Merge branch 'main' into force-routing-peer-dns-reverse-proxy

## 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 -->

.github/workflows/agent-network-e2e.yml

@@ -51,6 +51,9 @@ 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 }}

.github/workflows/frontend-ui.yml

@@ -86,7 +86,7 @@ jobs:
             ${{ runner.os }}-pnpm-
 
       - name: Install dependencies
-        run: pnpm install --frozen-lockfile
+        run: pnpm install --frozen-lockfile --ignore-scripts
 
       - name: Generate Wails bindings
         run: pnpm run bindings

.github/workflows/golangci-lint.yml

@@ -45,7 +45,7 @@ jobs:
             display_name: Linux
     name: ${{ matrix.display_name }}
     runs-on: ${{ matrix.os }}
-    timeout-minutes: 15
+    timeout-minutes: 25
     steps:
       - name: Checkout code
         uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
@@ -79,4 +79,4 @@ jobs:
           skip-cache: true
           skip-save-cache: true
           cache-invalidation-interval: 0
-          args: --timeout=12m
+          args: --timeout=20m

client/android/client.go

@@ -247,6 +247,9 @@ 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
 			}

client/cmd/root.go

@@ -6,7 +6,6 @@ import (
 	"fmt"
 	"io"
 	"io/fs"
-	"net"
 	"os"
 	"os/signal"
 	"path"
@@ -80,8 +79,6 @@ var (
 	updateSettingsDisabled bool
 	captureEnabled         bool
 	networksDisabled       bool
-	socketOwner            string
-	strictSocketDisabled   bool
 
 	rootCmd = &cobra.Command{
 		Use:          "netbird",
@@ -146,12 +143,10 @@ func init() {
 
 	defaultDaemonAddr := "unix:///var/run/netbird.sock"
 	if runtime.GOOS == "windows" {
-		// Named pipe (not loopback TCP): the pipe SDDL gates who may connect and
-		// the pipe client token carries the caller's SID for per-RPC authorization.
-		defaultDaemonAddr = "npipe://netbird"
+		defaultDaemonAddr = "tcp://127.0.0.1:41731"
 	}
 
-	rootCmd.PersistentFlags().StringVar(&daemonAddr, "daemon-addr", defaultDaemonAddr, "Daemon service address to serve CLI requests [unix|tcp|npipe]://[path|host:port|name]")
+	rootCmd.PersistentFlags().StringVar(&daemonAddr, "daemon-addr", defaultDaemonAddr, "Daemon service address to serve CLI requests [unix|tcp]://[path|host:port]")
 	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")
@@ -270,31 +265,16 @@ func FlagNameToEnvVar(cmdFlag string, prefix string) string {
 }
 
 // DialClientGRPCServer returns client connection to the daemon server.
-//
-// The daemon reads the caller's kernel identity from the transport (SO_PEERCRED
-// on a Unix socket, the client token on a Windows named pipe), so the client
-// side uses insecure (plaintext) credentials — it needs no cooperation to be
-// identified. For npipe addresses we install a context dialer since gRPC's
-// resolver does not understand Windows named pipes.
 func DialClientGRPCServer(ctx context.Context, addr string) (*grpc.ClientConn, error) {
 	ctx, cancel := context.WithTimeout(ctx, time.Second*10)
 	defer cancel()
 
-	opts := []grpc.DialOption{
+	return grpc.DialContext(
+		ctx,
+		strings.TrimPrefix(addr, "tcp://"),
 		grpc.WithTransportCredentials(insecure.NewCredentials()),
 		grpc.WithBlock(),
-	}
-
-	target := strings.TrimPrefix(addr, "tcp://")
-	if strings.HasPrefix(addr, "npipe://") {
-		path := pipePath(strings.TrimPrefix(addr, "npipe://"))
-		opts = append(opts, grpc.WithContextDialer(func(dialCtx context.Context, _ string) (net.Conn, error) {
-			return dialNamedPipe(dialCtx, path)
-		}))
-		target = "passthrough:///netbird-daemon-pipe"
-	}
-
-	return grpc.DialContext(ctx, target, opts...)
+	)
 }
 
 // WithBackOff execute function in backoff cycle.

client/cmd/service.go

@@ -56,9 +56,6 @@ func init() {
 	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")
 
-	serviceCmd.PersistentFlags().StringVar(&socketOwner, "socket-owner", "", "user to own the daemon control socket; restricts it to that user plus the netbird group (0660). If unset, the first client to connect claims ownership (trust-on-first-use). Persisted via: netbird service install --socket-owner")
-	serviceCmd.PersistentFlags().BoolVar(&strictSocketDisabled, "disable-strict-socket", false, "leave the daemon control socket world-writable (0666) instead of restricting it (root-only, discouraged). Persisted via: netbird service install --disable-strict-socket")
-
 	rootCmd.PersistentFlags().StringVarP(&serviceName, "service", "s", defaultServiceName, "Netbird system service name")
 	serviceEnvDesc := `Sets extra environment variables for the service. ` +
 		`You can specify a comma-separated list of KEY=VALUE pairs. ` +

client/cmd/service_controller.go

@@ -5,7 +5,6 @@ package cmd
 import (
 	"context"
 	"fmt"
-	"runtime"
 	"time"
 
 	"github.com/kardianos/service"
@@ -14,31 +13,12 @@ import (
 	"github.com/spf13/cobra"
 	"google.golang.org/grpc"
 
-	"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"
 )
 
