manager) reconcileExternalChains() error {
m.mutex.Lock()
defer m.mutex.Unlock()
var merr *multierror.Error
if m.router != nil {
if err := m.router.acceptExternalChainsRules(); err != nil {
if m.family4 != nil {
if err := m.family4.acceptExternalChainsRules(true); err != nil {
merr = multierror.Append(merr, fmt.Errorf("v4: %w", err))
}
}
if m.hasIPv6() {
if err := m.router6.acceptExternalChainsRules(); err != nil {
if err := m.family6.acceptExternalChainsRules(true); err != nil {
merr = multierror.Append(merr, fmt.Errorf("v6: %w", err))
}
}
@@ -187,12 +159,8 @@ func (m *Manager) initFirewall() (err error) {
}
}()
if err := m.router.init(workTable); err != nil {
return fmt.Errorf("router init: %w", err)
}
if err := m.aclManager.init(workTable); err != nil {
return fmt.Errorf("acl manager init: %w", err)
if err := m.family4.init(workTable); err != nil {
return fmt.Errorf("family init: %w", err)
}
if m.hasIPv6() {
@@ -220,7 +188,7 @@ func (m *Manager) persistState(stateManager *statemanager.Manager) {
InterfaceState: &InterfaceState{
NameStr: m.wgIface.Name(),
WGAddress: m.wgIface.Address(),
MTU: m.router.mtu,
MTU: m.family4.mtu,
},
}); err != nil {
log.Errorf("failed to update state: %v", err)
@@ -235,12 +203,12 @@ func (m *Manager) persistState(stateManager *statemanager.Manager) {
// rollbackInit performs best-effort cleanup of already-initialized state when Init fails partway through.
func (m *Manager) rollbackInit() {
if err := m.router.Reset(); err != nil {
log.Warnf("rollback router: %v", err)
if err := m.family4.Reset(); err != nil {
log.Warnf("rollback family: %v", err)
}
if m.hasIPv6() {
if err := m.router6.Reset(); err != nil {
log.Warnf("rollback v6 router: %v", err)
if err := m.family6.Reset(); err != nil {
log.Warnf("rollback v6 family: %v", err)
}
}
if err := m.cleanupNetbirdTables(); err != nil {
@@ -251,118 +219,82 @@ func (m *Manager) rollbackInit() {
}
}
// AddPeerFiltering rule to the firewall
// AddFilterRule installs a packet-filtering rule.
//
// If comment argument is empty firewall manager should set
// rule ID as comment for the rule
func (m *Manager) AddPeerFiltering(
id []byte,
ip net.IP,
proto firewall.Protocol,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
ipsetName string,
) ([]firewall.Rule, error) {
m.mutex.Lock()
defer m.mutex.Unlock()
if ip.To4() != nil {
return m.aclManager.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
}
if !m.hasIPv6() {
return nil, fmt.Errorf("add peer filtering for %s: %w", ip, firewall.ErrIPv6NotInitialized)
}
return m.aclManager6.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
}
func (m *Manager) AddRouteFiltering(
// Destination semantics: zero Network → input chain (peer ACL);
// set Network → forward chain (route ACL).
//
// Sources are a single address family; the rule is dispatched to the
// matching per-family backend.
func (m *Manager) AddFilterRule(
id []byte,
sources []netip.Prefix,
destination firewall.Network,
proto firewall.Protocol,
sPort, dPort *firewall.Port,
sPort *firewall.Port,
dPort *firewall.Port,
action firewall.Action,
) (firewall.Rule, error) {
if len(sources) == 0 {
return nil, firewall.ErrNoSources
}
m.mutex.Lock()
defer m.mutex.Unlock()
if isIPv6RouteRule(sources, destination) {
fam := m.family4
if isIPv6Rule(sources, destination) {
if !m.hasIPv6() {
return nil, fmt.Errorf("add route filtering: %w", firewall.ErrIPv6NotInitialized)
return nil, fmt.Errorf("add filtering: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
fam = m.family6
}
return m.router.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
return fam.AddFilterRule(id, sources, destination, proto, sPort, dPort, action)
}
// DeletePeerRule from the firewall by rule definition
func (m *Manager) DeletePeerRule(rule firewall.Rule) error {
// DeleteFilterRule removes a filtering rule. The owning family is found
// by id in the in-memory filter maps, which are the only tracking for
// filter rules. family.DeleteFilterRule is idempotent when the id is
// absent.
func (m *Manager) DeleteFilterRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
if m.hasIPv6() && isIPv6Rule(rule) {
return m.aclManager6.DeletePeerRule(rule)
}
return m.aclManager.DeletePeerRule(rule)
}
func isIPv6Rule(rule firewall.Rule) bool {
r, ok := rule.(*Rule)
return ok && r.nftRule != nil && r.nftRule.Table != nil && r.nftRule.Table.Family == nftables.TableFamilyIPv6
}
// isIPv6RouteRule determines whether a route rule belongs to the v6 table.
// For static routes, the destination prefix determines the family. For dynamic
// routes (DomainSet), the sources determine the family since management
// duplicates dynamic rules per family.
func isIPv6RouteRule(sources []netip.Prefix, destination firewall.Network) bool {
if destination.IsPrefix() {
return destination.Prefix.Addr().Is6()
}
return len(sources) > 0 && sources[0].Addr().Is6()
}
// DeleteRouteRule deletes a routing rule. Route rules live in exactly one
// router; the cached maps are normally authoritative, so the kernel is only
// consulted when neither map knows about the rule.
func (m *Manager) DeleteRouteRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
id := rule.ID()
r, err := m.routerForRuleID(id, (*router).hasRule)
fam, err := m.familyForRuleID(rule.ID(), (*family).hasRule, false)
if err != nil {
return err
}
return r.DeleteRouteRule(rule)
return fam.DeleteFilterRule(rule)
}
// routerForRuleID picks the router holding the rule with the given id, using
// the supplied lookup. If the cached maps disagree (or both miss), it refreshes
// from the kernel once and re-checks before falling back to the v4 router.
func (m *Manager) routerForRuleID(id string, has func(*router, string) bool) (*router, error) {
if has(m.router, id) {
return m.router, nil
}
if m.hasIPv6() && has(m.router6, id) {
return m.router6, nil
// familyForRuleID picks the family holding the rule with the given id, using
// the supplied lookup. With refresh set, a miss in both cached maps reloads
// the NAT/DNAT rule maps from the kernel once and re-checks before falling
// back to the v4 family. Filter rules are tracked only in memory and have no
// kernel-backed reload, so their callers pass refresh as false.
func (m *Manager) familyForRuleID(id firewall.RuleID, has func(*family, firewall.RuleID) bool, refresh bool) (*family, error) {
if has(m.family4, id) {
return m.family4, nil
}
if !m.hasIPv6() {
return m.router, nil
return m.family4, nil
}
if err := m.router.refreshRulesMap(); err != nil {
if has(m.family6, id) {
return m.family6, nil
}
if !refresh {
return m.family4, nil
}
if err := m.family4.refreshRulesMap(); err != nil {
return nil, fmt.Errorf("refresh v4 rules: %w", err)
}
if err := m.router6.refreshRulesMap(); err != nil {
if err := m.family6.refreshRulesMap(); err != nil {
return nil, fmt.Errorf("refresh v6 rules: %w", err)
}
if has(m.router6, id) && !has(m.router, id) {
return m.router6, nil
if has(m.family6, id) && !has(m.family4, id) {
return m.family6, nil
}
return m.router, nil
return m.family4, nil
}
func (m *Manager) IsServerRouteSupported() bool {
@@ -381,10 +313,10 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
if !m.hasIPv6() {
return fmt.Errorf("add NAT rule: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddNatRule(pair)
return m.family6.AddNatRule(pair)
}
if err := m.router.AddNatRule(pair); err != nil {
if err := m.family4.AddNatRule(pair); err != nil {
return err
}
@@ -396,7 +328,7 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
// so the eventual cleanup still works.
if m.hasIPv6() && pair.Dynamic {
v6Pair := firewall.ToV6NatPair(pair)
if err := m.router6.AddNatRule(v6Pair); err != nil {
if err := m.family6.AddNatRule(v6Pair); err != nil {
return fmt.Errorf("add v6 NAT rule: %w", err)
}
}
@@ -412,18 +344,18 @@ func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
if !m.hasIPv6() {
return nil
}
return m.router6.RemoveNatRule(pair)
return m.family6.RemoveNatRule(pair)
}
var merr *multierror.Error
if err := m.router.RemoveNatRule(pair); err != nil {
if err := m.family4.RemoveNatRule(pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove v4 NAT rule: %w", err))
}
if m.hasIPv6() && pair.Dynamic {
v6Pair := firewall.ToV6NatPair(pair)
if err := m.router6.RemoveNatRule(v6Pair); err != nil {
if err := m.family6.RemoveNatRule(v6Pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove v6 NAT rule: %w", err))
}
}
@@ -431,46 +363,13 @@ func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
return nberrors.FormatErrorOrNil(merr)
}
// AllowNetbird allows netbird interface traffic.
// This is called when USPFilter wraps the native firewall, adding blanket accept
// rules so that packet filtering is handled in userspace instead of by netfilter.
//
// TODO: In USP mode this only adds ACCEPT to the netbird table's own chains,
// which doesn't override DROP rules in external tables (e.g. firewalld).
// Should add passthrough rules to external chains (like the native mode router's
// addExternalChainsRules does) for both the netbird table family and inet tables.
// The netbird table itself is fine (routing chains already exist there), but
// non-netbird tables with INPUT/FORWARD hooks can still DROP our WG traffic.
func (m *Manager) AllowNetbird() error {
m.mutex.Lock()
defer m.mutex.Unlock()
if err := m.aclManager.createDefaultAllowRules(); err != nil {
return fmt.Errorf("create default allow rules: %w", err)
}
if m.hasIPv6() {
if err := m.aclManager6.createDefaultAllowRules(); err != nil {
return fmt.Errorf("create v6 default allow rules: %w", err)
}
}
if err := m.rConn.Flush(); err != nil {
return fmt.Errorf("flush allow input netbird rules: %w", err)
}
if err := firewalld.TrustInterface(m.wgIface.Name()); err != nil {
log.Warnf("failed to trust interface in firewalld: %v", err)
}
return nil
}
// SetLegacyManagement sets the route manager to use legacy management
func (m *Manager) SetLegacyManagement(isLegacy bool) error {
if err := firewall.SetLegacyManagement(m.router, isLegacy); err != nil {
if err := firewall.SetLegacyManagement(m.family4, isLegacy); err != nil {
return err
}
if m.hasIPv6() {
return firewall.SetLegacyManagement(m.router6, isLegacy)
return firewall.SetLegacyManagement(m.family6, isLegacy)
}
return nil
}
@@ -484,13 +383,13 @@ func (m *Manager) Close(stateManager *statemanager.Manager) error {
var merr *multierror.Error
if err := m.router.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset router: %v", err))
if err := m.family4.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset family: %w", err))
}
if m.hasIPv6() {
if err := m.router6.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset v6 router: %v", err))
if err := m.family6.Reset(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("reset v6 family: %w", err))
}
}
@@ -530,14 +429,14 @@ func (m *Manager) SetLogLevel(log.Level) {
}
func (m *Manager) EnableRouting() error {
if err := m.router.ipFwdState.RequestForwarding(); err != nil {
if err := m.family4.ipFwdState.RequestForwarding(); err != nil {
return fmt.Errorf("enable IP forwarding: %w", err)
}
return nil
}
func (m *Manager) DisableRouting() error {
if err := m.router.ipFwdState.ReleaseForwarding(); err != nil {
if err := m.family4.ipFwdState.ReleaseForwarding(); err != nil {
return fmt.Errorf("disable IP forwarding: %w", err)
}
return nil
@@ -551,13 +450,13 @@ func (m *Manager) Flush() error {
m.mutex.Lock()
defer m.mutex.Unlock()
if err := m.aclManager.Flush(); err != nil {
if err := m.family4.Flush(); err != nil {
return err
}
if m.hasIPv6() {
if err := m.aclManager6.Flush(); err != nil {
return fmt.Errorf("flush v6 acl: %w", err)
if err := m.family6.Flush(); err != nil {
return fmt.Errorf("flush v6 family: %w", err)
}
}
@@ -577,9 +476,9 @@ func (m *Manager) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error)
if !m.hasIPv6() {
return nil, fmt.Errorf("add DNAT rule: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddDNATRule(rule)
return m.family6.AddDNATRule(rule)
}
return m.router.AddDNATRule(rule)
return m.family4.AddDNATRule(rule)
}
// DeleteDNATRule deletes a DNAT rule
@@ -587,7 +486,7 @@ func (m *Manager) DeleteDNATRule(rule firewall.Rule) error {
m.mutex.Lock()
defer m.mutex.Unlock()
r, err := m.routerForRuleID(rule.ID(), (*router).hasDNATRule)
r, err := m.familyForRuleID(rule.ID(), (*family).hasDNATRule, true)
if err != nil {
return err
}
@@ -608,12 +507,12 @@ func (m *Manager) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
}
}
if err := m.router.UpdateSet(set, v4Prefixes); err != nil {
if err := m.family4.UpdateSet(set, v4Prefixes); err != nil {
return err
}
if m.hasIPv6() && len(v6Prefixes) > 0 {
if err := m.router6.UpdateSet(set, v6Prefixes); err != nil {
if err := m.family6.UpdateSet(set, v6Prefixes); err != nil {
return fmt.Errorf("update v6 set: %w", err)
}
}
@@ -630,9 +529,9 @@ func (m *Manager) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protoco
if !m.hasIPv6() {
return fmt.Errorf("add inbound DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.AddInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
// RemoveInboundDNAT removes an inbound DNAT rule.
@@ -644,9 +543,9 @@ func (m *Manager) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Prot
if !m.hasIPv6() {
return fmt.Errorf("remove inbound DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.RemoveInboundDNAT(localAddr, protocol, originalPort, translatedPort)
}
// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
@@ -658,9 +557,9 @@ func (m *Manager) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol
if !m.hasIPv6() {
return fmt.Errorf("add output DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.AddOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
@@ -672,9 +571,9 @@ func (m *Manager) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Proto
if !m.hasIPv6() {
return fmt.Errorf("remove output DNAT: %w", firewall.ErrIPv6NotInitialized)
}
return m.router6.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family6.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
return m.router.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
return m.family4.RemoveOutputDNAT(localAddr, protocol, originalPort, translatedPort)
}
const (
@@ -903,3 +802,14 @@ func getEstablishedExprs(register uint32) []expr.Any {
},
}
}
// isIPv6Rule reports whether the rule belongs to the v6 table. For a
// prefix destination the destination family decides; otherwise the
// (single-family) sources do, since management duplicates rules per
// family.
func isIPv6Rule(sources []netip.Prefix, destination firewall.Network) bool {
if destination.IsPrefix() {
return destination.Prefix.Addr().Is6()
}
return len(sources) > 0 && sources[0].Addr().Is6()
}