-// daemonServerOptions returns the gRPC server options that install peer-identity
-// transport credentials on the daemon control channel. Identity extraction is
-// only possible over a Unix socket (SO_PEERCRED) or Windows named pipe (client
-// token); over TCP, or on platforms without a peer-credential primitive, the
-// daemon runs without per-caller authorization and logs a warning.
-func daemonServerOptions(network string) []grpc.ServerOption {
-	creds := ipcauth.NewTransportCredentials()
-	if creds == nil {
-		log.Warnf("daemon control channel has no peer-identity primitive on %s; per-caller authorization is disabled", runtime.GOOS)
-		return nil
-	}
-	if network == "tcp" {
-		log.Warnf("daemon is listening on TCP (%s); peer identity cannot be authenticated over TCP, per-caller authorization is disabled", daemonAddr)
-		return nil
-	}
-	return []grpc.ServerOption{grpc.Creds(creds)}
-}
-
 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")
@@ -57,13 +37,8 @@ func (p *program) Start(svc service.Service) error {
 	// Collect static system and platform information
 	system.UpdateStaticInfoAsync()
 
-	network, _, err := parseListenAddress(daemonAddr)
-	if err != nil {
-		return fmt.Errorf("parse daemon address: %w", err)
-	}
-
-	// in any case, even if configuration does not exist we run daemon to serve the CLI gRPC API.
-	p.serv = grpc.NewServer(daemonServerOptions(network)...)
+	// in any case, even if configuration does not exists we run daemon to serve CLI gRPC API.
+	p.serv = grpc.NewServer()
 
 	daemonListener, err := listenOnAddress(daemonAddr)
 	if err != nil {
@@ -87,8 +62,7 @@ func (p *program) Start(svc service.Service) error {
 			defer jsonListener.Close()
 		}
 
-		serveListener, err := secureDaemonListener(daemonListener)
-		if err != nil {
+		if err := daemonListener.chmodUnixSocket("daemon"); err != nil {
 			log.Error(err)
 			return
 		}
@@ -110,7 +84,6 @@ func (p *program) Start(svc service.Service) error {
 		p.serverInstanceMu.Unlock()
 
 		if jsonListener != nil {
-			log.Warnf("JSON gateway (--enable-json-socket) re-dials the daemon locally as the daemon's own identity and BYPASSES per-caller authorization; restrict access to %s separately", jsonSocket)
 			if err := p.startJSONGateway(jsonListener, daemonAddr); err != nil {
 				log.Fatalf("failed to start daemon JSON server: %v", err)
 			}
@@ -119,7 +92,7 @@ func (p *program) Start(svc service.Service) error {
 		}
 
 		log.Printf("started daemon server: %v", daemonListener.address)
-		if err := p.serv.Serve(serveListener); err != nil {
+		if err := p.serv.Serve(daemonListener.Listener); err != nil {
 			log.Errorf("failed to serve daemon requests: %v", err)
 		}
 	}()

client/cmd/service_installer.go

@@ -71,14 +71,6 @@ func buildServiceArguments() []string {
 		args = append(args, "--enable-json-socket", "--json-socket", jsonSocket)
 	}
 
-	if socketOwner != "" {
-		args = append(args, "--socket-owner", socketOwner)
-	}
-
-	if strictSocketDisabled {
-		args = append(args, "--disable-strict-socket")
-	}
-
 	return args
 }
 

client/cmd/service_params.go

@@ -32,8 +32,6 @@ type serviceParams struct {
 	EnableCapture         bool              `json:"enable_capture,omitempty"`
 	DisableNetworks       bool              `json:"disable_networks,omitempty"`
 	EnableJSONSocket      bool              `json:"enable_json_socket,omitempty"`
-	SocketOwner           string            `json:"socket_owner,omitempty"`
-	DisableStrictSocket   bool              `json:"disable_strict_socket,omitempty"`
 	ServiceEnvVars        map[string]string `json:"service_env_vars,omitempty"`
 }
 
@@ -88,8 +86,6 @@ func currentServiceParams() *serviceParams {
 		EnableCapture:         captureEnabled,
 		DisableNetworks:       networksDisabled,
 		EnableJSONSocket:      enableJSONSocket,
-		SocketOwner:           socketOwner,
-		DisableStrictSocket:   strictSocketDisabled,
 	}
 
 	if len(serviceEnvVars) > 0 {
@@ -169,14 +165,6 @@ func applyServiceParams(cmd *cobra.Command, params *serviceParams) {
 		networksDisabled = params.DisableNetworks
 	}
 
-	if !serviceCmd.PersistentFlags().Changed("socket-owner") {
-		socketOwner = params.SocketOwner
-	}
-
-	if !serviceCmd.PersistentFlags().Changed("disable-strict-socket") {
-		strictSocketDisabled = params.DisableStrictSocket
-	}
-
 	applyServiceEnvParams(cmd, params)
 }
 

client/cmd/service_pipe_other.go

@@ -1,20 +0,0 @@
-//go:build !windows
-
-package cmd
-
-import (
-	"context"
-	"fmt"
-	"net"
-	"runtime"
-)
-
-// listenNamedPipe is unsupported off Windows; named pipes are a Windows-only transport.
-func listenNamedPipe(string) (net.Listener, error) {
-	return nil, fmt.Errorf("named pipe daemon socket is only supported on Windows, not %s", runtime.GOOS)
-}
-
-// dialNamedPipe is unsupported off Windows.
-func dialNamedPipe(context.Context, string) (net.Conn, error) {
-	return nil, fmt.Errorf("named pipe daemon socket is only supported on Windows, not %s", runtime.GOOS)
-}

client/cmd/service_pipe_windows.go

@@ -1,32 +0,0 @@
-//go:build windows
-
-package cmd
-
-import (
-	"context"
-	"net"
-	"time"
-
-	"github.com/Microsoft/go-winio"
-
-	"github.com/netbirdio/netbird/client/internal/ipcauth"
-)
-
-// listenNamedPipe creates the daemon control named pipe with a tight SDDL
-// (SYSTEM + Administrators + interactive users). ListenPipe fails if the pipe
-// already exists (first-instance semantics), which prevents a squatting process
-// from pre-creating it — we surface that error loudly rather than falling back.
-func listenNamedPipe(path string) (net.Listener, error) {
-	return winio.ListenPipe(path, &winio.PipeConfig{
-		SecurityDescriptor: ipcauth.DefaultPipeSDDL(),
-	})
-}
-
-// dialNamedPipe connects to the daemon control named pipe.
-func dialNamedPipe(ctx context.Context, path string) (net.Conn, error) {
-	if deadline, ok := ctx.Deadline(); ok {
-		timeout := time.Until(deadline)
-		return winio.DialPipe(path, &timeout)
-	}
-	return winio.DialPipeContext(ctx, path)
-}

client/cmd/service_socket.go

@@ -26,15 +26,6 @@ func listenOnAddress(addr string) (*socketListener, error) {
 		return nil, err
 	}
 
-	if network == "npipe" {
-		path := pipePath(address)
-		listener, err := listenNamedPipe(path)
-		if err != nil {
-			return nil, err
-		}
-		return &socketListener{Listener: listener, network: network, address: path}, nil
-	}
-
 	if network == "unix" {
 		removeStaleUnixSocket(address)
 	}
@@ -54,23 +45,13 @@ func parseListenAddress(addr string) (string, string, error) {
 	}
 
 	switch network {
-	case "unix", "tcp", "npipe":
+	case "unix", "tcp":
 		return network, address, nil
 	default:
 		return "", "", fmt.Errorf("unsupported daemon address protocol: %v", network)
 	}
 }
 
-// pipePath maps a daemon-addr npipe name (e.g. "netbird" from "npipe://netbird")
-// to a Windows named-pipe path (\\.\pipe\netbird). A caller may also pass a full
-// \\.\pipe\ path, which is returned unchanged.
-func pipePath(name string) string {
-	if strings.HasPrefix(name, `\\`) {
-		return name
-	}
-	return `\\.\pipe\` + name
-}
-
 func removeStaleUnixSocket(path string) {
 	stat, err := os.Lstat(path)
 	if err != nil {

client/cmd/service_socket_perms_other.go

@@ -1,11 +0,0 @@
-//go:build windows
-
-package cmd
-
-import "net"
-
-// secureDaemonListener is a no-op on Windows: the named-pipe SDDL gates who may
-// connect (Layer 1), and the pipe client token supplies per-RPC identity.
-func secureDaemonListener(l *socketListener) (net.Listener, error) {
-	return l.Listener, nil
-}