View File

@@ -72,13 +72,13 @@ func TestNftablesManager(t *testing.T) {
testClient := &nftables.Conn{}
rule, err := manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{53}}, fw.ActionDrop, "")
rule, err := manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{53}}, fw.ActionDrop)
require.NoError(t, err, "failed to add rule")
err = manager.Flush()
require.NoError(t, err, "failed to flush")
rules, err := testClient.GetRules(manager.aclManager.workTable, manager.aclManager.chainInputRules)
rules, err := testClient.GetRules(manager.family4.workTable, manager.family4.chainInputRules)
require.NoError(t, err, "failed to get rules")
require.Len(t, rules, 2, "expected 2 rules")
@@ -149,15 +149,12 @@ func TestNftablesManager(t *testing.T) {
// Compare connection tracking rule at position 1 (pushed down by DROP rule insertion)
compareExprsIgnoringCounters(t, rules[1].Exprs, expectedExprs1)
for _, r := range rule {
err = manager.DeletePeerRule(r)
require.NoError(t, err, "failed to delete rule")
}
require.NoError(t, manager.DeleteFilterRule(rule), "failed to delete rule")
err = manager.Flush()
require.NoError(t, err, "failed to flush")
rules, err = testClient.GetRules(manager.aclManager.workTable, manager.aclManager.chainInputRules)
rules, err = testClient.GetRules(manager.family4.workTable, manager.family4.chainInputRules)
require.NoError(t, err, "failed to get rules")
// established rule remains
require.Len(t, rules, 1, "expected 1 rules after deletion")
@@ -182,47 +179,39 @@ func TestNftablesManagerRuleOrder(t *testing.T) {
testClient := &nftables.Conn{}
// Add accept rule first
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept, "accept-http")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept)
require.NoError(t, err, "failed to add accept rule")
// Add deny rule second for the same traffic
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionDrop, "deny-http")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionDrop)
require.NoError(t, err, "failed to add deny rule")
err = manager.Flush()
require.NoError(t, err, "failed to flush")
rules, err := testClient.GetRules(manager.aclManager.workTable, manager.aclManager.chainInputRules)
rules, err := testClient.GetRules(manager.family4.workTable, manager.family4.chainInputRules)
require.NoError(t, err, "failed to get rules")
t.Logf("Found %d rules in nftables chain", len(rules))
// Find the accept and deny rules and verify deny comes before accept
// Single-source rules emit a direct payload+cmp on the source IP
// (no set lookup). Match by source-IP + port + verdict instead of
// the legacy per-(action,port) set names ("deny-http"/"accept-http")
// that this test predates.
wantSrc := ip.AsSlice()
var acceptRuleIndex, denyRuleIndex = -1, -1
for i, rule := range rules {
hasAcceptHTTPSet := false
hasDenyHTTPSet := false
hasPort80 := false
var hasSrc, hasPort80 bool
var action string
for _, e := range rule.Exprs {
// Check for set lookup
if lookup, ok := e.(*expr.Lookup); ok {
switch lookup.SetName {
case "accept-http":
hasAcceptHTTPSet = true
case "deny-http":
hasDenyHTTPSet = true
if cmp, ok := e.(*expr.Cmp); ok && cmp.Op == expr.CmpOpEq {
if bytes.Equal(cmp.Data, wantSrc) {
hasSrc = true
}
}
// Check for port 80
if cmp, ok := e.(*expr.Cmp); ok {
if cmp.Op == expr.CmpOpEq && len(cmp.Data) == 2 && binary.BigEndian.Uint16(cmp.Data) == 80 {
if len(cmp.Data) == 2 && binary.BigEndian.Uint16(cmp.Data) == 80 {
hasPort80 = true
}
}
// Check for verdict
if verdict, ok := e.(*expr.Verdict); ok {
switch verdict.Kind {
case expr.VerdictAccept:
@@ -233,11 +222,15 @@ func TestNftablesManagerRuleOrder(t *testing.T) {
}
}
if hasAcceptHTTPSet && hasPort80 && action == "ACCEPT" {
t.Logf("Rule [%d]: accept-http set + Port 80 + ACCEPT", i)
if !hasSrc || !hasPort80 {
continue
}
switch action {
case "ACCEPT":
t.Logf("Rule [%d]: src=%s port=80 ACCEPT", i, ip)
acceptRuleIndex = i
} else if hasDenyHTTPSet && hasPort80 && action == "DROP" {
t.Logf("Rule [%d]: deny-http set + Port 80 + DROP", i)
case "DROP":
t.Logf("Rule [%d]: src=%s port=80 DROP", i, ip)
denyRuleIndex = i
}
}
@@ -281,7 +274,7 @@ func TestNFtablesCreatePerformance(t *testing.T) {
start := time.Now()
for i := 0; i < testMax; i++ {
port := &fw.Port{Values: []uint16{uint16(1000 + i)}}
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), "tcp", nil, port, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, "tcp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
if i%100 == 0 {
@@ -363,10 +356,10 @@ func TestNftablesManagerCompatibilityWithIptables(t *testing.T) {
})
ip := netip.MustParseAddr("100.96.0.1")
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept)
require.NoError(t, err, "failed to add peer filtering rule")
_, err = manager.AddRouteFiltering(
_, err = manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("192.168.2.0/24")},
fw.Network{Prefix: netip.MustParsePrefix("10.1.0.0/24")},
@@ -439,10 +432,10 @@ func TestNftablesManagerIPv6CompatibilityWithIp6tables(t *testing.T) {
})
ip := netip.MustParseAddr("fd00::2")
_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept)
require.NoError(t, err, "add v6 peer filtering rule")
_, err = manager.AddRouteFiltering(
_, err = manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("fd00:1::/64")},
fw.Network{Prefix: netip.MustParsePrefix("2001:db8::/48")},
@@ -552,7 +545,7 @@ func TestNftablesManagerCompatibilityWithIptablesFor6kPrefixes(t *testing.T) {
prefixes = append(prefixes, netip.PrefixFrom(addr, 24))
}
}
_, err = manager.AddRouteFiltering(
_, err = manager.AddFilterRule(
nil,
prefixes,
fw.Network{Prefix: netip.MustParsePrefix("10.2.0.0/24")},
@@ -567,7 +560,7 @@ func TestNftablesManagerCompatibilityWithIptablesFor6kPrefixes(t *testing.T) {
verifyIptablesOutput(t, stdout, stderr)
}
func TestNftablesManagerCompatibilityWithIptablesForEmptyPrefixes(t *testing.T) {
func TestNftablesManagerCompatibilityWithIptablesForWildcardSource(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
}
@@ -593,9 +586,9 @@ func TestNftablesManagerCompatibilityWithIptablesForEmptyPrefixes(t *testing.T)
verifyIptablesOutput(t, stdout, stderr)
})
_, err = manager.AddRouteFiltering(
_, err = manager.AddFilterRule(
nil,
[]netip.Prefix{},
[]netip.Prefix{netip.MustParsePrefix("0.0.0.0/0")},
fw.Network{Prefix: netip.MustParsePrefix("10.2.0.0/24")},
fw.ProtocolTCP,
nil,
@@ -608,6 +601,73 @@ func TestNftablesManagerCompatibilityWithIptablesForEmptyPrefixes(t *testing.T)
verifyIptablesOutput(t, stdout, stderr)
}
func TestNftablesManagerMultiPortFilter(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
}
manager, err := Create(ifaceMock, iface.DefaultMTU)
require.NoError(t, err)
require.NoError(t, manager.Init(nil))
t.Cleanup(func() {
require.NoError(t, manager.Close(nil), "failed to reset manager state")
})
ip := netip.MustParseAddr("100.96.0.1")
rule, err := manager.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80, 443}}, fw.ActionAccept)
require.NoError(t, err, "failed to add multi-port rule")
testClient := &nftables.Conn{}
rules, err := testClient.GetRules(manager.family4.workTable, manager.family4.chainInputRules)
require.NoError(t, err, "failed to get rules")
var lookup *expr.Lookup
for _, kernelRule := range rules {
if string(kernelRule.UserData) != string(rule.ID()) {
continue
}
for _, e := range kernelRule.Exprs {
if l, ok := e.(*expr.Lookup); ok {
lookup = l
}
}
}
require.NotNil(t, lookup, "multi-port rule must match ports via a set lookup")
sets, err := testClient.GetSets(manager.family4.workTable)
require.NoError(t, err, "failed to get sets")
var portSet *nftables.Set
for _, s := range sets {
if s.Name == lookup.SetName {
portSet = s
}
}
require.NotNil(t, portSet, "anonymous port set not found in kernel")
portSet.Table = manager.family4.workTable
elements, err := testClient.GetSetElements(portSet)
require.NoError(t, err, "failed to get set elements")
ports := make(map[uint16]bool)
for _, e := range elements {
require.Len(t, e.Key, 2, "port set element key should be 2 bytes")
ports[binary.BigEndian.Uint16(e.Key)] = true
}
require.True(t, ports[80], "port set should contain port 80")
require.True(t, ports[443], "port set should contain port 443")
require.NoError(t, manager.DeleteFilterRule(rule), "failed to delete rule")
rules, err = testClient.GetRules(manager.family4.workTable, manager.family4.chainInputRules)
require.NoError(t, err, "failed to get rules after delete")
for _, kernelRule := range rules {
require.NotEqual(t, string(rule.ID()), string(kernelRule.UserData), "rule should be removed from kernel")
}
}
func compareExprsIgnoringCounters(t *testing.T, got, want []expr.Any) {
t.Helper()
require.Equal(t, len(got), len(want), "expression count mismatch")

File diff suppressed because it is too large Load Diff

View File

@@ -37,7 +37,7 @@ func TestNftablesManager_AddNatRule(t *testing.T) {
for _, testCase := range test.InsertRuleTestCases {
t.Run(testCase.Name, func(t *testing.T) {
// need fw manager to init both acl mgr and router for all chains to be present
// need fw manager to init both acl mgr and family for all chains to be present
manager, err := Create(ifaceMock, iface.DefaultMTU)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -47,7 +47,7 @@ func TestNftablesManager_AddNatRule(t *testing.T) {
nftablesTestingClient := &nftables.Conn{}
rtr := manager.router
rtr := manager.family4
err = rtr.AddNatRule(testCase.InputPair)
require.NoError(t, err, "pair should be inserted")
@@ -90,9 +90,9 @@ func TestNftablesManager_AddNatRule(t *testing.T) {
}
// Build CIDR matching expressions
testRouter := &router{af: afIPv4}
sourceExp := testRouter.applyPrefix(testCase.InputPair.Source.Prefix, true)
destExp := testRouter.applyPrefix(testCase.InputPair.Destination.Prefix, false)
testRouter := &family{af: afIPv4}
sourceExp := prefixMatchExprs(testRouter.af, testCase.InputPair.Source.Prefix, true)
destExp := prefixMatchExprs(testRouter.af, testCase.InputPair.Destination.Prefix, false)
// Combine all expressions in the correct order
// nolint:gocritic
@@ -100,14 +100,14 @@ func TestNftablesManager_AddNatRule(t *testing.T) {
testingExpression = append(testingExpression, sourceExp...)
testingExpression = append(testingExpression, destExp...)
natRuleKey := firewall.GenKey(firewall.PreroutingFormat, testCase.InputPair)
natRuleKey := testCase.InputPair.GenKey(firewall.PreroutingFormat)
found := 0
for _, chain := range rtr.chains {
if chain.Name == chainNameManglePrerouting {
rules, err := nftablesTestingClient.GetRules(chain.Table, chain)
require.NoError(t, err, "should list rules for %s table and %s chain", chain.Table.Name, chain.Name)
for _, rule := range rules {
if len(rule.UserData) > 0 && string(rule.UserData) == natRuleKey {
if len(rule.UserData) > 0 && firewall.RuleID(rule.UserData) == natRuleKey {
// Compare expressions up to the mark setting expressions
require.ElementsMatchf(t, rule.Exprs[:len(testingExpression)], testingExpression, "prerouting nat rule elements should match")
found = 1
@@ -135,19 +135,19 @@ func TestNftablesManager_RemoveNatRule(t *testing.T) {
require.NoError(t, err)
require.NoError(t, manager.Init(nil))
rtr := manager.router
rtr := manager.family4
// First add the NAT rule using the router's method
// First add the NAT rule using the family's method
err = rtr.AddNatRule(testCase.InputPair)
require.NoError(t, err, "should add NAT rule")
// Verify the rule was added
natRuleKey := firewall.GenKey(firewall.PreroutingFormat, testCase.InputPair)
natRuleKey := testCase.InputPair.GenKey(firewall.PreroutingFormat)
found := false
rules, err := rtr.conn.GetRules(rtr.workTable, rtr.chains[chainNameManglePrerouting])
require.NoError(t, err, "should list rules")
for _, rule := range rules {
if len(rule.UserData) > 0 && string(rule.UserData) == natRuleKey {
if len(rule.UserData) > 0 && firewall.RuleID(rule.UserData) == natRuleKey {
found = true
break
}
@@ -163,7 +163,7 @@ func TestNftablesManager_RemoveNatRule(t *testing.T) {
rules, err = rtr.conn.GetRules(rtr.workTable, rtr.chains[chainNameManglePrerouting])
require.NoError(t, err, "should list rules after removal")
for _, rule := range rules {
if len(rule.UserData) > 0 && string(rule.UserData) == natRuleKey {
if len(rule.UserData) > 0 && firewall.RuleID(rule.UserData) == natRuleKey {
found = true
break
}
@@ -200,11 +200,10 @@ func TestRouter_AddRouteFiltering(t *testing.T) {
defer deleteWorkTable()
r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "Failed to create router")
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func(r *router) {
defer func(r *family) {
require.NoError(t, r.Reset(), "Failed to reset rules")
}(r)
@@ -314,16 +313,16 @@ func TestRouter_AddRouteFiltering(t *testing.T) {
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
ruleKey, err := r.AddRouteFiltering(nil, tt.sources, firewall.Network{Prefix: tt.destination}, tt.proto, tt.sPort, tt.dPort, tt.action)
require.NoError(t, err, "AddRouteFiltering failed")
ruleKey, err := r.AddFilterRule(nil, tt.sources, firewall.Network{Prefix: tt.destination}, tt.proto, tt.sPort, tt.dPort, tt.action)
require.NoError(t, err, "AddFilterRule failed")
t.Cleanup(func() {
require.NoError(t, r.DeleteRouteRule(ruleKey), "Failed to delete rule")
require.NoError(t, r.DeleteFilterRule(ruleKey), "Failed to delete rule")
})
// Check if the rule is in the internal map
rule, ok := r.rules[ruleKey.ID()]
assert.True(t, ok, "Rule not found in internal map")
stored, ok := r.filters[id.RuleID(ruleKey.ID())]
require.True(t, ok, "Rule not found in filters map")
rule := stored.nftRule
t.Log("Internal rule expressions:")
for i, expr := range rule.Exprs {
@@ -339,7 +338,7 @@ func TestRouter_AddRouteFiltering(t *testing.T) {
var nftRule *nftables.Rule
for _, rule := range rules {
if string(rule.UserData) == ruleKey.ID() {
if firewall.RuleID(rule.UserData) == ruleKey.ID() {
nftRule = rule
break
}
@@ -367,12 +366,11 @@ func TestNftablesCreateIpSet(t *testing.T) {
defer deleteWorkTable()
r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "Failed to create router")
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() {
require.NoError(t, r.Reset(), "Failed to reset router")
require.NoError(t, r.Reset(), "Failed to reset family")
}()
tests := []struct {
@@ -509,6 +507,41 @@ func TestNftablesCreateIpSet(t *testing.T) {
}
}
// TestNftablesUpdateSetMergesOverlapping verifies that UpdateSet merges
// overlapping prefixes before adding them. An interval set rejects
// overlapping elements, so without the merge a batch holding a /32 already
// covered by a /24, or a duplicate address as DNS resolution can produce,
// would fail.
func TestNftablesUpdateSetMergesOverlapping(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
}
workTable, err := createWorkTable()
require.NoError(t, err, "create work table")
defer deleteWorkTable()
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() {
require.NoError(t, r.Reset(), "reset family")
}()
initial := []netip.Prefix{netip.MustParsePrefix("10.0.0.0/24")}
set := firewall.NewPrefixSet(initial)
created, err := r.createIpSet(set.HashedName(), setInput{prefixes: initial})
require.NoError(t, err, "create ip set")
require.NotNil(t, created)
overlapping := []netip.Prefix{
netip.MustParsePrefix("192.168.1.0/24"),
netip.MustParsePrefix("192.168.1.1/32"),
netip.MustParsePrefix("192.168.1.1/32"),
}
require.NoError(t, r.UpdateSet(set, overlapping), "UpdateSet must merge overlapping prefixes")
}
func TestNftablesCreateIpSet_IPv6(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
@@ -518,11 +551,10 @@ func TestNftablesCreateIpSet_IPv6(t *testing.T) {
require.NoError(t, err, "Failed to create v6 work table")
defer deleteWorkTableIPv6()
r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, err, "Failed to create router")
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() {
require.NoError(t, r.Reset(), "Failed to reset router")
require.NoError(t, r.Reset(), "Failed to reset family")
}()
tests := []struct {
@@ -748,6 +780,14 @@ func containsPort(exprs []expr.Any, port *firewall.Port, isSource bool) bool {
}
}
}
case *expr.Lookup:
// Multiple discrete ports compile to an anonymous set lookup
// rather than a chain of comparisons. The set's id and name are
// assigned dynamically, so matching the lookup is enough here;
// the set elements are verified separately.
if !port.IsRange && len(port.Values) > 1 {
portMatchFound = true
}
}
if payloadFound && portMatchFound {
return true
@@ -861,13 +901,12 @@ func TestRouter_RefreshRulesMap_RemovesStaleEntries(t *testing.T) {
require.NoError(t, err)
defer deleteWorkTable()
r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, err)
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() { require.NoError(t, r.Reset()) }()
// Add a real rule to the kernel
ruleKey, err := r.AddRouteFiltering(
ruleKey, err := r.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("192.168.1.0/24")},
firewall.Network{Prefix: netip.MustParsePrefix("10.0.0.0/24")},
@@ -878,11 +917,11 @@ func TestRouter_RefreshRulesMap_RemovesStaleEntries(t *testing.T) {
)
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, r.DeleteRouteRule(ruleKey))
require.NoError(t, r.DeleteFilterRule(ruleKey))
})
// Inject a stale entry with Handle=0 (simulates store-before-flush failure)
staleKey := "stale-rule-that-does-not-exist"
staleKey := firewall.RuleID("stale-rule-that-does-not-exist")
r.rules[staleKey] = &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingFw],
@@ -902,6 +941,54 @@ func TestRouter_RefreshRulesMap_RemovesStaleEntries(t *testing.T) {
assert.NotZero(t, realRule.Handle, "real rule should have a valid handle")
}
// TestRouter_DeleteRouteRule_RemovesKernelRule verifies a route filter
// rule is actually removed from the kernel on delete. The route chain is
// not refreshed by Flush, so the stored rule carries a zero handle;
// DeleteFilterRule must pull live handles itself before issuing the
// delete or the kernel rule leaks. Regression test for that path.
func TestRouter_DeleteRouteRule_RemovesKernelRule(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
}
workTable, err := createWorkTable()
require.NoError(t, err)
defer deleteWorkTable()
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() { require.NoError(t, r.Reset()) }()
ruleKey, err := r.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("192.168.1.0/24")},
firewall.Network{Prefix: netip.MustParsePrefix("10.0.0.0/24")},
firewall.ProtocolTCP,
nil,
&firewall.Port{Values: []uint16{80}},
firewall.ActionAccept,
)
require.NoError(t, err)
countKernelRules := func() int {
list, err := r.conn.GetRules(r.workTable, r.chains[chainNameRoutingFw])
require.NoError(t, err)
n := 0
for _, rule := range list {
if string(rule.UserData) == string(ruleKey.ID()) {
n++
}
}
return n
}
require.Equal(t, 1, countKernelRules(), "rule should be present in the kernel after add")
require.NoError(t, r.DeleteFilterRule(ruleKey))
assert.Equal(t, 0, countKernelRules(), "rule must be removed from the kernel after delete")
assert.NotContains(t, r.filters, ruleKey.ID(), "filters map entry should be cleared")
}
func TestRouter_DeleteRouteRule_StaleHandle(t *testing.T) {
if check() != NFTABLES {
t.Skip("nftables not supported on this system")
@@ -911,24 +998,27 @@ func TestRouter_DeleteRouteRule_StaleHandle(t *testing.T) {
require.NoError(t, err)
defer deleteWorkTable()
r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, err)
r := newFamily(workTable, ifaceMock, iface.DefaultMTU)
require.NoError(t, r.init(workTable))
defer func() { require.NoError(t, r.Reset()) }()
// Inject a stale entry with Handle=0
staleKey := "stale-route-rule"
r.rules[staleKey] = &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingFw],
Handle: 0,
UserData: []byte(staleKey),
staleKey := id.RuleID("stale-route-rule")
staleRule := &Rule{
nftRule: &nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingFw],
Handle: 0,
UserData: []byte(staleKey),
},
id: staleKey,
}
r.filters[staleKey] = staleRule
// DeleteRouteRule should not return an error for stale handles
err = r.DeleteRouteRule(id.RuleID(staleKey))
// DeleteFilterRule should not return an error for stale handles
err = r.DeleteFilterRule(staleRule)
assert.NoError(t, err, "deleting a stale rule should not error")
assert.NotContains(t, r.rules, staleKey, "stale entry should be cleaned up")
assert.NotContains(t, r.filters, staleKey, "stale entry should be cleaned up")
}
func TestRouter_AddNatRule_WithStaleEntry(t *testing.T) {
@@ -950,7 +1040,7 @@ func TestRouter_AddNatRule_WithStaleEntry(t *testing.T) {
Masquerade: true,
}
rtr := manager.router
rtr := manager.family4
// First add succeeds
err = rtr.AddNatRule(pair)
@@ -960,11 +1050,11 @@ func TestRouter_AddNatRule_WithStaleEntry(t *testing.T) {
})
// Corrupt the handle to simulate stale state
natRuleKey := firewall.GenKey(firewall.PreroutingFormat, pair)
natRuleKey := pair.GenKey(firewall.PreroutingFormat)
if rule, exists := rtr.rules[natRuleKey]; exists {
rule.Handle = 0
}
inverseKey := firewall.GenKey(firewall.PreroutingFormat, firewall.GetInversePair(pair))
inverseKey := firewall.GetInversePair(pair).GenKey(firewall.PreroutingFormat)
if rule, exists := rtr.rules[inverseKey]; exists {
rule.Handle = 0
}
@@ -979,7 +1069,7 @@ func TestRouter_AddNatRule_WithStaleEntry(t *testing.T) {
found := 0
for _, rule := range rules {
if len(rule.UserData) > 0 && string(rule.UserData) == natRuleKey {
if len(rule.UserData) > 0 && firewall.RuleID(rule.UserData) == natRuleKey {
found++
}
}
@@ -1010,7 +1100,7 @@ func TestCalculateLastIP(t *testing.T) {
}
func TestConvertPrefixesToSet_IPv6(t *testing.T) {
r := &router{af: afIPv6}
r := &family{af: afIPv6}
prefixes := []netip.Prefix{
netip.MustParsePrefix("fd00::/64"),
netip.MustParsePrefix("2001:db8::1/128"),