client/cmd/service_socket_perms_unix.go

@@ -1,225 +0,0 @@
-//go:build !windows && !ios && !android
-
-package cmd
-
-import (
-	"errors"
-	"fmt"
-	"net"
-	"os"
-	"os/exec"
-	"os/user"
-	"strconv"
-	"sync"
-
-	log "github.com/sirupsen/logrus"
-
-	"github.com/netbirdio/netbird/client/internal/ipcauth"
-	"github.com/netbirdio/netbird/client/internal/shell"
-)
-
-// secureDaemonListener applies the Layer-1 access control to the daemon control
-// socket and returns the listener to serve on. For a Unix socket this restricts
-// the socket to an owner (plus the netbird group); for anything else it is a
-// no-op (TCP is legacy/unauthenticated; named pipes are gated by their SDDL).
-func secureDaemonListener(l *socketListener) (net.Listener, error) {
-	if l.network != "unix" {
-		return l.Listener, nil
-	}
-
-	owner := effectiveSocketOwner()
-	switch {
-	case strictSocketDisabled:
-		// Root-only opt-out (via service.json): leave it world-writable.
-		if err := os.Chmod(l.address, 0666); err != nil {
-			return nil, fmt.Errorf("set daemon socket permissions: %w", err)
-		}
-		log.Warnf("daemon control socket left world-writable (0666) by --disable-strict-socket")
-		return l.Listener, nil
-
-	case owner != "":
-		// Seeded owner (flag, MDM, or persisted TOFU result): restrict before
-		// serving so there is no open window.
-		uid, err := lookupUser(owner)
-		if err != nil {
-			return nil, fmt.Errorf("lookup socket owner %q: %w", owner, err)
-		}
-		if err := restrictSocket(l.address, uid); err != nil {
-			return nil, fmt.Errorf("restrict socket to %q: %w", owner, err)
-		}
-		return l.Listener, nil
-
-	default:
-		// Trust-on-first-use: open the socket now; tofuListener locks it to the
-		// first caller's uid on the first connection.
-		if err := os.Chmod(l.address, 0666); err != nil {
-			return nil, fmt.Errorf("set daemon socket permissions: %w", err)
-		}
-		return &tofuListener{Listener: l.Listener, path: l.address, owner: -1}, nil
-	}
-}
-
-func lookupUser(username string) (int, error) {
-	u, err := shell.LookupWithGetent(username)
-	if err != nil {
-		return -1, fmt.Errorf("lookup user %s: %w", username, err)
-	}
-	uid, err := strconv.Atoi(u.Uid)
-	if err != nil {
-		return -1, fmt.Errorf("parse uid %s: %w", u.Uid, err)
-	}
-	return uid, nil
-}
-
-// addGroup creates a system group if it doesn't already exist and returns the gid.
-// Must run as root.
-func addGroup(name string) (int, error) {
-	group, err := shell.LookupGroupWithGetent(name)
-	if err == nil {
-		gid, err := strconv.ParseInt(group.Gid, 10, 64)
-		return int(gid), err
-	}
-
-	groupadd, err := exec.LookPath("groupadd")
-	if err != nil {
-		// Fallback for Alpine/BusyBox systems.
-		if groupadd, err = exec.LookPath("addgroup"); err != nil {
-			return -1, errors.New("neither groupadd nor addgroup found")
-		}
-	}
-
-	// Use --system for a service/daemon group (no login, low GID).
-	out, err := exec.Command(groupadd, "--system", name).CombinedOutput()
-	if err != nil {
-		return -1, fmt.Errorf("create group %q: %w: %s", name, err, out)
-	}
-	if group, err := shell.LookupGroupWithGetent(name); err == nil {
-		gid, err := strconv.ParseInt(group.Gid, 10, 64)
-		return int(gid), err
-	}
-	return -1, fmt.Errorf("lookup group %q: %w", name, err)
-}
-
-// restrictSocket locks the unix socket down to the owner uid plus the netbird
-// group (0660). If the group cannot be created or applied, it fails closed to
-// owner-only 0600 — it never leaves the socket world-writable.
-func restrictSocket(path string, uid int) error {
-	gid, err := addGroup("netbird")
-	if err != nil {
-		log.Errorf("create netbird group, failing closed to owner-only 0600: %v", err)
-		return chownChmod(path, uid, -1, 0600)
-	}
-	if err := chownChmod(path, uid, gid, 0660); err != nil {
-		log.Errorf("apply netbird group to socket, failing closed to owner-only 0600: %v", err)
-		return chownChmod(path, uid, -1, 0600)
-	}
-	return nil
-}
-
-// chownChmod sets ownership and mode on the socket. A gid of -1 leaves the
-// group unchanged.
-func chownChmod(path string, uid, gid int, mode os.FileMode) error {
-	if err := os.Chown(path, uid, gid); err != nil {
-		return fmt.Errorf("chown socket %s: %w", path, err)
-	}
-	if err := os.Chmod(path, mode); err != nil {
-		return fmt.Errorf("chmod socket %s: %w", path, err)
-	}
-	return nil
-}
-
-// tofuListener implements trust-on-first-use for the daemon control socket.
-// The socket starts world-writable; the first caller's uid (read via SO_PEERCRED)
-// becomes the owner. On that first connection the socket is restricted and the
-// owner persisted so the open window never reopens on later starts. Connections
-// that raced in during the open window and are neither the owner nor root are
-// dropped. Changing the socket mode does not disturb the already-open
-// connection, so the first caller's request is served normally.
-type tofuListener struct {
-	net.Listener
-	path  string
-	mu    sync.Mutex
-	owner int // -1 until claimed
-}
-
-func (l *tofuListener) Accept() (net.Conn, error) {
-	for {
-		c, err := l.Listener.Accept()
-		if err != nil {
-			return nil, err
-		}
-
-		id, err := ipcauth.PeerIdentity(c)
-		if err != nil {
-			log.Errorf("read peer credentials, dropping connection: %v", err)
-			_ = c.Close()
-			continue
-		}
-		uid := int(id.UID)
-
-		l.mu.Lock()
-		if l.owner == -1 {
-			if err := restrictSocket(l.path, uid); err != nil {
-				l.mu.Unlock()
-				_ = c.Close()
-				// Refuse to serve on a socket we could not lock down.
-				return nil, fmt.Errorf("restrict socket on first connection: %w", err)
-			}
-			l.owner = uid
-			persistSocketOwner(uid)
-			log.Infof("control socket restricted to first caller (uid %d)", uid)
-			l.mu.Unlock()
-			return c, nil
-		}
-		owner := l.owner
-		l.mu.Unlock()
-
-		// New connects are already gated by the 0660 perms set above; this only
-		// drops anything that slipped in during the brief open window.
-		if uid != owner && uid != 0 {
-			log.Warnf("dropping non-owner connection (uid %d) during socket bootstrap", uid)
-			_ = c.Close()
-			continue
-		}
-		return c, nil
-	}
-}
-
-// effectiveSocketOwner returns the configured socket owner: the --socket-owner
-// flag when set, otherwise the owner persisted by a previous TOFU migration.
-func effectiveSocketOwner() string {
-	if socketOwner != "" {
-		return socketOwner
-	}
-	params, err := loadServiceParams()
-	if err != nil {
-		log.Errorf("load service params for socket owner: %v", err)
-		return ""
-	}
-	if params != nil {
-		return params.SocketOwner
-	}
-	return ""
-}
-
-// persistSocketOwner records the TOFU-selected owner (by username) so the next
-// daemon start restricts the socket immediately, with no open window.
-func persistSocketOwner(uid int) {
-	u, err := user.LookupId(strconv.Itoa(uid))
-	if err != nil {
-		log.Errorf("resolve uid %d to username for persistence: %v", uid, err)
-		return
-	}
-	params, err := loadServiceParams()
-	if err != nil {
-		log.Errorf("load service params to persist socket owner: %v", err)
-		return
-	}
-	if params == nil {
-		params = currentServiceParams()
-	}
-	params.SocketOwner = u.Username
-	if err := saveServiceParams(params); err != nil {
-		log.Errorf("persist socket owner: %v", err)
-	}
-}