View File

@@ -0,0 +1,505 @@
//go:build !android
package nftables
import (
"fmt"
"strings"
"github.com/google/nftables"
"github.com/google/nftables/binaryutil"
"github.com/google/nftables/expr"
"github.com/hashicorp/go-multierror"
log "github.com/sirupsen/logrus"
"golang.org/x/sys/unix"
nberrors "github.com/netbirdio/netbird/client/errors"
firewall "github.com/netbirdio/netbird/client/firewall/manager"
nbnet "github.com/netbirdio/netbird/client/net"
)
func (r *family) AddNatRule(pair firewall.RouterPair) error {
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
if r.legacyManagement {
log.Warnf("This peer is connected to a NetBird Management service with an older version. Allowing all traffic for %s", pair.Destination)
if err := r.addLegacyRouteRule(pair); err != nil {
r.rollbackRules(pair)
return fmt.Errorf("add legacy routing rule: %w", err)
}
}
if pair.Masquerade {
if err := r.addNatRule(pair); err != nil {
r.rollbackRules(pair)
return fmt.Errorf("add nat rule: %w", err)
}
if err := r.addNatRule(firewall.GetInversePair(pair)); err != nil {
r.rollbackRules(pair)
return fmt.Errorf("add inverse nat rule: %w", err)
}
}
if err := r.conn.Flush(); err != nil {
r.rollbackRules(pair)
return fmt.Errorf("insert rules for %s: %w", pair.Destination, err)
}
return nil
}
// rollbackRules cleans up unflushed rules and their set counters after a flush failure.
func (r *family) rollbackRules(pair firewall.RouterPair) {
keys := []firewall.RuleID{
pair.GenKey(firewall.ForwardingFormat),
pair.GenKey(firewall.PreroutingFormat),
firewall.GetInversePair(pair).GenKey(firewall.PreroutingFormat),
}
for _, key := range keys {
rule, ok := r.rules[key]
if !ok {
continue
}
if err := r.decrementSetCounter(rule); err != nil {
log.Warnf("rollback set counter for %s: %v", key, err)
}
delete(r.rules, key)
}
}
// addNatRule inserts a nftables rule to the conn client flush queue
func (r *family) addNatRule(pair firewall.RouterPair) error {
sourceExp, err := r.applyNetwork(pair.Source, nil, true)
if err != nil {
return fmt.Errorf("apply source: %w", err)
}
destExp, err := r.applyNetwork(pair.Destination, nil, false)
if err != nil {
r.dropNetworkMatch(sourceExp)
return fmt.Errorf("apply destination: %w", err)
}
op := expr.CmpOpEq
if pair.Inverse {
op = expr.CmpOpNeq
}
exprs := []expr.Any{
&expr.Meta{
Key: expr.MetaKeyIIFNAME,
Register: 1,
},
&expr.Cmp{
Op: op,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
}
// We only care about NEW connections to mark them and later identify them in the postrouting chain for masquerading.
// Masquerading will take care of the conntrack state, which means we won't need to mark established connections.
exprs = append(exprs, getCtNewExprs()...)
exprs = append(exprs, sourceExp...)
exprs = append(exprs, destExp...)
var markValue uint32 = nbnet.PreroutingFwmarkMasquerade
if pair.Inverse {
markValue = nbnet.PreroutingFwmarkMasqueradeReturn
}
exprs = append(exprs,
&expr.Immediate{
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(markValue),
},
&expr.Meta{
Key: expr.MetaKeyMARK,
SourceRegister: true,
Register: 1,
},
)
ruleID := pair.GenKey(firewall.PreroutingFormat)
if _, exists := r.rules[ruleID]; exists {
if err := r.removeNatRule(pair); err != nil {
r.dropNetworkMatch(sourceExp)
r.dropNetworkMatch(destExp)
return fmt.Errorf("remove prerouting rule: %w", err)
}
}
// Ensure nat rules come first, so the mark can be overwritten.
// Currently overwritten by the dst-type LOCAL rules for redirected traffic.
r.rules[ruleID] = r.conn.InsertRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameManglePrerouting],
Exprs: exprs,
UserData: []byte(ruleID),
})
return nil
}
func (r *family) addPostroutingRules() {
// First masquerade rule for traffic coming in from WireGuard interface
exprs := []expr.Any{
// Match on the first fwmark
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkMasquerade),
},
// We need to exclude the loopback interface as this changes the ebpf proxy port
&expr.Meta{
Key: expr.MetaKeyOIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpNeq,
Register: 1,
Data: ifname("lo"),
},
&expr.Counter{},
&expr.Masq{},
}
r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingNat],
Exprs: exprs,
})
// Second masquerade rule for traffic going out through WireGuard interface
exprs2 := []expr.Any{
// Match on the second fwmark
&expr.Meta{
Key: expr.MetaKeyMARK,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: binaryutil.NativeEndian.PutUint32(nbnet.PreroutingFwmarkMasqueradeReturn),
},
// Match WireGuard interface
&expr.Meta{
Key: expr.MetaKeyOIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Counter{},
&expr.Masq{},
}
r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingNat],
Exprs: exprs2,
})
}
// addMSSClampingRules adds MSS clamping rules to prevent fragmentation for forwarded traffic.
func (r *family) addMSSClampingRules() error {
overhead := uint16(ipv4TCPHeaderSize)
if r.af.tableFamily == nftables.TableFamilyIPv6 {
overhead = ipv6TCPHeaderSize
}
if r.mtu <= overhead {
log.Debugf("MTU %d too small for MSS clamping (overhead %d), skipping", r.mtu, overhead)
return nil
}
mss := r.mtu - overhead
exprsOut := []expr.Any{
&expr.Meta{
Key: expr.MetaKeyOIFNAME,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: ifname(r.wgIface.Name()),
},
&expr.Meta{
Key: expr.MetaKeyL4PROTO,
Register: 1,
},
&expr.Cmp{
Op: expr.CmpOpEq,
Register: 1,
Data: []byte{unix.IPPROTO_TCP},
},
&expr.Payload{
DestRegister: 1,
Base: expr.PayloadBaseTransportHeader,
Offset: 13,
Len: 1,
},
&expr.Bitwise{
DestRegister: 1,
SourceRegister: 1,
Len: 1,
Mask: []byte{0x02},
Xor: []byte{0x00},
},
&expr.Cmp{
Op: expr.CmpOpNeq,
Register: 1,
Data: []byte{0x00},
},
&expr.Counter{},
&expr.Exthdr{
DestRegister: 1,
Type: 2,
Offset: 2,
Len: 2,
Op: expr.ExthdrOpTcpopt,
},
&expr.Cmp{
Op: expr.CmpOpGt,
Register: 1,
Data: binaryutil.BigEndian.PutUint16(uint16(mss)),
},
&expr.Immediate{
Register: 1,
Data: binaryutil.BigEndian.PutUint16(uint16(mss)),
},
&expr.Exthdr{
SourceRegister: 1,
Type: 2,
Offset: 2,
Len: 2,
Op: expr.ExthdrOpTcpopt,
},
}
r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameMangleForward],
Exprs: exprsOut,
})
return r.conn.Flush()
}
func buildLegacyRouteRuleExpressions(sourceExp, destExp []expr.Any) []expr.Any {
exprs := make([]expr.Any, 0, len(sourceExp)+len(destExp)+2)
exprs = append(exprs, sourceExp...)
exprs = append(exprs, destExp...)
exprs = append(exprs,
&expr.Counter{},
&expr.Verdict{Kind: expr.VerdictAccept},
)
return exprs
}
func (r *family) addLegacyRouteRule(pair firewall.RouterPair) error {
sourceExp, err := r.applyNetwork(pair.Source, nil, true)
if err != nil {
return fmt.Errorf("apply source: %w", err)
}
destExp, err := r.applyNetwork(pair.Destination, nil, false)
if err != nil {
r.dropNetworkMatch(sourceExp)
return fmt.Errorf("apply destination: %w", err)
}
exprs := buildLegacyRouteRuleExpressions(sourceExp, destExp)
ruleID := pair.GenKey(firewall.ForwardingFormat)
if _, exists := r.rules[ruleID]; exists {
if err := r.removeLegacyRouteRule(pair); err != nil {
r.dropNetworkMatch(sourceExp)
r.dropNetworkMatch(destExp)
return fmt.Errorf("remove legacy routing rule: %w", err)
}
}
r.rules[ruleID] = r.conn.AddRule(&nftables.Rule{
Table: r.workTable,
Chain: r.chains[chainNameRoutingFw],
Exprs: exprs,
UserData: []byte(ruleID),
})
return nil
}
// removeLegacyRouteRule removes a legacy routing rule for mgmt servers pre route acls
func (r *family) removeLegacyRouteRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.ForwardingFormat)
rule, exists := r.rules[ruleID]
if !exists {
return nil
}
return r.deleteLegacyRuleEntry(ruleID, rule)
}
// deleteLegacyRuleEntry removes one legacy forwarding rule and drops its
// ipset references. It also clears stale entries that never got a handle.
func (r *family) deleteLegacyRuleEntry(ruleID firewall.RuleID, rule *nftables.Rule) error {
if rule.Handle == 0 {
log.Warnf("legacy forwarding rule %s has no handle, removing stale entry", ruleID)
if err := r.decrementSetCounter(rule); err != nil {
log.Warnf("decrement set counter for stale rule %s: %v", ruleID, err)
}
delete(r.rules, ruleID)
return nil
}
if err := r.conn.DelRule(rule); err != nil {
return fmt.Errorf("remove legacy forwarding rule %s: %w", ruleID, err)
}
delete(r.rules, ruleID)
if err := r.decrementSetCounter(rule); err != nil {
return fmt.Errorf("decrement set counter: %w", err)
}
return nil
}
// GetLegacyManagement returns the route manager's legacy management mode
func (r *family) GetLegacyManagement() bool {
return r.legacyManagement
}
// SetLegacyManagement sets the route manager to use legacy management mode
func (r *family) SetLegacyManagement(isLegacy bool) {
r.legacyManagement = isLegacy
}
// RemoveAllLegacyRouteRules removes all legacy routing rules for mgmt servers pre route acls
func (r *family) RemoveAllLegacyRouteRules() error {
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
var merr *multierror.Error
for k, rule := range r.rules {
if !strings.HasPrefix(string(k), firewall.ForwardingFormatPrefix) {
continue
}
if err := r.deleteLegacyRuleEntry(k, rule); err != nil {
merr = multierror.Append(merr, err)
}
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) removeNatPreroutingRules() error {
table := &nftables.Table{
Name: tableNat,
Family: r.af.tableFamily,
}
chain := &nftables.Chain{
Name: chainNameNatPrerouting,
Table: table,
Hooknum: nftables.ChainHookPrerouting,
Priority: nftables.ChainPriorityNATDest,
Type: nftables.ChainTypeNAT,
}
rules, err := r.conn.GetRules(table, chain)
if err != nil {
return fmt.Errorf("get rules from nat table: %w", err)
}
var merr *multierror.Error
// Delete rules that have our UserData suffix
for _, rule := range rules {
if len(rule.UserData) == 0 || !strings.HasSuffix(string(rule.UserData), string(dnatSuffix)) {
continue
}
if err := r.conn.DelRule(rule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("delete rule %s: %w", rule.UserData, err))
}
}
if err := r.conn.Flush(); err != nil {
merr = multierror.Append(merr, fmt.Errorf(flushError, err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) RemoveNatRule(pair firewall.RouterPair) error {
if err := r.refreshRulesMap(); err != nil {
return fmt.Errorf(refreshRulesMapError, err)
}
var merr *multierror.Error
if pair.Masquerade {
if err := r.removeNatRule(pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove prerouting rule: %w", err))
}
if err := r.removeNatRule(firewall.GetInversePair(pair)); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove inverse prerouting rule: %w", err))
}
}
if err := r.removeLegacyRouteRule(pair); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove legacy routing rule: %w", err))
}
// Set counters are decremented in the sub-methods above before flush. If flush fails,
// counters will be off until the next successful removal or refresh cycle.
if err := r.conn.Flush(); err != nil {
merr = multierror.Append(merr, fmt.Errorf("flush remove nat rules %s: %w", pair.Destination, err))
}
return nberrors.FormatErrorOrNil(merr)
}
func (r *family) removeNatRule(pair firewall.RouterPair) error {
ruleID := pair.GenKey(firewall.PreroutingFormat)
rule, exists := r.rules[ruleID]
if !exists {
log.Debugf("prerouting rule %s not found", ruleID)
return nil
}
if rule.Handle == 0 {
log.Warnf("prerouting rule %s has no handle, removing stale entry", ruleID)
if err := r.decrementSetCounter(rule); err != nil {
log.Warnf("decrement set counter for stale rule %s: %v", ruleID, err)
}
delete(r.rules, ruleID)
return nil
}
if err := r.conn.DelRule(rule); err != nil {
return fmt.Errorf("remove prerouting rule %s -> %s: %w", pair.Source, pair.Destination, err)
}
log.Debugf("removed prerouting rule %s -> %s", pair.Source, pair.Destination)
delete(r.rules, ruleID)
if err := r.decrementSetCounter(rule); err != nil {
return fmt.Errorf("decrement set counter: %w", err)
}
return nil
}

View File

@@ -1,21 +1,26 @@
package nftables
import (
"net"
"net/netip"
"github.com/google/nftables"
"github.com/netbirdio/netbird/client/firewall/manager"
)
// Rule to handle management of rules
// Rule wraps an installed filter rule (peer or route). Source set
// membership is encoded in the rule's expressions; DeleteFilterRule
// recovers the set name via findSets so the refcounter can drop the
// right reference. mangleRule is set only for peer rules.
type Rule struct {
nftRule *nftables.Rule
mangleRule *nftables.Rule
nftSet *nftables.Set
ruleID string
ip net.IP
// sources is the canonical source list this rule was created for.
sources []netip.Prefix
id manager.RuleID
}
// GetRuleID returns the rule id
func (r *Rule) ID() string {
return r.ruleID
// ID returns the rule id
func (r *Rule) ID() manager.RuleID {
return r.id
}

View File

@@ -0,0 +1,27 @@
//go:build privileged
package nftables
import (
"fmt"
"net"
"net/netip"
)
func pfx(ip net.IP) []netip.Prefix {
if ip == nil {
return []netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)}
}
if ip.IsUnspecified() {
if ip.To4() != nil {
return []netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)}
}
return []netip.Prefix{netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
}
a, ok := netip.AddrFromSlice(ip)
if !ok {
panic(fmt.Sprintf("invalid IP length: %d", len(ip)))
}
a = a.Unmap()
return []netip.Prefix{netip.PrefixFrom(a, a.BitLen())}
}

View File

@@ -1,37 +0,0 @@
//go:build !windows
package uspfilter
import (
log "github.com/sirupsen/logrus"
"github.com/netbirdio/netbird/client/firewall/firewalld"
"github.com/netbirdio/netbird/client/internal/statemanager"
)
// Close cleans up the firewall manager by removing all rules and closing trackers
func (m *Manager) Close(stateManager *statemanager.Manager) error {
m.mutex.Lock()
defer m.mutex.Unlock()
m.resetState()
if m.nativeFirewall != nil {
return m.nativeFirewall.Close(stateManager)
}
if err := firewalld.UntrustInterface(m.wgIface.Name()); err != nil {
log.Warnf("failed to untrust interface in firewalld: %v", err)
}
return nil
}
// AllowNetbird allows netbird interface traffic
func (m *Manager) AllowNetbird() error {
if m.nativeFirewall != nil {
return m.nativeFirewall.AllowNetbird()
}
if err := firewalld.TrustInterface(m.wgIface.Name()); err != nil {
log.Warnf("failed to trust interface in firewalld: %v", err)
}
return nil
}

View File

@@ -1,17 +0,0 @@
package common
import (
wgdevice "golang.zx2c4.com/wireguard/device"
"github.com/netbirdio/netbird/client/iface/device"
"github.com/netbirdio/netbird/client/iface/wgaddr"
)
// IFaceMapper defines subset methods of interface required for manager
type IFaceMapper interface {
Name() string
SetFilter(device.PacketFilter) error
Address() wgaddr.Address
GetWGDevice() *wgdevice.Device
GetDevice() *device.FilteredDevice
}

File diff suppressed because it is too large Load Diff

View File

@@ -94,7 +94,7 @@ func BenchmarkCoreFiltering(b *testing.B) {
stateful: false,
setupFunc: func(m *Manager) {
// Single rule allowing all traffic
_, err := m.AddPeerFiltering(nil, net.ParseIP("0.0.0.0"), fw.ProtocolALL, nil, nil, fw.ActionAccept, "")
_, err := m.AddFilterRule(nil, pfx(net.ParseIP("0.0.0.0")), fw.Network{}, fw.ProtocolALL, nil, nil, fw.ActionAccept)
require.NoError(b, err)
},
desc: "Baseline: Single 'allow all' rule without connection tracking",
@@ -114,15 +114,13 @@ func BenchmarkCoreFiltering(b *testing.B) {
// Add explicit rules matching return traffic pattern
for i := 0; i < 1000; i++ { // Simulate realistic ruleset size
ip := generateRandomIPs(1)[0]
_, err := m.AddPeerFiltering(
_, err := m.AddFilterRule(
nil,
ip,
pfx(ip), fw.Network{},
fw.ProtocolTCP,
&fw.Port{Values: []uint16{uint16(1024 + i)}},
&fw.Port{Values: []uint16{80}},
fw.ActionAccept,
"",
)
fw.ActionAccept)
require.NoError(b, err)
}
},
@@ -133,15 +131,13 @@ func BenchmarkCoreFiltering(b *testing.B) {
stateful: true,
setupFunc: func(m *Manager) {
// Add some basic rules but rely on state for established connections
_, err := m.AddPeerFiltering(
_, err := m.AddFilterRule(
nil,
net.ParseIP("0.0.0.0"),
pfx(net.ParseIP("0.0.0.0")), fw.Network{},
fw.ProtocolTCP,
nil,
nil,
fw.ActionDrop,
"",
)
fw.ActionDrop)
require.NoError(b, err)
},
desc: "Connection tracking with established connections",
@@ -168,9 +164,12 @@ func BenchmarkCoreFiltering(b *testing.B) {
}
// Create manager and basic setup
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -208,9 +207,12 @@ func BenchmarkStateScaling(b *testing.B) {
for _, count := range connCounts {
b.Run(fmt.Sprintf("conns_%d", count), func(b *testing.B) {
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -251,9 +253,12 @@ func BenchmarkEstablishmentOverhead(b *testing.B) {
for _, sc := range scenarios {
b.Run(sc.name, func(b *testing.B) {
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -409,9 +414,12 @@ func BenchmarkRoutedNetworkReturn(b *testing.B) {
for _, sc := range scenarios {
b.Run(sc.name, func(b *testing.B) {
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -536,9 +544,12 @@ func BenchmarkLongLivedConnections(b *testing.B) {
require.NoError(b, os.Unsetenv("NB_DISABLE_CONNTRACK"))
}
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -546,7 +557,7 @@ func BenchmarkLongLivedConnections(b *testing.B) {
// Setup initial state based on scenario
if sc.rules {
// Single rule to allow all return traffic from port 80
_, err := manager.AddPeerFiltering(nil, net.ParseIP("0.0.0.0"), fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept, "")
_, err := manager.AddFilterRule(nil, pfx(net.ParseIP("0.0.0.0")), fw.Network{}, fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept)
require.NoError(b, err)
}
@@ -619,9 +630,12 @@ func BenchmarkShortLivedConnections(b *testing.B) {
require.NoError(b, os.Unsetenv("NB_DISABLE_CONNTRACK"))
}
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
@@ -629,7 +643,7 @@ func BenchmarkShortLivedConnections(b *testing.B) {
// Setup initial state based on scenario
if sc.rules {
// Single rule to allow all return traffic from port 80
_, err := manager.AddPeerFiltering(nil, net.ParseIP("0.0.0.0"), fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept, "")
_, err := manager.AddFilterRule(nil, pfx(net.ParseIP("0.0.0.0")), fw.Network{}, fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept)
require.NoError(b, err)
}
@@ -730,16 +744,19 @@ func BenchmarkParallelLongLivedConnections(b *testing.B) {
require.NoError(b, os.Unsetenv("NB_DISABLE_CONNTRACK"))
}
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
// Setup initial state based on scenario
if sc.rules {
_, err := manager.AddPeerFiltering(nil, net.ParseIP("0.0.0.0"), fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept, "")
_, err := manager.AddFilterRule(nil, pfx(net.ParseIP("0.0.0.0")), fw.Network{}, fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept)
require.NoError(b, err)
}
@@ -810,15 +827,18 @@ func BenchmarkParallelShortLivedConnections(b *testing.B) {
require.NoError(b, os.Unsetenv("NB_DISABLE_CONNTRACK"))
}
manager, _ := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer b.Cleanup(func() {
require.NoError(b, manager.Close(nil))
})
if sc.rules {
_, err := manager.AddPeerFiltering(nil, net.ParseIP("0.0.0.0"), fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept, "")
_, err := manager.AddFilterRule(nil, pfx(net.ParseIP("0.0.0.0")), fw.Network{}, fw.ProtocolTCP, &fw.Port{Values: []uint16{80}}, nil, fw.ActionAccept)
require.NoError(b, err)
}
@@ -931,7 +951,7 @@ func BenchmarkRouteACLs(b *testing.B) {
for _, r := range rules {
dst := fw.Network{Prefix: r.dest}
_, err := manager.AddRouteFiltering(nil, r.sources, dst, r.proto, nil, r.port, fw.ActionAccept)
_, err := manager.AddFilterRule(nil, r.sources, dst, r.proto, nil, r.port, fw.ActionAccept)
if err != nil {
b.Fatal(err)
}
@@ -1014,9 +1034,11 @@ func BenchmarkMSSClamping(b *testing.B) {
for _, sc := range scenarios {
b.Run(sc.name, func(b *testing.B) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer func() {
require.NoError(b, manager.Close(nil))
@@ -1079,9 +1101,11 @@ func BenchmarkMSSClampingOverhead(b *testing.B) {
for _, sc := range scenarios {
b.Run(sc.name, func(b *testing.B) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer func() {
require.NoError(b, manager.Close(nil))
@@ -1134,9 +1158,11 @@ func BenchmarkMSSClampingMemory(b *testing.B) {
for _, sc := range scenarios {
b.Run(sc.name, func(b *testing.B) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(b, err)
defer func() {
require.NoError(b, manager.Close(nil))

View File

@@ -32,7 +32,7 @@ func TestPeerACLFiltering(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
require.NotNil(t, manager)
@@ -496,40 +496,32 @@ func TestPeerACLFiltering(t *testing.T) {
t.Run(tc.name, func(t *testing.T) {
if tc.ruleAction == fw.ActionDrop {
// add general accept rule for the same IP to test drop rule precedence
rules, err := manager.AddPeerFiltering(
rules, err := manager.AddFilterRule(
nil,
net.ParseIP(tc.ruleIP),
pfx(net.ParseIP(tc.ruleIP)), fw.Network{},
fw.ProtocolALL,
nil,
nil,
fw.ActionAccept,
"",
)
fw.ActionAccept)
require.NoError(t, err)
require.NotEmpty(t, rules)
require.NotNil(t, rules)
t.Cleanup(func() {
for _, rule := range rules {
require.NoError(t, manager.DeletePeerRule(rule))
}
require.NoError(t, manager.DeleteFilterRule(rules))
})
}
rules, err := manager.AddPeerFiltering(
rules, err := manager.AddFilterRule(
nil,
net.ParseIP(tc.ruleIP),
pfx(net.ParseIP(tc.ruleIP)), fw.Network{},
tc.ruleProto,
tc.ruleSrcPort,
tc.ruleDstPort,
tc.ruleAction,
"",
)
tc.ruleAction)
require.NoError(t, err)
require.NotEmpty(t, rules)
require.NotNil(t, rules)
t.Cleanup(func() {
for _, rule := range rules {
require.NoError(t, manager.DeletePeerRule(rule))
}
require.NoError(t, manager.DeleteFilterRule(rules))
})
packet := createTestPacket(t, tc.srcIP, tc.dstIP, tc.proto, tc.srcPort, tc.dstPort)
@@ -557,7 +549,7 @@ func TestPeerACLFilteringIPv6(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
@@ -652,14 +644,24 @@ func TestPeerACLFilteringIPv6(t *testing.T) {
shouldBeBlocked: false,
},
{
name: "IPv6: v4 wildcard ICMP rule matches ICMPv6 via protoLayerMatches",
name: "IPv6: v4 wildcard ICMP rule does not match ICMPv6",
srcIP: "fd00::1",
dstIP: "fd00::100",
proto: fw.ProtocolICMP,
ruleIP: "0.0.0.0",
ruleProto: fw.ProtocolICMP,
ruleAction: fw.ActionAccept,
shouldBeBlocked: false,
shouldBeBlocked: true,
},
{
name: "IPv4: v6 wildcard ICMP rule does not match ICMPv4",
srcIP: "100.10.0.1",
dstIP: "100.10.0.100",
proto: fw.ProtocolICMP,
ruleIP: "::",
ruleProto: fw.ProtocolICMP,
ruleAction: fw.ActionAccept,
shouldBeBlocked: true,
},
}
@@ -672,22 +674,18 @@ func TestPeerACLFilteringIPv6(t *testing.T) {
for _, tc := range testCases {
t.Run(tc.name, func(t *testing.T) {
if tc.ruleAction == fw.ActionDrop {
rules, err := manager.AddPeerFiltering(nil, net.ParseIP(tc.ruleIP), fw.ProtocolALL, nil, nil, fw.ActionAccept, "")
rules, err := manager.AddFilterRule(nil, pfx(net.ParseIP(tc.ruleIP)), fw.Network{}, fw.ProtocolALL, nil, nil, fw.ActionAccept)
require.NoError(t, err)
t.Cleanup(func() {
for _, rule := range rules {
require.NoError(t, manager.DeletePeerRule(rule))
}
require.NoError(t, manager.DeleteFilterRule(rules))
})
}
rules, err := manager.AddPeerFiltering(nil, net.ParseIP(tc.ruleIP), tc.ruleProto, nil, tc.ruleDstPort, tc.ruleAction, "")
rules, err := manager.AddFilterRule(nil, pfx(net.ParseIP(tc.ruleIP)), fw.Network{}, tc.ruleProto, nil, tc.ruleDstPort, tc.ruleAction)
require.NoError(t, err)
require.NotEmpty(t, rules)
require.NotNil(t, rules)
t.Cleanup(func() {
for _, rule := range rules {
require.NoError(t, manager.DeletePeerRule(rule))
}
require.NoError(t, manager.DeleteFilterRule(rules))
})
packet := createTestPacket(t, tc.srcIP, tc.dstIP, tc.proto, tc.srcPort, tc.dstPort)
@@ -800,7 +798,7 @@ func setupRoutedManager(tb testing.TB, network string) *Manager {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(tb, err)
require.NoError(tb, manager.EnableRouting())
require.NotNil(tb, manager)
@@ -1405,7 +1403,7 @@ func TestRouteACLFiltering(t *testing.T) {
t.Run(tc.name, func(t *testing.T) {
if tc.rule.action == fw.ActionDrop {
// add general accept rule to test drop rule
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("0.0.0.0/0")},
fw.Network{Prefix: netip.MustParsePrefix("0.0.0.0/0")},
@@ -1415,13 +1413,13 @@ func TestRouteACLFiltering(t *testing.T) {
fw.ActionAccept,
)
require.NoError(t, err)
require.NotNil(t, rule)
require.NotEmpty(t, rule)
t.Cleanup(func() {
require.NoError(t, manager.DeleteRouteRule(rule))
require.NoError(t, manager.DeleteFilterRule(rule))
})
}
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
tc.rule.sources,
tc.rule.dest,
@@ -1431,10 +1429,10 @@ func TestRouteACLFiltering(t *testing.T) {
tc.rule.action,
)
require.NoError(t, err)
require.NotNil(t, rule)
require.NotEmpty(t, rule)
t.Cleanup(func() {
require.NoError(t, manager.DeleteRouteRule(rule))
require.NoError(t, manager.DeleteFilterRule(rule))
})
srcIP := netip.MustParseAddr(tc.srcIP)
@@ -1602,9 +1600,9 @@ func TestRouteACLOrder(t *testing.T) {
for _, tc := range testCases {
t.Run(tc.name, func(t *testing.T) {
var rules []fw.Rule
var addedRules []fw.Rule
for _, r := range tc.rules {
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
r.sources,
r.dest,
@@ -1615,12 +1613,12 @@ func TestRouteACLOrder(t *testing.T) {
)
require.NoError(t, err)
require.NotNil(t, rule)
rules = append(rules, rule)
addedRules = append(addedRules, rule)
}
t.Cleanup(func() {
for _, rule := range rules {
require.NoError(t, manager.DeleteRouteRule(rule))
for _, rule := range addedRules {
require.NoError(t, manager.DeleteFilterRule(rule))
}
})
@@ -1646,7 +1644,7 @@ func TestRouteACLSet(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -1655,7 +1653,7 @@ func TestRouteACLSet(t *testing.T) {
set := fw.NewDomainSet(domain.List{"example.org"})
// Add rule that uses the set (initially empty)
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("0.0.0.0/0")},
fw.Network{Set: set},
@@ -1689,7 +1687,7 @@ func TestRouteACLFilteringIPv6(t *testing.T) {
manager := setupRoutedManager(t, "10.10.0.100/16")
v6Dst := netip.MustParsePrefix("fd00:dead:beef::/48")
_, err := manager.AddRouteFiltering(
_, err := manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("fd00::/16")},
fw.Network{Prefix: v6Dst},
@@ -1700,7 +1698,7 @@ func TestRouteACLFilteringIPv6(t *testing.T) {
)
require.NoError(t, err)
_, err = manager.AddRouteFiltering(
_, err = manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("fd00::/16")},
fw.Network{Prefix: netip.MustParsePrefix("fd00:dead:beef:1::/64")},