client/iface/iface_test.go

@@ -569,17 +569,17 @@ func Test_ConnectPeers(t *testing.T) {
 	if err != nil {
 		t.Fatal(err)
 	}
-	// todo: investigate why in some tests execution we need 30s
+	// The peers use userspace WireGuard (stdnet transport). A tight busy-loop
+	// here starves the wireguard-go goroutines that process the handshake, so
+	// poll on a ticker instead and yield the CPU between checks. WireGuard also
+	// only retries a lost handshake initiation every REKEY_TIMEOUT (5s), which
+	// is why the overall wait can occasionally stretch to tens of seconds.
 	timeout := 30 * time.Second
 	timeoutChannel := time.After(timeout)
+	ticker := time.NewTicker(500 * time.Millisecond)
+	defer ticker.Stop()
 
 	for {
-		select {
-		case <-timeoutChannel:
-			t.Fatalf("waiting for peer handshake timeout after %s", timeout.String())
-		default:
-		}
-
 		peer, gpErr := getPeer(peer1ifaceName, peer2Key.PublicKey().String())
 		if gpErr != nil {
 			t.Fatal(gpErr)
@@ -588,6 +588,12 @@ func Test_ConnectPeers(t *testing.T) {
 			t.Log("peers successfully handshake")
 			break
 		}
+
+		select {
+		case <-timeoutChannel:
+			t.Fatalf("waiting for peer handshake timeout after %s", timeout.String())
+		case <-ticker.C:
+		}
 	}
 
 }

client/internal/auth/auth.go

@@ -351,6 +351,7 @@ func (a *Auth) setSystemInfoFlags(info *system.Info) {
 		a.config.BlockLANAccess,
 		a.config.BlockInbound,
 		a.config.DisableIPv6,
+		a.config.SyncMessageVersion,
 		a.config.EnableSSHRoot,
 		a.config.EnableSSHSFTP,
 		a.config.EnableSSHLocalPortForwarding,

client/internal/auth/sessionwatch/watcher.go

@@ -24,11 +24,7 @@ import (
 )
 
 const (
-	// Skew tolerates a small clock difference between the management
-	// server and this peer before treating a deadline as "in the past".
-	// Slightly above typical NTP drift; tight enough that the UI doesn't
-	// paint a stale expiry as if it were valid.
-	Skew = 30 * time.Second
+	maxPastHorizon = 30 * 24 * time.Hour
 
 	// maxDeadlineHorizon caps how far in the future an accepted deadline
 	// can sit. A timestamp beyond this is almost certainly a protocol
@@ -57,7 +53,7 @@ var (
 	ErrDeadlineTooFarFuture = errors.New("session deadline too far in the future")
 
 	// ErrDeadlineInPast is returned by Update when the supplied deadline
-	// is more than Skew in the past.
+	// is more than maxPastHorizon in the past.
 	ErrDeadlineInPast = errors.New("session deadline in the past")
 )
 
@@ -66,15 +62,14 @@ var (
 // for deadline change/clear, PublishEvent for the two warnings); tests pass
 // a fake recorder so the same surface is observable without an engine.
 //
-// The watcher is the single owner of the deadline propagated to the
-// recorder: every set, clear, sanity-check rejection and Close routes the
-// value through SetSessionExpiresAt, so the SubscribeStatus snapshot the UI
-// reads can never drift from the watcher's timer state. (SetSessionExpiresAt
-// fans out its own state-change notification, so no separate notify is
-// needed.) The recorder is server-scoped and outlives this engine-scoped
-// watcher — without the Close-time clear a teardown (Down, or the Down+Up of
-// a profile switch) would leave the next session showing the previous one's
-// stale "expires in" value.
+// While the watcher runs, it owns the deadline propagated to the recorder:
+// every set, clear and sanity-check rejection routes the value through
+// SetSessionExpiresAt, so the SubscribeStatus snapshot the UI reads can
+// never drift from the watcher's timer state. (SetSessionExpiresAt fans
+// out its own state-change notification, so no separate notify is needed.)
+// The recorder is server-scoped and outlives this engine-scoped watcher;
+// Close deliberately leaves the recorder value in place so transient engine
+// restarts don't blank it — the client run loop clears it on real teardown.
 //
 // PublishEvent's signature mirrors peer.Status.PublishEvent: the watcher
 // composes the metadata internally so the wire format (MetaSession*) is
@@ -135,10 +130,13 @@ func NewWithLeads(lead, final time.Duration, recorder StatusRecorder) *Watcher {
 // was disabled).
 //
 // Same-value updates are no-ops. A different non-zero value cancels any
-// pending timer, resets the "already fired" guard, and arms a new one.
+// pending timer, resets the "already fired" guards, and — when the
+// deadline lies in the future — arms fresh warning timers. A deadline
+// already in the past (within maxPastHorizon) is recorded as-is with no
+// timers: the session has expired and consumers render it that way.
 //
 // Returns one of the sentinel Err* values when the deadline fails the
-// sanity checks (pre-epoch, far future, or in the past beyond Skew).
+// sanity checks (pre-epoch, far future, or past beyond maxPastHorizon).
 // In every error case the watcher first clears its state so it stays
 // consistent with what the caller will push into its other sinks (e.g.
 // applySessionDeadline forces a zero deadline into the status recorder
@@ -163,7 +161,7 @@ func (w *Watcher) Update(deadline time.Time) error {
 	case deadline.After(now.Add(maxDeadlineHorizon)):
 		w.clearLocked()
 		return fmt.Errorf("%w: %v", ErrDeadlineTooFarFuture, deadline)
-	case deadline.Before(now.Add(-Skew)):
+	case deadline.Before(now.Add(-maxPastHorizon)):
 		w.clearLocked()
 		return fmt.Errorf("%w: %v (now=%v)", ErrDeadlineInPast, deadline, now)
 	}
@@ -183,7 +181,9 @@ func (w *Watcher) Update(deadline time.Time) error {
 	w.finalFiredAt = time.Time{}
 	w.dismissedAt = time.Time{}
 