View File

@@ -29,7 +29,7 @@ func TestAddRouteFilteringReturnsExistingRule(t *testing.T) {
destination := fw.Network{Prefix: netip.MustParsePrefix("192.168.1.0/24")}
// Add rule first time
rule1, err := manager.AddRouteFiltering(
rule1, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -42,7 +42,7 @@ func TestAddRouteFilteringReturnsExistingRule(t *testing.T) {
require.NotNil(t, rule1)
// Add the same rule again
rule2, err := manager.AddRouteFiltering(
rule2, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -74,7 +74,7 @@ func TestAddRouteFilteringDifferentRulesGetDifferentIDs(t *testing.T) {
sources := []netip.Prefix{netip.MustParsePrefix("100.64.1.0/24")}
// Add first rule
rule1, err := manager.AddRouteFiltering(
rule1, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
fw.Network{Prefix: netip.MustParsePrefix("192.168.1.0/24")},
@@ -86,7 +86,7 @@ func TestAddRouteFilteringDifferentRulesGetDifferentIDs(t *testing.T) {
require.NoError(t, err)
// Add different rule (different destination)
rule2, err := manager.AddRouteFiltering(
rule2, err := manager.AddFilterRule(
[]byte("policy-2"),
sources,
fw.Network{Prefix: netip.MustParsePrefix("192.168.2.0/24")}, // Different!
@@ -115,7 +115,7 @@ func TestRouteRuleUpdateDoesNotCauseGap(t *testing.T) {
sources := []netip.Prefix{netip.MustParsePrefix("100.64.1.0/24")}
destination := fw.Network{Prefix: netip.MustParsePrefix("192.168.1.0/24")}
rule1, err := manager.AddRouteFiltering(
rule1, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -132,7 +132,7 @@ func TestRouteRuleUpdateDoesNotCauseGap(t *testing.T) {
require.True(t, pass, "Traffic should pass with rule in place")
// Re-add same rule (simulates network map update)
rule2, err := manager.AddRouteFiltering(
rule2, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -147,7 +147,7 @@ func TestRouteRuleUpdateDoesNotCauseGap(t *testing.T) {
// won't delete rule1 during cleanup. If IDs differed, deleting rule1
// would remove the only matching rule and cause a traffic gap.
if rule1.ID() != rule2.ID() {
err = manager.DeleteRouteRule(rule1)
err = manager.DeleteFilterRule(rule1)
require.NoError(t, err)
}
@@ -156,6 +156,59 @@ func TestRouteRuleUpdateDoesNotCauseGap(t *testing.T) {
"Traffic should still pass after rule update - no gap should occur")
}
// TestBlockInvalidRoutedDualStack verifies that when the interface has an
// IPv6 overlay address, blockInvalidRouted installs a block rule for both
// the v4 and v6 WG prefixes and that routed traffic to the v6 prefix is
// denied. The v4-only soft-skip path is covered by
// TestBlockInvalidRoutedIdempotent, whose mock leaves IPv6Net invalid.
func TestBlockInvalidRoutedDualStack(t *testing.T) {
ctrl := gomock.NewController(t)
dev := mocks.NewMockDevice(ctrl)
dev.EXPECT().MTU().Return(1500, nil).AnyTimes()
wgNet := netip.MustParsePrefix("100.64.0.1/16")
wgNet6 := netip.MustParsePrefix("fd00:1234::1/64")
ifaceMock := &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
AddressFunc: func() wgaddr.Address {
return wgaddr.Address{
IP: wgNet.Addr(),
Network: wgNet,
IPv6: wgNet6.Addr(),
IPv6Net: wgNet6,
}
},
GetDeviceFunc: func() *device.FilteredDevice {
return &device.FilteredDevice{Device: dev}
},
GetWGDeviceFunc: func() *wgdevice.Device {
return &wgdevice.Device{}
},
}
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
})
rules, err := manager.blockInvalidRouted(ifaceMock)
require.NoError(t, err)
require.Len(t, rules, 2, "dual-stack interface must produce a v4 and a v6 block rule")
manager.mutex.RLock()
ruleCount := len(manager.routeRules)
manager.mutex.RUnlock()
assert.Equal(t, 2, ruleCount, "should have one block rule per family")
// v6 routed traffic to the WG prefix must be denied.
srcIP := netip.MustParseAddr("2001:db8::1")
dstIP := netip.MustParseAddr("fd00:1234::50")
_, pass := manager.routeACLsPass(srcIP, dstIP, layers.LayerTypeTCP, 12345, 80)
assert.False(t, pass, "block rule should deny routed traffic to the v6 WG prefix")
}
// TestBlockInvalidRoutedIdempotent verifies that blockInvalidRouted creates
// exactly one drop rule for the WireGuard network prefix, and calling it again
// returns the same rule without duplicating.
@@ -182,7 +235,7 @@ func TestBlockInvalidRoutedIdempotent(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -245,7 +298,7 @@ func TestBlockRuleNotAccumulatedOnRepeatedEnableRouting(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -274,7 +327,7 @@ func TestRouteRuleCountStableAcrossUpdates(t *testing.T) {
// Simulate 5 network map updates with the same route rule
for i := 0; i < 5; i++ {
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -304,7 +357,7 @@ func TestDeleteRouteRuleAfterIdempotentAdd(t *testing.T) {
destination := fw.Network{Prefix: netip.MustParsePrefix("192.168.1.0/24")}
// Add same rule twice
rule1, err := manager.AddRouteFiltering(
rule1, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -315,7 +368,7 @@ func TestDeleteRouteRuleAfterIdempotentAdd(t *testing.T) {
)
require.NoError(t, err)
rule2, err := manager.AddRouteFiltering(
rule2, err := manager.AddFilterRule(
[]byte("policy-1"),
sources,
destination,
@@ -329,7 +382,7 @@ func TestDeleteRouteRuleAfterIdempotentAdd(t *testing.T) {
require.Equal(t, rule1.ID(), rule2.ID(), "Should return same rule ID")
// Delete using first reference
err = manager.DeleteRouteRule(rule1)
err = manager.DeleteFilterRule(rule1)
require.NoError(t, err)
// Verify traffic no longer passes
@@ -364,7 +417,7 @@ func setupTestManager(t *testing.T) *Manager {
},
}
manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: iface.DefaultMTU})
require.NoError(t, err)
require.NoError(t, manager.EnableRouting())

View File

@@ -78,18 +78,19 @@ func TestManagerCreate(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Errorf("failed to create Manager: %v", err)
return
}
t.Cleanup(func() { require.NoError(t, m.Close(nil)) })
if m == nil {
t.Error("Manager is nil")
}
}
func TestManagerAddPeerFiltering(t *testing.T) {
func TestManagerAddFilterRule(t *testing.T) {
isSetFilterCalled := false
ifaceMock := &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error {
@@ -98,18 +99,19 @@ func TestManagerAddPeerFiltering(t *testing.T) {
},
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Errorf("failed to create Manager: %v", err)
return
}
t.Cleanup(func() { require.NoError(t, m.Close(nil)) })
ip := net.ParseIP("192.168.1.1")
proto := fw.ProtocolTCP
port := &fw.Port{Values: []uint16{80}}
action := fw.ActionDrop
rule, err := m.AddPeerFiltering(nil, ip, proto, nil, port, action, "")
rule, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, proto, nil, port, action)
if err != nil {
t.Errorf("failed to add filtering: %v", err)
return
@@ -131,74 +133,47 @@ func TestManagerDeleteRule(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Errorf("failed to create Manager: %v", err)
return
}
t.Cleanup(func() { require.NoError(t, m.Close(nil)) })
ip := netip.MustParseAddr("192.168.1.1")
proto := fw.ProtocolTCP
port := &fw.Port{Values: []uint16{80}}
action := fw.ActionDrop
rule2, err := m.AddPeerFiltering(nil, ip.AsSlice(), proto, nil, port, action, "")
rule2, err := m.AddFilterRule(nil, pfx(ip.AsSlice()), fw.Network{}, proto, nil, port, action)
if err != nil {
t.Errorf("failed to add filtering: %v", err)
return
}
// Check rules exist in appropriate maps
for _, r := range rule2 {
peerRule, ok := r.(*PeerRule)
if !ok {
t.Errorf("rule should be a PeerRule")
continue
}
// Check if rule exists in deny or allow maps based on action
var found bool
if peerRule.drop {
_, found = m.incomingDenyRules[ip][r.ID()]
} else {
_, found = m.incomingRules[ip][r.ID()]
}
if !found {
t.Errorf("rule2 is not in the expected rules map")
peerRule, ok := rule2.(*PeerRule)
require.True(t, ok, "rule should be a peer rule")
inMap := func() bool {
if peerRule.action == fw.ActionDrop {
return findRuleByID(m.incomingDenyRules, ip, rule2.ID())
}
return findRuleByID(m.incomingAcceptRules, ip, rule2.ID())
}
for _, r := range rule2 {
err = m.DeletePeerRule(r)
if err != nil {
t.Errorf("failed to delete rule: %v", err)
return
}
}
require.True(t, inMap(), "rule2 should be in the expected rules list")
// Check rules are removed from appropriate maps
for _, r := range rule2 {
peerRule, ok := r.(*PeerRule)
if !ok {
t.Errorf("rule should be a PeerRule")
continue
}
// Check if rule is removed from deny or allow maps based on action
var found bool
if peerRule.drop {
_, found = m.incomingDenyRules[ip][r.ID()]
} else {
_, found = m.incomingRules[ip][r.ID()]
}
if found {
t.Errorf("rule2 should be removed from the rules map")
}
}
require.NoError(t, m.DeleteFilterRule(rule2), "failed to delete rule")
require.False(t, inMap(), "rule2 should be removed from the rules list")
}
func TestSetUDPPacketHook(t *testing.T) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
@@ -220,9 +195,11 @@ func TestSetUDPPacketHook(t *testing.T) {
}
func TestSetTCPPacketHook(t *testing.T) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
@@ -250,7 +227,7 @@ func TestPeerRuleLifecycleDenyRules(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
defer func() {
require.NoError(t, m.Close(nil))
@@ -260,36 +237,34 @@ func TestPeerRuleLifecycleDenyRules(t *testing.T) {
addr := netip.MustParseAddr("192.168.1.1")
// Add multiple deny rules for different ports
rule1, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{22}}, fw.ActionDrop, "")
rule1, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{22}}, fw.ActionDrop)
require.NoError(t, err)
rule2, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{80}}, fw.ActionDrop, "")
rule2, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionDrop)
require.NoError(t, err)
m.mutex.RLock()
denyCount := len(m.incomingDenyRules[addr])
denyCount := countRulesForAddr(m.incomingDenyRules, addr)
m.mutex.RUnlock()
require.Equal(t, 2, denyCount, "Should have exactly 2 deny rules")
// Delete the first deny rule
err = m.DeletePeerRule(rule1[0])
err = m.DeleteFilterRule(rule1)
require.NoError(t, err)
m.mutex.RLock()
denyCount = len(m.incomingDenyRules[addr])
denyCount = countRulesForAddr(m.incomingDenyRules, addr)
m.mutex.RUnlock()
require.Equal(t, 1, denyCount, "Should have 1 deny rule after deleting first")
// Delete the second deny rule
err = m.DeletePeerRule(rule2[0])
err = m.DeleteFilterRule(rule2)
require.NoError(t, err)
m.mutex.RLock()
_, exists := m.incomingDenyRules[addr]
exists := countRulesForAddr(m.incomingDenyRules, addr) > 0
m.mutex.RUnlock()
require.False(t, exists, "Deny rules IP entry should be cleaned up when empty")
require.False(t, exists, "Deny rules should be cleaned up when empty")
}
// TestPeerRuleAddAndDeleteDontLeak verifies that repeatedly adding and deleting
@@ -299,7 +274,7 @@ func TestPeerRuleAddAndDeleteDontLeak(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
defer func() {
require.NoError(t, m.Close(nil))
@@ -311,27 +286,21 @@ func TestPeerRuleAddAndDeleteDontLeak(t *testing.T) {
// Simulate 10 network map updates: add rule, delete old, add new
for i := 0; i < 10; i++ {
// Add a deny rule
rules, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{22}}, fw.ActionDrop, "")
rules, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{22}}, fw.ActionDrop)
require.NoError(t, err)
// Add an allow rule
allowRules, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{80}}, fw.ActionAccept, "")
allowRules, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept)
require.NoError(t, err)
// Delete them (simulating ACL manager cleanup)
for _, r := range rules {
require.NoError(t, m.DeletePeerRule(r))
}
for _, r := range allowRules {
require.NoError(t, m.DeletePeerRule(r))
}
require.NoError(t, m.DeleteFilterRule(rules))
require.NoError(t, m.DeleteFilterRule(allowRules))
}
m.mutex.RLock()
denyCount := len(m.incomingDenyRules[addr])
allowCount := len(m.incomingRules[addr])
denyCount := countRulesForAddr(m.incomingDenyRules, addr)
allowCount := countRulesForAddr(m.incomingAcceptRules, addr)
m.mutex.RUnlock()
require.Equal(t, 0, denyCount, "No deny rules should remain after cleanup")
@@ -345,7 +314,7 @@ func TestMixedAllowDenyRulesSameIP(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
defer func() {
require.NoError(t, m.Close(nil))
@@ -354,41 +323,39 @@ func TestMixedAllowDenyRulesSameIP(t *testing.T) {
ip := net.ParseIP("192.168.1.1")
// Add allow rule for port 80
allowRule, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{80}}, fw.ActionAccept, "")
allowRule, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept)
require.NoError(t, err)
// Add deny rule for port 22
denyRule, err := m.AddPeerFiltering(nil, ip, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{22}}, fw.ActionDrop, "")
denyRule, err := m.AddFilterRule(nil, pfx(ip), fw.Network{}, fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{22}}, fw.ActionDrop)
require.NoError(t, err)
addr := netip.MustParseAddr("192.168.1.1")
m.mutex.RLock()
allowCount := len(m.incomingRules[addr])
denyCount := len(m.incomingDenyRules[addr])
allowCount := countRulesForAddr(m.incomingAcceptRules, addr)
denyCount := countRulesForAddr(m.incomingDenyRules, addr)
m.mutex.RUnlock()
require.Equal(t, 1, allowCount, "Should have 1 allow rule")
require.Equal(t, 1, denyCount, "Should have 1 deny rule")
// Delete allow rule should not affect deny rule
err = m.DeletePeerRule(allowRule[0])
err = m.DeleteFilterRule(allowRule)
require.NoError(t, err)
m.mutex.RLock()
denyCountAfter := len(m.incomingDenyRules[addr])
denyCountAfter := countRulesForAddr(m.incomingDenyRules, addr)
m.mutex.RUnlock()
require.Equal(t, 1, denyCountAfter, "Deny rule should still exist after deleting allow rule")
// Delete deny rule
err = m.DeletePeerRule(denyRule[0])
err = m.DeleteFilterRule(denyRule)
require.NoError(t, err)
m.mutex.RLock()
_, denyExists := m.incomingDenyRules[addr]
_, allowExists := m.incomingRules[addr]
denyExists := countRulesForAddr(m.incomingDenyRules, addr) > 0
allowExists := countRulesForAddr(m.incomingAcceptRules, addr) > 0
m.mutex.RUnlock()
require.False(t, denyExists, "Deny rules should be empty")
@@ -400,7 +367,7 @@ func TestManagerReset(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Errorf("failed to create Manager: %v", err)
return
@@ -411,7 +378,7 @@ func TestManagerReset(t *testing.T) {
port := &fw.Port{Values: []uint16{80}}
action := fw.ActionDrop
_, err = m.AddPeerFiltering(nil, ip, proto, nil, port, action, "")
_, err = m.AddFilterRule(nil, pfx(ip), fw.Network{}, proto, nil, port, action)
if err != nil {
t.Errorf("failed to add filtering: %v", err)
return
@@ -423,7 +390,7 @@ func TestManagerReset(t *testing.T) {
return
}
if len(m.outgoingRules) != 0 || len(m.incomingRules) != 0 || len(m.incomingDenyRules) != 0 {
if len(m.incomingAcceptRules) != 0 || len(m.incomingDenyRules) != 0 {
t.Errorf("rules are not empty")
}
}
@@ -439,7 +406,7 @@ func TestNotMatchByIP(t *testing.T) {
},
}
m, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Errorf("failed to create Manager: %v", err)
return
@@ -449,7 +416,7 @@ func TestNotMatchByIP(t *testing.T) {
proto := fw.ProtocolUDP
action := fw.ActionAccept
_, err = m.AddPeerFiltering(nil, ip, proto, nil, nil, action, "")
_, err = m.AddFilterRule(nil, pfx(ip), fw.Network{}, proto, nil, nil, action)
if err != nil {
t.Errorf("failed to add filtering: %v", err)
return
@@ -502,7 +469,7 @@ func TestRemovePacketHook(t *testing.T) {
}
// creating manager instance
manager, err := Create(iface, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{IFace: iface, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
if err != nil {
t.Fatalf("Failed to create Manager: %s", err)
}
@@ -519,9 +486,11 @@ func TestRemovePacketHook(t *testing.T) {
}
func TestProcessOutgoingHooks(t *testing.T) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
manager.udpTracker.Close()
@@ -606,7 +575,7 @@ func TestUSPFilterCreatePerformance(t *testing.T) {
ifaceMock := &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
manager, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
time.Sleep(time.Second)
@@ -621,7 +590,7 @@ func TestUSPFilterCreatePerformance(t *testing.T) {
start := time.Now()
for i := 0; i < testMax; i++ {
port := &fw.Port{Values: []uint16{uint16(1000 + i)}}
_, err = manager.AddPeerFiltering(nil, ip, "tcp", nil, port, fw.ActionAccept, "")
_, err = manager.AddFilterRule(nil, pfx(ip), fw.Network{}, "tcp", nil, port, fw.ActionAccept)
require.NoError(t, err, "failed to add rule")
}
@@ -631,9 +600,11 @@ func TestUSPFilterCreatePerformance(t *testing.T) {
}
func TestStatefulFirewall_UDPTracking(t *testing.T) {
manager, err := Create(&IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
}, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{
IFace: &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
},
FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
manager.udpTracker.Close() // Close the existing tracker
@@ -845,7 +816,7 @@ func TestUpdateSetMerge(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
manager, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -858,7 +829,7 @@ func TestUpdateSetMerge(t *testing.T) {
netip.MustParsePrefix("192.168.1.0/24"),
}
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("0.0.0.0/0")},
fw.Network{Set: set},
@@ -931,7 +902,7 @@ func TestUpdateSetDeduplication(t *testing.T) {
SetFilterFunc: func(device.PacketFilter) error { return nil },
}
manager, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
t.Cleanup(func() {
require.NoError(t, manager.Close(nil))
@@ -939,7 +910,7 @@ func TestUpdateSetDeduplication(t *testing.T) {
set := fw.NewDomainSet(domain.List{"example.org"})
rule, err := manager.AddRouteFiltering(
rule, err := manager.AddFilterRule(
nil,
[]netip.Prefix{netip.MustParsePrefix("0.0.0.0/0")},
fw.Network{Set: set},
@@ -1051,7 +1022,7 @@ func TestMSSClamping(t *testing.T) {
},
}
manager, err := Create(ifaceMock, false, flowLogger, 1280)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: 1280})
require.NoError(t, err)
defer func() {
require.NoError(t, manager.Close(nil))
@@ -1243,7 +1214,7 @@ func TestShouldForward(t *testing.T) {
return wgaddr.Address{IP: wgIP, Network: netip.PrefixFrom(wgIP, 24)}
}
manager, err := Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
manager, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
defer func() {
require.NoError(t, manager.Close(nil))
@@ -1358,7 +1329,7 @@ func TestShouldForward(t *testing.T) {
// Re-create manager to pick up the new address with IPv6
require.NoError(t, manager.Close(nil))
manager, err = Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
manager, err = Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(t, err)
v6Cases := []struct {

View File

@@ -5,11 +5,14 @@ import (
"fmt"
"net"
"net/netip"
"os"
"runtime"
"strconv"
"sync"
"time"
log "github.com/sirupsen/logrus"
wgdevice "golang.zx2c4.com/wireguard/device"
"gvisor.dev/gvisor/pkg/buffer"
"gvisor.dev/gvisor/pkg/tcpip"
"gvisor.dev/gvisor/pkg/tcpip/header"
@@ -20,9 +23,9 @@ import (
"gvisor.dev/gvisor/pkg/tcpip/transport/tcp"
"gvisor.dev/gvisor/pkg/tcpip/transport/udp"
"github.com/netbirdio/netbird/client/firewall/uspfilter/common"
"github.com/netbirdio/netbird/client/firewall/uspfilter/conntrack"
nblog "github.com/netbirdio/netbird/client/firewall/uspfilter/log"
"github.com/netbirdio/netbird/client/iface/wgaddr"
nftypes "github.com/netbirdio/netbird/client/internal/netflow/types"
)
@@ -31,8 +34,19 @@ const (
defaultMaxInFlight = 1024
iosReceiveWindow = 16384
iosMaxInFlight = 256
// envForceTCPRACK overrides the platform default for gVisor's RACK loss
// detection. Set to a truthy value to force RACK on, or a falsy value to
// force it off, on any platform.
envForceTCPRACK = "NB_FORCE_TCP_RACK"
)
// IFace provides the WireGuard device and overlay addresses the forwarder needs.
type IFace interface {
GetWGDevice() *wgdevice.Device
Address() wgaddr.Address
}
type Forwarder struct {
logger *nblog.Logger
flowLogger nftypes.FlowLogger
@@ -51,7 +65,7 @@ type Forwarder struct {
pingSemaphore chan struct{}
}
func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.FlowLogger, netstack bool, mtu uint16) (*Forwarder, error) {
func New(iface IFace, logger *nblog.Logger, flowLogger nftypes.FlowLogger, netstack bool, mtu uint16) (*Forwarder, error) {
s := stack.New(stack.Options{
NetworkProtocols: []stack.NetworkProtocolFactory{
ipv4.NewProtocol,
@@ -152,6 +166,8 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
maxInFlight = iosMaxInFlight
}
configureTCPRecovery(s)
tcpForwarder := tcp.NewForwarder(s, receiveWindow, maxInFlight, f.handleTCP)
s.SetTransportProtocolHandler(tcp.ProtocolNumber, tcpForwarder.HandlePacket)
@@ -466,3 +482,31 @@ func probeRawICMP(network, addr string, logger *nblog.Logger) bool {
logger.Debug1("forwarder: raw %s socket access available", network)
return true
}
// configureTCPRecovery disables gVisor's RACK loss detection on Windows, where
// it interacts poorly with the host and collapses throughput on routed TCP
// connections (gVisor issue #9778). Other platforms keep the default. The
// EnvForceTCPRACK environment variable overrides the platform default.
func configureTCPRecovery(s *stack.Stack) {
disableRACK := runtime.GOOS == "windows"
if val := os.Getenv(envForceTCPRACK); val != "" {
force, err := strconv.ParseBool(val)
if err != nil {
log.Warnf("parse %s: %v", envForceTCPRACK, err)
} else {
disableRACK = !force
}
}
if !disableRACK {
return
}
opt := tcpip.TCPRecovery(0)
if err := s.SetTransportProtocolOption(tcp.ProtocolNumber, &opt); err != nil {
log.Warnf("disable TCP RACK loss detection: %v", err)
return
}
log.Info("forwarder: TCP RACK loss detection disabled")
}

View File

@@ -0,0 +1,204 @@
package uspfilter
import (
"context"
"net/netip"
"os"
"strconv"
"sync"
"time"
nblog "github.com/netbirdio/netbird/client/firewall/uspfilter/log"
)
const (
// defaultFragmentTimeout bounds how long a first-fragment verdict is kept
// while the remaining fragments arrive. It mirrors the Linux IP reassembly
// timeout (net.ipv4.ipfrag_time).
defaultFragmentTimeout = 30 * time.Second
// fragmentCleanupInterval is how often expired verdicts are purged.
fragmentCleanupInterval = 10 * time.Second
// defaultMaxFragmentEntries caps the number of concurrently tracked
// fragmented datagrams. The table stays bounded because each datagram is a
// single small entry regardless of how many fragments it is split into, and
// the 13-bit IPv4 fragment-offset field limits any datagram to 64 KiB.
defaultMaxFragmentEntries = 16384
// EnvFragmentMaxEntries overrides defaultMaxFragmentEntries.
EnvFragmentMaxEntries = "NB_FRAGMENT_MAX_ENTRIES"
)
// fragmentVerdict is the decision for a trailing (headerless) fragment.
type fragmentVerdict int
const (
// fragmentDeny drops the fragment: no allowed first fragment is on record.
fragmentDeny fragmentVerdict = iota
// fragmentAllow passes the fragment: it belongs to an allowed datagram and
// does not overlap the already-inspected transport header.
fragmentAllow
// fragmentOverlap drops the fragment and poisons its datagram: it overlaps
// the transport header the ACL inspected (RFC 1858 §4, RFC 3128; RFC 5722
// requires discarding the whole datagram on overlap for IPv6).
fragmentOverlap
)
// fragmentKey identifies a fragmented datagram. It matches the RFC 791 / RFC
// 8200 reassembly key: source, destination, protocol and identification. The id
// is 32-bit to hold both the IPv4 (16-bit) and IPv6 (32-bit) identification.
type fragmentKey struct {
srcIP netip.Addr
dstIP netip.Addr
id uint32
proto uint8
}
// fragmentEntry records the verdict of an allowed first fragment.
type fragmentEntry struct {
// headerEndOctets is the offset, in 8-byte units, at which the first
// fragment's payload ended. A trailing fragment starting before this
// overlaps bytes the ACL already inspected and is rejected.
headerEndOctets uint16
// recordedAt is when the first fragment was accepted. The verdict expires a
// fixed timeout later and is not refreshed, mirroring the kernel reassembly
// timer so a trailing-fragment flood can't keep a datagram alive.
recordedAt time.Time
}
// fragmentTracker records the ACL verdict of a datagram's first fragment so the
// remaining fragments, which carry no L4 header, can inherit the decision
// without reassembling the datagram. Only allowed first fragments are stored;
// anything that cannot be tied to an allowed, non-overlapping first fragment is
// dropped (fail closed).
type fragmentTracker struct {
logger *nblog.Logger
mutex sync.Mutex
entries map[fragmentKey]fragmentEntry
timeout time.Duration
// maxEntries caps the table; atCapacity dedups the capacity warning until
// the table drains below the cap again.
maxEntries int
atCapacity bool
cleanupTicker *time.Ticker
cancel context.CancelFunc
}
func newFragmentTracker(logger *nblog.Logger) *fragmentTracker {
ctx, cancel := context.WithCancel(context.Background())
t := &fragmentTracker{
logger: logger,
entries: make(map[fragmentKey]fragmentEntry),
timeout: defaultFragmentTimeout,
maxEntries: fragmentMaxEntries(logger),
cleanupTicker: time.NewTicker(fragmentCleanupInterval),
cancel: cancel,
}
go t.cleanupRoutine(ctx)
return t
}
func fragmentMaxEntries(logger *nblog.Logger) int {
v := os.Getenv(EnvFragmentMaxEntries)
if v == "" {
return defaultMaxFragmentEntries
}
n, err := strconv.Atoi(v)
if err != nil || n <= 0 {
logger.Warn2("invalid %s=%q, using default", EnvFragmentMaxEntries, v)
return defaultMaxFragmentEntries
}
return n
}
// recordAllowed stores the verdict of an allowed first fragment. headerEndOctets
// is the first fragment's payload length in 8-byte units. When the table is full
// the record is dropped, which fails closed: the datagram's trailing fragments
// will be denied.
func (t *fragmentTracker) recordAllowed(key fragmentKey, headerEndOctets uint16) {
t.mutex.Lock()
defer t.mutex.Unlock()
if t.entries == nil {
return
}
if _, ok := t.entries[key]; !ok && len(t.entries) >= t.maxEntries {
if !t.atCapacity {
t.atCapacity = true
t.logger.Warn2("fragment verdict table at capacity (%d/%d): trailing fragments of new datagrams will be dropped",
len(t.entries), t.maxEntries)
}
return
}
t.entries[key] = fragmentEntry{
headerEndOctets: headerEndOctets,
recordedAt: time.Now(),
}
}
// poison drops any recorded verdict for a datagram, so its later fragments are
// denied until a new allowed first fragment is recorded. Called on every
// offset-zero fragment to defeat offset-zero overlap rewrites (RFC 3128).
func (t *fragmentTracker) poison(key fragmentKey) {
t.mutex.Lock()
defer t.mutex.Unlock()
delete(t.entries, key)
}
// verdict decides the fate of a trailing fragment at fragOffsetOctets (the IPv4
// fragment offset, in 8-byte units). A fragment overlapping the inspected
// header poisons the datagram: the entry is removed so all further fragments of
// that datagram are denied too.
func (t *fragmentTracker) verdict(key fragmentKey, fragOffsetOctets uint16) fragmentVerdict {
t.mutex.Lock()
defer t.mutex.Unlock()
entry, ok := t.entries[key]
if !ok {
return fragmentDeny
}
if time.Since(entry.recordedAt) > t.timeout {
delete(t.entries, key)
return fragmentDeny
}
if fragOffsetOctets < entry.headerEndOctets {
delete(t.entries, key)
return fragmentOverlap
}
return fragmentAllow
}
func (t *fragmentTracker) cleanupRoutine(ctx context.Context) {
defer t.cleanupTicker.Stop()
for {
select {
case <-t.cleanupTicker.C:
t.cleanup()
case <-ctx.Done():
return
}
}
}
func (t *fragmentTracker) cleanup() {
t.mutex.Lock()
defer t.mutex.Unlock()
for key, entry := range t.entries {
if time.Since(entry.recordedAt) > t.timeout {
delete(t.entries, key)
}
}
if len(t.entries) < t.maxEntries {
t.atCapacity = false
}
}
// Close stops the cleanup routine and releases resources.
func (t *fragmentTracker) Close() {
t.cancel()
t.mutex.Lock()
t.entries = nil
t.mutex.Unlock()
}