-	w.armTimerLocked(deadline)
+	if deadline.After(now) {
+		w.armTimerLocked(deadline)
+	}
 	recorder := w.recorder
 	w.mu.Unlock()
 	if recorder != nil {
@@ -227,30 +227,25 @@ func (w *Watcher) Dismiss() {
 	log.Infof("auth session final-warning dismissed for deadline %s", w.current.Format(time.RFC3339))
 }
 
-// Close stops any pending timer and drops the deadline on the status
-// recorder. Update calls after Close are ignored. Clearing the recorder
-// here is what keeps a teardown (Down, or the Down+Up of a profile switch)
-// from leaving the next session showing this one's stale "expires in"
-// value — the recorder is server-scoped and outlives this engine-scoped
-// watcher, so nothing else drops the anchor on teardown.
+// Close stops any pending timer. Update calls after Close are ignored.
+// The recorder keeps its deadline: the watcher is engine-scoped and closes
+// on every engine restart (network change, sleep/wake, stream errors)
+// while the SSO deadline stays valid across those, so clearing here would
+// blank the UI's "expires in" row on every transient reconnect. The
+// client run loop clears the server-scoped recorder when it exits for
+// real (Down, profile switch, permanent login failure).
 func (w *Watcher) Close() {
 	w.mu.Lock()
+	defer w.mu.Unlock()
 	if w.closed {
-		w.mu.Unlock()
 		return
 	}
 	w.closed = true
 	w.stopTimerLocked()
-	hadDeadline := !w.current.IsZero()
 	w.current = time.Time{}
 	w.firedAt = time.Time{}
 	w.finalFiredAt = time.Time{}
 	w.dismissedAt = time.Time{}
-	recorder := w.recorder
-	w.mu.Unlock()
-	if recorder != nil && hadDeadline {
-		recorder.SetSessionExpiresAt(time.Time{})
-	}
 }
 
 // clearLocked drops the tracked deadline and notifies the recorder so

client/internal/auth/sessionwatch/watcher_test.go

@@ -224,11 +224,13 @@ func TestNewDeadlineCancelsPriorTimer(t *testing.T) {
 
 func TestRefreshAfterFireArmsNewWarning(t *testing.T) {
 	r := &fakeRecorder{}
-	lead := 30 * time.Millisecond
+	lead := 150 * time.Millisecond
 	w := newWatcher(lead, r)
 	defer w.Close()
 
-	first := time.Now().Add(50 * time.Millisecond)
+	// Warning fires ~20ms in; the deadline itself stays 150ms away so the
+	// replacement below lands well before it.
+	first := time.Now().Add(170 * time.Millisecond)
 	_ = w.Update(first)
 
 	// Wait for stateChange + warning of the first cycle.
@@ -306,7 +308,29 @@ func TestUpdateRejectsTooFarFuture(t *testing.T) {
 	}
 }
 
-func TestUpdateInPastClearsDeadline(t *testing.T) {
+func TestUpdateRecentPastRecordedAsExpired(t *testing.T) {
+	r := &fakeRecorder{}
+	w := newWatcher(50*time.Millisecond, r)
+	defer w.Close()
+
+	d := time.Now().Add(-1 * time.Hour)
+	if err := w.Update(d); err != nil {
+		t.Fatalf("recent-past Update should succeed, got %v", err)
+	}
+	if !w.Deadline().Equal(d) {
+		t.Fatalf("expected deadline to be recorded, got %v want %v", w.Deadline(), d)
+	}
+	if got := r.deadline(); !got.Equal(d) {
+		t.Fatalf("recorder deadline = %v, want %v", got, d)
+	}
+
+	time.Sleep(80 * time.Millisecond)
+	if n := countWhere(r.snapshot(), func(e event) bool { return e.kind == publish }); n != 0 {
+		t.Fatalf("no warning events may fire for an already-past deadline, got %+v", r.snapshot())
+	}
+}
+
+func TestUpdateAncientPastRejected(t *testing.T) {
 	r := &fakeRecorder{}
 	w := newWatcher(50*time.Millisecond, r)
 	defer w.Close()
@@ -318,12 +342,12 @@ func TestUpdateInPastClearsDeadline(t *testing.T) {
 	// Drain the stateChange from the seed.
 	waitForEvents(t, r, 1)
 
-	err := w.Update(time.Now().Add(-1 * time.Hour))
+	err := w.Update(time.Now().Add(-31 * 24 * time.Hour))
 	if !errors.Is(err, ErrDeadlineInPast) {
 		t.Fatalf("want ErrDeadlineInPast, got %v", err)
 	}
 	if !w.Deadline().IsZero() {
-		t.Fatalf("in-past update must clear the deadline, got %v", w.Deadline())
+		t.Fatalf("rejected ancient-past update must clear the deadline, got %v", w.Deadline())
 	}
 	events := waitForEvents(t, r, 2)
 	if events[1].kind != stateChange {
@@ -331,39 +355,25 @@ func TestUpdateInPastClearsDeadline(t *testing.T) {
 	}
 }
 