View File

@@ -0,0 +1,115 @@
package uspfilter
import (
"encoding/binary"
"testing"
)
// benchFilterInbound drives filterInbound over a fixed packet in a tight loop.
// Packets are built once, outside the timed region, so the benchmark measures
// only pipeline cost, which is what an attacker can amplify.
func benchFilterInbound(b *testing.B, pkt []byte) {
b.Helper()
b.ReportAllocs()
b.SetBytes(int64(len(pkt)))
b.ResetTimer()
for i := 0; i < b.N; i++ {
m := benchManager
m.filterInbound(pkt, len(pkt))
}
}
// benchManager is a package-level manager reused across fragment benchmarks so
// setup cost stays out of the timed region.
var benchManager *Manager
func setupBenchManager(b *testing.B) *Manager {
b.Helper()
m := newFragmentTestManager(b)
allowUDP(b, m, 8080)
// Disable conntrack so the allowed-first-fragment path measures transport
// decode + ACL every iteration instead of matching the connection tracked
// on the first iteration.
m.stateful = false
benchManager = m
return m
}
// BenchmarkInbound_NormalPacket is the baseline: a full, non-fragmented UDP
// packet that passes the ACL. Fragment paths should stay comparable to this.
func BenchmarkInbound_NormalPacket(b *testing.B) {
setupBenchManager(b)
pkt := normalUDPPacket(b, 8080, 32)
benchFilterInbound(b, pkt)
}
// BenchmarkInbound_FirstFragmentAllowed measures the first-fragment path:
// transport decode + ACL evaluation + verdict record.
func BenchmarkInbound_FirstFragmentAllowed(b *testing.B) {
setupBenchManager(b)
pkt := firstFragmentUDP(b, 0x2000, 8080, 32)
benchFilterInbound(b, pkt)
}
// BenchmarkInbound_TrailingFragmentAllowed measures the common trailing-fragment
// path: a single map lookup after the first fragment is on record.
func BenchmarkInbound_TrailingFragmentAllowed(b *testing.B) {
m := setupBenchManager(b)
first := firstFragmentUDP(b, 0x3000, 8080, 32)
m.filterInbound(first, len(first))
pkt := trailingFragment(b, 0x3000, 5, false, 24)
benchFilterInbound(b, pkt)
}
// BenchmarkInbound_TrailingFragmentNoFirst is the primary DoS vector: an
// attacker floods trailing fragments with no first fragment on record. Each is
// a map miss and must be cheap.
func BenchmarkInbound_TrailingFragmentNoFirst(b *testing.B) {
setupBenchManager(b)
pkt := trailingFragment(b, 0x4000, 185, false, 40)
benchFilterInbound(b, pkt)
}
// BenchmarkInbound_TinyFirstFragment measures the tiny-fragment drop path: a
// first fragment too small to decode a transport header.
func BenchmarkInbound_TinyFirstFragment(b *testing.B) {
setupBenchManager(b)
pkt := trailingFragment(b, 0x5000, 0, true, 4)
benchFilterInbound(b, pkt)
}
// BenchmarkInbound_TrailingFragmentDistinctIDs is the worst case for the
// verdict table: an attacker varies the datagram id on every packet so no first
// fragment ever matches. Verdict lookups always miss and nothing is recorded,
// so the table cannot grow. Each iteration rewrites the id field in place.
func BenchmarkInbound_TrailingFragmentDistinctIDs(b *testing.B) {
setupBenchManager(b)
pkt := trailingFragment(b, 0x6000, 185, false, 40)
m := benchManager
b.ReportAllocs()
b.SetBytes(int64(len(pkt)))
b.ResetTimer()
for i := 0; i < b.N; i++ {
// IPv4 identification field is at bytes 4:6.
binary.BigEndian.PutUint16(pkt[4:6], uint16(i))
m.filterInbound(pkt, len(pkt))
}
}
// BenchmarkInbound_FirstFragmentDistinctIDs measures sustained first-fragment
// pressure with distinct ids: transport decode + ACL + verdict insert until the
// table caps, exercising the map growth and capacity guard.
func BenchmarkInbound_FirstFragmentDistinctIDs(b *testing.B) {
setupBenchManager(b)
pkt := firstFragmentUDP(b, 0x7000, 8080, 32)
m := benchManager
b.ReportAllocs()
b.SetBytes(int64(len(pkt)))
b.ResetTimer()
for i := 0; i < b.N; i++ {
binary.BigEndian.PutUint16(pkt[4:6], uint16(i))
m.filterInbound(pkt, len(pkt))
}
}