-func TestUpdateWithinSkewAccepted(t *testing.T) {
-	r := &fakeRecorder{}
-	w := newWatcher(50*time.Millisecond, r)
-	defer w.Close()
-
-	// 5 seconds in the past is within the 30s Skew tolerance — accept it.
-	d := time.Now().Add(-5 * time.Second)
-	if err := w.Update(d); err != nil {
-		t.Fatalf("within-skew Update should succeed, got %v", err)
-	}
-	if !w.Deadline().Equal(d) {
-		t.Fatalf("expected deadline to be applied, got %v want %v", w.Deadline(), d)
-	}
-}
-
 func TestCloseSilencesUpdates(t *testing.T) {
 	r := &fakeRecorder{}
 	w := newWatcher(50*time.Millisecond, r)
 	w.Close()
 
-	_ = w.Update(time.Now().Add(time.Hour))
-
-	time.Sleep(20 * time.Millisecond)
+	if err := w.Update(time.Now().Add(time.Hour)); err != nil {
+		t.Fatalf("Update after Close: want nil, got %v", err)
+	}
 	if got := r.snapshot(); len(got) != 0 {
 		t.Fatalf("expected no events after Close, got %+v", got)
 	}
 }
 
-// TestCloseClearsRecorderDeadline pins the profile-switch fix: a watcher
-// holding a live deadline must zero the recorder on Close so the next
-// engine's watcher (and the UI reading the shared server-scoped recorder)
-// doesn't start out showing the previous session's stale "expires in".
-func TestCloseClearsRecorderDeadline(t *testing.T) {
+// TestCloseKeepsRecorderDeadline pins the reconnect-flap fix: the watcher
+// closes on every engine restart (network change, sleep/wake) while the
+// SSO deadline stays valid across those, so Close must leave the
+// server-scoped recorder's value in place. The client run loop clears the
+// recorder when it exits for real.
+func TestCloseKeepsRecorderDeadline(t *testing.T) {
 	r := &fakeRecorder{}
 	w := newWatcher(time.Hour, r)
 
@@ -377,8 +387,8 @@ func TestCloseClearsRecorderDeadline(t *testing.T) {
 
 	w.Close()
 
-	if got := r.deadline(); !got.IsZero() {
-		t.Fatalf("recorder deadline after Close = %v, want zero", got)
+	if got := r.deadline(); !got.Equal(d) {
+		t.Fatalf("recorder deadline after Close = %v, want %v", got, d)
 	}
 }
 

client/internal/conn_mgr.go

@@ -34,6 +34,8 @@ const (
 // - Handling connection establishment based on peer signaling
 //
 // The implementation is not thread-safe; it is protected by engine.syncMsgMux.
+// The only exception is ActivatePeer, which is safe for concurrent use so the
+// DNS warm-up path can call it without contending on the engine mutex.
 type ConnMgr struct {
 	peerStore        *peerstore.Store
 	statusRecorder   *peer.Status
@@ -42,12 +44,26 @@ type ConnMgr struct {
 	rosenpassEnabled bool
 
 	lazyConnMgr *manager.Manager
+	// lazyConnMgrMu guards the lazyConnMgr pointer for readers outside the
+	// engine loop (ActivatePeer). Writers hold it in addition to
+	// engine.syncMsgMux; all other reads stay under engine.syncMsgMux only.
+	lazyConnMgrMu sync.RWMutex
+
+	// reconcileRoutedIPs re-applies a peer's routed allowed IPs after its lazy wake endpoint is
+	// (re)armed (Mode A at arm time). Injected by the engine; nil disables the reconcile.
+	reconcileRoutedIPs func(peerKey string) error
 
 	wg            sync.WaitGroup
 	lazyCtx       context.Context
 	lazyCtxCancel context.CancelFunc
 }
 
+// SetRoutedIPsReconciler injects the callback used to re-apply a peer's routed allowed IPs when
+// its lazy wake endpoint is (re)armed. Must be called before the lazy manager starts.
+func (e *ConnMgr) SetRoutedIPsReconciler(fn func(peerKey string) error) {
+	e.reconcileRoutedIPs = fn
+}
+
 func NewConnMgr(engineConfig *EngineConfig, statusRecorder *peer.Status, peerStore *peerstore.Store, iface lazyconn.WGIface) *ConnMgr {
 	e := &ConnMgr{
 		peerStore:        peerStore,
@@ -238,12 +254,20 @@ func (e *ConnMgr) RemovePeerConn(peerKey string) {
 	conn.Log.Infof("removed peer from lazy conn manager")
 }
 
+// ActivatePeer wakes an idle lazy connection. Unlike the rest of ConnMgr it is
+// safe for concurrent use: the lazy manager pointer is read under lazyConnMgrMu
+// and the manager itself is internally synchronized, so callers outside the
+// engine loop (DNS warm-up) do not need engine.syncMsgMux.
 func (e *ConnMgr) ActivatePeer(ctx context.Context, conn *peer.Conn) {
-	if !e.isStartedWithLazyMgr() {
+	e.lazyConnMgrMu.RLock()
+	lazyConnMgr := e.lazyConnMgr
+	started := lazyConnMgr != nil && e.lazyCtxCancel != nil
+	e.lazyConnMgrMu.RUnlock()
+	if !started {
 		return
 	}
 