View File

@@ -0,0 +1,554 @@
package uspfilter
import (
"encoding/binary"
"net"
"net/netip"
"testing"
"time"
"github.com/google/gopacket"
"github.com/google/gopacket/layers"
"github.com/stretchr/testify/require"
fw "github.com/netbirdio/netbird/client/firewall/manager"
nbiface "github.com/netbirdio/netbird/client/iface"
"github.com/netbirdio/netbird/client/iface/device"
"github.com/netbirdio/netbird/client/iface/wgaddr"
)
const (
fragTestSrc = "100.10.0.1"
fragTestDst = "100.10.0.100"
fragTestSrcV6 = "fd00::1"
fragTestDstV6 = "fd00::100"
)
func newFragmentTestManager(tb testing.TB) *Manager {
tb.Helper()
ifaceMock := &IFaceMock{
SetFilterFunc: func(device.PacketFilter) error { return nil },
AddressFunc: func() wgaddr.Address {
return wgaddr.Address{
IP: netip.MustParseAddr(fragTestDst),
Network: netip.MustParsePrefix("100.10.0.0/16"),
IPv6: netip.MustParseAddr(fragTestDstV6),
IPv6Net: netip.MustParsePrefix("fd00::/64"),
}
},
}
m, err := Create(Config{IFace: ifaceMock, FlowLogger: flowLogger, MTU: nbiface.DefaultMTU})
require.NoError(tb, err)
require.NoError(tb, m.UpdateLocalIPs())
tb.Cleanup(func() { require.NoError(tb, m.Close(nil)) })
return m
}
// firstFragmentUDPTo builds the first fragment of a fragmented UDP datagram to
// the given destination: it carries the full UDP header plus payloadLen bytes
// of data, with the More Fragments flag set and offset zero.
func firstFragmentUDPTo(tb testing.TB, dst string, id uint16, dstPort uint16, payloadLen int) []byte {
tb.Helper()
ip := &layers.IPv4{
Version: 4,
TTL: 64,
Id: id,
Protocol: layers.IPProtocolUDP,
SrcIP: net.ParseIP(fragTestSrc),
DstIP: net.ParseIP(dst),
Flags: layers.IPv4MoreFragments,
}
udp := &layers.UDP{SrcPort: 40000, DstPort: layers.UDPPort(dstPort)}
require.NoError(tb, udp.SetNetworkLayerForChecksum(ip))
buf := gopacket.NewSerializeBuffer()
opts := gopacket.SerializeOptions{ComputeChecksums: true, FixLengths: true}
require.NoError(tb, gopacket.SerializeLayers(buf, opts, ip, udp, gopacket.Payload(make([]byte, payloadLen))))
return buf.Bytes()
}
func firstFragmentUDP(tb testing.TB, id uint16, dstPort uint16, payloadLen int) []byte {
tb.Helper()
return firstFragmentUDPTo(tb, fragTestDst, id, dstPort, payloadLen)
}
// firstFragmentTCP builds the first fragment of a fragmented TCP datagram: the
// full 20-byte TCP header plus 12 bytes of data, with the More Fragments flag
// set and offset zero.
func firstFragmentTCP(tb testing.TB, id uint16, dstPort uint16) []byte {
tb.Helper()
ip := &layers.IPv4{
Version: 4,
TTL: 64,
Id: id,
Protocol: layers.IPProtocolTCP,
SrcIP: net.ParseIP(fragTestSrc),
DstIP: net.ParseIP(fragTestDst),
Flags: layers.IPv4MoreFragments,
}
tcp := &layers.TCP{SrcPort: 40000, DstPort: layers.TCPPort(dstPort), SYN: true, Window: 64240}
require.NoError(tb, tcp.SetNetworkLayerForChecksum(ip))
buf := gopacket.NewSerializeBuffer()
opts := gopacket.SerializeOptions{ComputeChecksums: true, FixLengths: true}
require.NoError(tb, gopacket.SerializeLayers(buf, opts, ip, tcp, gopacket.Payload(make([]byte, 12))))
return buf.Bytes()
}
// trailingFragmentTo builds a non-first fragment to the given destination: an
// IPv4 header at the given fragment offset (in 8-byte units) carrying raw
// payload and no L4 header.
func trailingFragmentTo(tb testing.TB, dst string, proto layers.IPProtocol, id uint16, fragOffsetOctets uint16, moreFragments bool, payloadLen int) []byte {
tb.Helper()
ip := &layers.IPv4{
Version: 4,
TTL: 64,
Id: id,
Protocol: proto,
SrcIP: net.ParseIP(fragTestSrc),
DstIP: net.ParseIP(dst),
FragOffset: fragOffsetOctets,
}
if moreFragments {
ip.Flags = layers.IPv4MoreFragments
}
buf := gopacket.NewSerializeBuffer()
opts := gopacket.SerializeOptions{FixLengths: true}
require.NoError(tb, gopacket.SerializeLayers(buf, opts, ip, gopacket.Payload(make([]byte, payloadLen))))
return buf.Bytes()
}
func trailingFragment(tb testing.TB, id uint16, fragOffsetOctets uint16, moreFragments bool, payloadLen int) []byte {
tb.Helper()
return trailingFragmentTo(tb, fragTestDst, layers.IPProtocolUDP, id, fragOffsetOctets, moreFragments, payloadLen)
}
// outboundUDPPacket builds a complete outbound UDP packet from the local
// address, used to establish conntrack state for reply-direction tests.
func outboundUDPPacket(tb testing.TB, srcPort, dstPort uint16) []byte {
tb.Helper()
ip := &layers.IPv4{
Version: 4,
TTL: 64,
Id: 1,
Protocol: layers.IPProtocolUDP,
SrcIP: net.ParseIP(fragTestDst),
DstIP: net.ParseIP(fragTestSrc),
}
udp := &layers.UDP{SrcPort: layers.UDPPort(srcPort), DstPort: layers.UDPPort(dstPort)}
require.NoError(tb, udp.SetNetworkLayerForChecksum(ip))
buf := gopacket.NewSerializeBuffer()
opts := gopacket.SerializeOptions{ComputeChecksums: true, FixLengths: true}
require.NoError(tb, gopacket.SerializeLayers(buf, opts, ip, udp, gopacket.Payload(make([]byte, 16))))
return buf.Bytes()
}
// normalUDPPacket builds a complete, non-fragmented UDP packet for baseline
// comparisons against the fragment paths.
func normalUDPPacket(tb testing.TB, dstPort uint16, payloadLen int) []byte {
tb.Helper()
ip := &layers.IPv4{
Version: 4,
TTL: 64,
Id: 1,
Protocol: layers.IPProtocolUDP,
SrcIP: net.ParseIP(fragTestSrc),
DstIP: net.ParseIP(fragTestDst),
}
udp := &layers.UDP{SrcPort: 40000, DstPort: layers.UDPPort(dstPort)}
require.NoError(tb, udp.SetNetworkLayerForChecksum(ip))
buf := gopacket.NewSerializeBuffer()
opts := gopacket.SerializeOptions{ComputeChecksums: true, FixLengths: true}
require.NoError(tb, gopacket.SerializeLayers(buf, opts, ip, udp, gopacket.Payload(make([]byte, payloadLen))))
return buf.Bytes()
}
func allowUDP(tb testing.TB, m *Manager, dstPort uint16) {
tb.Helper()
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrc)), fw.Network{}, fw.ProtocolUDP, nil,
&fw.Port{Values: []uint16{dstPort}}, fw.ActionAccept)
require.NoError(tb, err)
}
// TestFragment_TrailingWithoutFirstDropped is the core bypass repro: a trailing
// fragment with no allowed first fragment on record must be dropped. Before the
// fix, filterInbound returned false (allow) for any fragment.
func TestFragment_TrailingWithoutFirstDropped(t *testing.T) {
m := newFragmentTestManager(t)
frag := trailingFragment(t, 0x1234, 185, false, 40)
require.True(t, m.filterInbound(frag, len(frag)),
"trailing fragment without an allowed first fragment must be dropped")
}
// TestFragment_AllowedFirstPassesTrailing verifies that once a first fragment
// passes the ACL, its trailing fragments inherit the allow verdict.
func TestFragment_AllowedFirstPassesTrailing(t *testing.T) {
m := newFragmentTestManager(t)
allowUDP(t, m, 8080)
// First fragment: UDP header (8) + 32 payload = 40 octets -> headerEnd = 5.
first := firstFragmentUDP(t, 0x2222, 8080, 32)
require.False(t, m.filterInbound(first, len(first)),
"allowed first fragment should pass and be recorded")
trailing := trailingFragment(t, 0x2222, 5, false, 24)
require.False(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of an allowed datagram should pass")
}
// TestFragment_DeniedFirstDropsTrailing verifies that a first fragment blocked
// by the ACL leaves no verdict, so its trailing fragments are dropped.
func TestFragment_DeniedFirstDropsTrailing(t *testing.T) {
m := newFragmentTestManager(t)
// No accept rule: local traffic defaults to deny.
first := firstFragmentUDP(t, 0x3333, 9999, 32)
require.True(t, m.filterInbound(first, len(first)),
"first fragment to a blocked port should be dropped by the ACL")
trailing := trailingFragment(t, 0x3333, 5, false, 24)
require.True(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of a denied datagram must be dropped")
}
// TestFragment_OverlappingHeaderDropped covers the RFC 1858 §4 / RFC 3128
// overlapping-fragment rewrite: a trailing fragment starting inside the range
// the ACL already inspected is dropped and poisons the datagram. TCP is used so
// the overlap lands on real header bytes (the flags at byte 13).
func TestFragment_OverlappingHeaderDropped(t *testing.T) {
m := newFragmentTestManager(t)
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrc)), fw.Network{}, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{8080}}, fw.ActionAccept)
require.NoError(t, err)
// First fragment: TCP header (20) + 12 data = 32 bytes -> headerEnd = 4 octets.
first := firstFragmentTCP(t, 0x4444, 8080)
require.False(t, m.filterInbound(first, len(first)))
// Overlapping fragment at offset 1 (byte 8) falls inside the inspected TCP
// header, so it could rewrite the flags or port on reassembly.
overlap := trailingFragmentTo(t, fragTestDst, layers.IPProtocolTCP, 0x4444, 1, true, 32)
require.True(t, m.filterInbound(overlap, len(overlap)),
"fragment overlapping the inspected header must be dropped")
// The datagram is now poisoned: a later, non-overlapping fragment is also
// dropped because the verdict was removed.
later := trailingFragmentTo(t, fragTestDst, layers.IPProtocolTCP, 0x4444, 4, false, 24)
require.True(t, m.filterInbound(later, len(later)),
"fragments after an overlap must be dropped (datagram poisoned)")
}
// TestFragment_OffsetZeroOverlapPoisons covers the RFC 3128 offset-zero rewrite:
// an allowed first fragment followed by a denied offset-zero fragment for the
// same datagram must not leave the earlier allow verdict in place.
func TestFragment_OffsetZeroOverlapPoisons(t *testing.T) {
m := newFragmentTestManager(t)
allowUDP(t, m, 8080)
allowed := firstFragmentUDP(t, 0x5A5A, 8080, 32)
require.False(t, m.filterInbound(allowed, len(allowed)),
"allowed first fragment should pass and be recorded")
// A second offset-zero fragment to a denied port supersedes the datagram's
// verdict; it is dropped and must not leave the allow in place.
denied := firstFragmentUDP(t, 0x5A5A, 9999, 32)
require.True(t, m.filterInbound(denied, len(denied)),
"denied offset-zero fragment must be dropped")
trailing := trailingFragment(t, 0x5A5A, 5, false, 24)
require.True(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment must be denied after the datagram was poisoned")
}
// TestFragment_TinyFirstDropped covers the tiny-fragment attack: a first
// fragment too small to contain the full transport header can't be
// ACL-evaluated and must be dropped.
func TestFragment_TinyFirstDropped(t *testing.T) {
m := newFragmentTestManager(t)
allowUDP(t, m, 8080)
// IPv4 header + 4 raw bytes, MF set, offset 0: too small for the 8-byte UDP
// header, so it decodes to L3 only.
tiny := trailingFragment(t, 0x5555, 0, true, 4)
require.True(t, m.filterInbound(tiny, len(tiny)),
"tiny first fragment without a full L4 header must be dropped")
}
// TestFragment_TCPFirstFragment verifies the TCP arm of the transport decode: a
// first fragment carrying the full 20-byte TCP header is ACL-evaluated and its
// trailing fragments inherit the verdict.
func TestFragment_TCPFirstFragment(t *testing.T) {
m := newFragmentTestManager(t)
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrc)), fw.Network{}, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{8080}}, fw.ActionAccept)
require.NoError(t, err)
// TCP header (20) + 12 data = 32 bytes -> headerEnd = 4 octets.
first := firstFragmentTCP(t, 0x6666, 8080)
require.False(t, m.filterInbound(first, len(first)),
"allowed TCP first fragment should pass and be recorded")
trailing := trailingFragmentTo(t, fragTestDst, layers.IPProtocolTCP, 0x6666, 4, false, 24)
require.False(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of an allowed TCP datagram should pass")
}
// TestFragment_TCPTinyFirstDropped verifies the TCP minimum header length: 12
// bytes would satisfy a UDP header but falls short of the 20-byte TCP header.
func TestFragment_TCPTinyFirstDropped(t *testing.T) {
m := newFragmentTestManager(t)
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrc)), fw.Network{}, fw.ProtocolTCP, nil,
&fw.Port{Values: []uint16{8080}}, fw.ActionAccept)
require.NoError(t, err)
tiny := trailingFragmentTo(t, fragTestDst, layers.IPProtocolTCP, 0x7777, 0, true, 12)
require.True(t, m.filterInbound(tiny, len(tiny)),
"first fragment shorter than the TCP header must be dropped")
}
// TestFragment_ConntrackAllowsFirstFragment verifies the conntrack branch: reply
// fragments of an outbound-established UDP flow pass without any inbound rule.
func TestFragment_ConntrackAllowsFirstFragment(t *testing.T) {
m := newFragmentTestManager(t)
out := outboundUDPPacket(t, 12345, 40000)
require.False(t, m.filterOutbound(out, len(out)))
first := firstFragmentUDP(t, 0x8888, 12345, 32)
require.False(t, m.filterInbound(first, len(first)),
"reply first fragment should pass via conntrack")
trailing := trailingFragment(t, 0x8888, 5, false, 24)
require.False(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of a tracked flow should pass")
}
// TestFragment_RoutingDisabledDropsFragment verifies routed first fragments are
// dropped when routing is disabled.
func TestFragment_RoutingDisabledDropsFragment(t *testing.T) {
m := newFragmentTestManager(t)
m.routingEnabled.Store(false)
first := firstFragmentUDPTo(t, "198.51.100.10", 0x9999, 8080, 32)
require.True(t, m.filterInbound(first, len(first)),
"routed first fragment must be dropped when routing is disabled")
}
// TestFragment_RouteACL verifies the route-ACL branch: fragments to a non-local
// destination follow the route rules, allowed datagrams pass their trailing
// fragments and denied ones don't.
func TestFragment_RouteACL(t *testing.T) {
m := newFragmentTestManager(t)
m.routingEnabled.Store(true)
m.nativeRouter.Store(false)
_, err := m.AddFilterRule(
[]byte("rt-1"),
[]netip.Prefix{netip.MustParsePrefix("100.10.0.0/16")},
fw.Network{Prefix: netip.MustParsePrefix("198.51.100.0/24")},
fw.ProtocolUDP,
nil,
&fw.Port{Values: []uint16{8080}},
fw.ActionAccept,
)
require.NoError(t, err)
first := firstFragmentUDPTo(t, "198.51.100.10", 0xAAAA, 8080, 32)
require.False(t, m.filterInbound(first, len(first)),
"route-ACL-allowed first fragment should pass")
trailing := trailingFragmentTo(t, "198.51.100.10", layers.IPProtocolUDP, 0xAAAA, 5, false, 24)
require.False(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of an allowed routed datagram should pass")
denied := firstFragmentUDPTo(t, "198.51.100.10", 0xBBBB, 9999, 32)
require.True(t, m.filterInbound(denied, len(denied)),
"route-ACL-denied first fragment must be dropped")
deniedTrailing := trailingFragmentTo(t, "198.51.100.10", layers.IPProtocolUDP, 0xBBBB, 5, false, 24)
require.True(t, m.filterInbound(deniedTrailing, len(deniedTrailing)),
"trailing fragment of a denied routed datagram must be dropped")
}
// TestFragment_ExpiredVerdictDropsTrailing verifies a verdict older than the
// tracker timeout no longer admits trailing fragments.
func TestFragment_ExpiredVerdictDropsTrailing(t *testing.T) {
m := newFragmentTestManager(t)
allowUDP(t, m, 8080)
first := firstFragmentUDP(t, 0xCCCC, 8080, 32)
require.False(t, m.filterInbound(first, len(first)))
m.fragments.mutex.Lock()
for key, entry := range m.fragments.entries {
entry.recordedAt = time.Now().Add(-defaultFragmentTimeout - time.Second)
m.fragments.entries[key] = entry
}
m.fragments.mutex.Unlock()
trailing := trailingFragment(t, 0xCCCC, 5, false, 24)
require.True(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment after verdict expiry must be dropped")
}
// TestFragment_CapacityFailsClosed verifies the table cap: at capacity, new
// datagram verdicts are not recorded (their trailing fragments are dropped)
// while already-recorded datagrams keep working.
func TestFragment_CapacityFailsClosed(t *testing.T) {
m := newFragmentTestManager(t)
allowUDP(t, m, 8080)
m.fragments.mutex.Lock()
m.fragments.maxEntries = 1
m.fragments.mutex.Unlock()
first1 := firstFragmentUDP(t, 0x0101, 8080, 32)
require.False(t, m.filterInbound(first1, len(first1)))
first2 := firstFragmentUDP(t, 0x0202, 8080, 32)
require.False(t, m.filterInbound(first2, len(first2)),
"first fragment itself still passes at capacity")
trailing2 := trailingFragment(t, 0x0202, 5, false, 24)
require.True(t, m.filterInbound(trailing2, len(trailing2)),
"trailing fragment of an unrecorded datagram must be dropped at capacity")
trailing1 := trailingFragment(t, 0x0101, 5, false, 24)
require.False(t, m.filterInbound(trailing1, len(trailing1)),
"already-recorded datagram should keep passing at capacity")
}
// v6FragmentHeader builds the 8-byte IPv6 fragment extension header for the
// given inner protocol, offset (8-byte units), More Fragments bit and id.
func v6FragmentHeader(proto layers.IPProtocol, offsetOctets uint16, moreFragments bool, id uint32) []byte {
offsetFlags := offsetOctets << 3
if moreFragments {
offsetFlags |= 1
}
hdr := make([]byte, 8)
hdr[0] = uint8(proto)
binary.BigEndian.PutUint16(hdr[2:4], offsetFlags)
binary.BigEndian.PutUint32(hdr[4:8], id)
return hdr
}
func v6UDPHeader(dstPort uint16, dataLen int) []byte {
hdr := make([]byte, 8)
binary.BigEndian.PutUint16(hdr[0:2], 40000)
binary.BigEndian.PutUint16(hdr[2:4], dstPort)
binary.BigEndian.PutUint16(hdr[4:6], uint16(8+dataLen))
return hdr
}
// firstFragmentUDPv6 builds the first fragment of a fragmented IPv6 UDP
// datagram: fragment header (offset 0, More Fragments set) + full UDP header +
// data.
func firstFragmentUDPv6(tb testing.TB, id uint32, dstPort uint16, dataLen int) []byte {
tb.Helper()
return fragmentUDPv6(tb, id, dstPort, dataLen, true)
}
// fragmentUDPv6 builds an offset-zero IPv6 UDP fragment. With moreFragments
// false it is an atomic fragment (a complete datagram, RFC 6946).
func fragmentUDPv6(tb testing.TB, id uint32, dstPort uint16, dataLen int, moreFragments bool) []byte {
tb.Helper()
ip := &layers.IPv6{
Version: 6,
NextHeader: layers.IPProtocolIPv6Fragment,
HopLimit: 64,
SrcIP: net.ParseIP(fragTestSrcV6),
DstIP: net.ParseIP(fragTestDstV6),
}
payload := append(v6FragmentHeader(layers.IPProtocolUDP, 0, moreFragments, id), v6UDPHeader(dstPort, dataLen)...)
payload = append(payload, make([]byte, dataLen)...)
buf := gopacket.NewSerializeBuffer()
require.NoError(tb, gopacket.SerializeLayers(buf, gopacket.SerializeOptions{FixLengths: true}, ip, gopacket.Payload(payload)))
return buf.Bytes()
}
// trailingFragmentV6 builds a non-first IPv6 fragment: fragment header at the
// given offset carrying raw data and no transport header.
func trailingFragmentV6(tb testing.TB, id uint32, offsetOctets uint16, moreFragments bool, dataLen int) []byte {
tb.Helper()
ip := &layers.IPv6{
Version: 6,
NextHeader: layers.IPProtocolIPv6Fragment,
HopLimit: 64,
SrcIP: net.ParseIP(fragTestSrcV6),
DstIP: net.ParseIP(fragTestDstV6),
}
payload := append(v6FragmentHeader(layers.IPProtocolUDP, offsetOctets, moreFragments, id), make([]byte, dataLen)...)
buf := gopacket.NewSerializeBuffer()
require.NoError(tb, gopacket.SerializeLayers(buf, gopacket.SerializeOptions{FixLengths: true}, ip, gopacket.Payload(payload)))
return buf.Bytes()
}
// TestFragmentV6_TrailingWithoutFirstDropped verifies the IPv6 bypass is closed:
// a trailing fragment with no allowed first fragment is dropped.
func TestFragmentV6_TrailingWithoutFirstDropped(t *testing.T) {
m := newFragmentTestManager(t)
frag := trailingFragmentV6(t, 0xAABBCCDD, 100, false, 40)
require.True(t, m.filterInbound(frag, len(frag)),
"IPv6 trailing fragment without an allowed first fragment must be dropped")
}
// TestFragmentV6_AllowedFirstPassesTrailing verifies IPv6 fragments are
// evaluated like IPv4: an allowed first fragment lets its trailing fragments
// through.
func TestFragmentV6_AllowedFirstPassesTrailing(t *testing.T) {
m := newFragmentTestManager(t)
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrcV6)), fw.Network{}, fw.ProtocolUDP, nil,
&fw.Port{Values: []uint16{8080}}, fw.ActionAccept)
require.NoError(t, err)
// First fragment: UDP header (8) + 32 data = 40 octets -> headerEnd = 5.
first := firstFragmentUDPv6(t, 0xAABBCCDD, 8080, 32)
require.False(t, m.filterInbound(first, len(first)),
"allowed IPv6 first fragment should pass and be recorded")
trailing := trailingFragmentV6(t, 0xAABBCCDD, 5, false, 24)
require.False(t, m.filterInbound(trailing, len(trailing)),
"trailing fragment of an allowed IPv6 datagram should pass")
}
// TestFragmentV6_AtomicNotCached verifies an IPv6 atomic fragment (fragment
// header with offset 0 and no More Fragments, a complete datagram per RFC 6946)
// is evaluated but not recorded, so a flood of allowed atomic fragments can't
// exhaust the verdict table.
func TestFragmentV6_AtomicNotCached(t *testing.T) {
m := newFragmentTestManager(t)
_, err := m.AddFilterRule(nil, pfx(net.ParseIP(fragTestSrcV6)), fw.Network{}, fw.ProtocolUDP, nil,
&fw.Port{Values: []uint16{8080}}, fw.ActionAccept)
require.NoError(t, err)
atomic := fragmentUDPv6(t, 0xA70301C, 8080, 16, false)
require.False(t, m.filterInbound(atomic, len(atomic)),
"allowed IPv6 atomic fragment should pass")
m.fragments.mutex.Lock()
n := len(m.fragments.entries)
m.fragments.mutex.Unlock()
require.Zero(t, n, "atomic fragment must not create a verdict entry")
// A genuine fragmented datagram (More Fragments set) is still recorded.
first := fragmentUDPv6(t, 0xBEEF, 8080, 32, true)
require.False(t, m.filterInbound(first, len(first)))
m.fragments.mutex.Lock()
n = len(m.fragments.entries)
m.fragments.mutex.Unlock()
require.Equal(t, 1, n, "genuine first fragment must record a verdict")
}

View File

@@ -9,7 +9,6 @@ import (
log "github.com/sirupsen/logrus"
nberrors "github.com/netbirdio/netbird/client/errors"
"github.com/netbirdio/netbird/client/internal/statemanager"
)
type action string
@@ -20,35 +19,20 @@ const (
firewallRuleName = "Netbird"
)
// Close cleans up the firewall manager by removing all rules and closing trackers
func (m *Manager) Close(*statemanager.Manager) error {
m.mutex.Lock()
defer m.mutex.Unlock()
m.resetState()
if !isWindowsFirewallReachable() {
return nil
}
var merr *multierror.Error
if isFirewallRuleActive(firewallRuleName) {
if err := manageFirewallRule(firewallRuleName, deleteRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove windows firewall rule: %w", err))
}
}
if isFirewallRuleActive(firewallRuleName + "-v6") {
if err := manageFirewallRule(firewallRuleName+"-v6", deleteRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove windows v6 firewall rule: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
// WindowsInterfaceAllower opens the NetBird interface in the Windows firewall
// via netsh advfirewall rules. It implements InterfaceAllower for the userspace
// firewall on Windows.
type WindowsInterfaceAllower struct {
iface Iface
}
// AllowNetbird allows netbird interface traffic
func (m *Manager) AllowNetbird() error {
// NewWindowsInterfaceAllower builds the Windows netsh-based interface allower.
func NewWindowsInterfaceAllower(iface Iface) *WindowsInterfaceAllower {
return &WindowsInterfaceAllower{iface: iface}
}
// Apply adds inbound-allow netsh rules for the interface's addresses.
func (a *WindowsInterfaceAllower) Apply() error {
if !isWindowsFirewallReachable() {
return nil
}
@@ -60,13 +44,13 @@ func (m *Manager) AllowNetbird() error {
"enable=yes",
"action=allow",
"profile=any",
"localip="+m.wgIface.Address().IP.String(),
"localip="+a.iface.Address().IP.String(),
); err != nil {
return err
}
}
if v6 := m.wgIface.Address().IPv6; v6.IsValid() && !isFirewallRuleActive(firewallRuleName+"-v6") {
if v6 := a.iface.Address().IPv6; v6.IsValid() && !isFirewallRuleActive(firewallRuleName+"-v6") {
if err := manageFirewallRule(firewallRuleName+"-v6",
addRule,
"dir=in",
@@ -82,8 +66,27 @@ func (m *Manager) AllowNetbird() error {
return nil
}
func manageFirewallRule(ruleName string, action action, extraArgs ...string) error {
// Close removes the netsh rules added by Apply.
func (a *WindowsInterfaceAllower) Close() error {
if !isWindowsFirewallReachable() {
return nil
}
var merr *multierror.Error
if isFirewallRuleActive(firewallRuleName) {
if err := manageFirewallRule(firewallRuleName, deleteRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove windows firewall rule: %w", err))
}
}
if isFirewallRuleActive(firewallRuleName + "-v6") {
if err := manageFirewallRule(firewallRuleName+"-v6", deleteRule); err != nil {
merr = multierror.Append(merr, fmt.Errorf("remove windows v6 firewall rule: %w", err))
}
}
return nberrors.FormatErrorOrNil(merr)
}
func manageFirewallRule(ruleName string, action action, extraArgs ...string) error {
args := []string{"advfirewall", "firewall", string(action), "rule", "name=" + ruleName}
if action == addRule {
args = append(args, extraArgs...)

View File

@@ -7,8 +7,6 @@ import (
"sync/atomic"
log "github.com/sirupsen/logrus"
"github.com/netbirdio/netbird/client/firewall/uspfilter/common"
)
// localIPSnapshot is an immutable snapshot of local IP addresses, swapped
@@ -60,7 +58,7 @@ func processInterface(iface net.Interface, ips map[netip.Addr]struct{}, addresse
}
// UpdateLocalIPs rebuilds the local IP snapshot and swaps it in atomically.
func (m *localIPManager) UpdateLocalIPs(iface common.IFaceMapper) (err error) {
func (m *localIPManager) UpdateLocalIPs(iface Iface) (err error) {
defer func() {
if r := recover(); r != nil {
err = fmt.Errorf("panic: %v", r)

Some files were not shown because too many files have changed in this diff Show More