-	if found := e.lazyConnMgr.ActivatePeer(conn.GetKey()); found {
+	if found := lazyConnMgr.ActivatePeer(conn.GetKey()); found {
 		if err := conn.Open(ctx); err != nil {
 			conn.Log.Errorf("failed to open connection: %v", err)
 		}
@@ -268,16 +292,22 @@ func (e *ConnMgr) Close() {
 
 	e.lazyCtxCancel()
 	e.wg.Wait()
+
+	e.lazyConnMgrMu.Lock()
 	e.lazyConnMgr = nil
+	e.lazyConnMgrMu.Unlock()
 }
 
 func (e *ConnMgr) initLazyManager(engineCtx context.Context) {
 	cfg := manager.Config{
 		InactivityThreshold: inactivityThresholdEnv(),
+		ReconcileAllowedIPs: e.reconcileRoutedIPs,
 	}
-	e.lazyConnMgr = manager.NewManager(cfg, engineCtx, e.peerStore, e.iface)
 
+	e.lazyConnMgrMu.Lock()
+	e.lazyConnMgr = manager.NewManager(cfg, engineCtx, e.peerStore, e.iface)
 	e.lazyCtx, e.lazyCtxCancel = context.WithCancel(engineCtx)
+	e.lazyConnMgrMu.Unlock()
 
 	e.wg.Add(1)
 	go func() {
@@ -316,7 +346,10 @@ func (e *ConnMgr) closeManager(ctx context.Context) {
 
 	e.lazyCtxCancel()
 	e.wg.Wait()
+
+	e.lazyConnMgrMu.Lock()
 	e.lazyConnMgr = nil
+	e.lazyConnMgrMu.Unlock()
 
 	for _, peerID := range e.peerStore.PeersPubKey() {
 		e.peerStore.PeerConnOpen(ctx, peerID)

client/internal/conn_mgr_test.go

@@ -1,10 +1,21 @@
 package internal
 
 import (
+	"context"
+	"net"
+	"net/netip"
 	"os"
+	"sync"
 	"testing"
+	"time"
 
+	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
+
+	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/internal/lazyconn"
+	"github.com/netbirdio/netbird/client/internal/peer"
+	"github.com/netbirdio/netbird/client/internal/peerstore"
+	"github.com/netbirdio/netbird/monotime"
 )
 
 func TestResolveLazyForce(t *testing.T) {
@@ -38,3 +49,58 @@ func TestResolveLazyForce(t *testing.T) {
 		})
 	}
 }
+
+type mockLazyWGIface struct{}
+
+func (mockLazyWGIface) RemovePeer(string) error { return nil }
+func (mockLazyWGIface) UpdatePeer(string, []netip.Prefix, time.Duration, *net.UDPAddr, *wgtypes.Key) error {
+	return nil
+}
+func (mockLazyWGIface) IsUserspaceBind() bool                    { return false }
+func (mockLazyWGIface) Address() wgaddr.Address                  { return wgaddr.Address{} }
+func (mockLazyWGIface) LastActivities() map[string]monotime.Time { return nil }
+func (mockLazyWGIface) MTU() uint16                              { return 1280 }
+
+// TestConnMgr_ActivatePeerConcurrentWithLifecycle exercises ActivatePeer from
+// non-engine goroutines (the DNS warm-up path) racing the manager lifecycle,
+// which stays on the engine loop. Run with -race: it fails if ActivatePeer
+// still requires engine.syncMsgMux for safety.
+func TestConnMgr_ActivatePeerConcurrentWithLifecycle(t *testing.T) {
+	t.Setenv(lazyconn.EnvLazyConn, "on")
+
+	status := peer.NewRecorder("https://mgm")
+	store := peerstore.NewConnStore()
+	connMgr := NewConnMgr(&EngineConfig{}, status, store, mockLazyWGIface{})
+
+	conn := newTestPeerConn(t, "peerA")
+
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+	connMgr.Start(ctx)
+
+	done := make(chan struct{})
+	var wg sync.WaitGroup
+	for range 4 {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			for {
+				select {
+				case <-done:
+					return
+				default:
+					connMgr.ActivatePeer(ctx, conn)
+				}
+			}
+		}()
+	}
+
+	// Let the activators spin against the started manager, then tear it down
+	// underneath them and let them spin against the stopped manager.
+	time.Sleep(100 * time.Millisecond)
+	connMgr.Close()
+	time.Sleep(50 * time.Millisecond)
+
+	close(done)
+	wg.Wait()
+}

client/internal/connect.go

@@ -257,7 +257,10 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		log.Errorf("failed to clean up temporary installer file: %v", err)
 	}
 
-	defer c.statusRecorder.ClientStop()
+	defer func() {
+		c.statusRecorder.SetSessionExpiresAt(time.Time{})
+		c.statusRecorder.ClientStop()
+	}()
 	operation := func() error {
 		// if context cancelled we not start new backoff cycle
 		if c.ctx.Err() != nil {
@@ -618,6 +621,7 @@ func createEngineConfig(key wgtypes.Key, config *profilemanager.Config, peerConf
 		BlockLANAccess:      config.BlockLANAccess,
 		BlockInbound:        config.BlockInbound,
 		DisableIPv6:         config.DisableIPv6,
+		SyncMessageVersion:  config.SyncMessageVersion,
 
 		LazyConnection: lazyconn.ParseState(config.LazyConnection),
 
@@ -693,6 +697,7 @@ func loginToManagement(ctx context.Context, client mgm.Client, pubSSHKey []byte,
 		config.BlockLANAccess,
 		config.BlockInbound,
 		config.DisableIPv6,
+		config.SyncMessageVersion,
 		config.EnableSSHRoot,
 		config.EnableSSHSFTP,
 		config.EnableSSHLocalPortForwarding,

client/internal/debug/debug.go

@@ -676,6 +676,7 @@ func (g *BundleGenerator) addCommonConfigFields(configContent *strings.Builder)
 	configContent.WriteString(fmt.Sprintf("BlockLANAccess: %v\n", g.internalConfig.BlockLANAccess))
 	configContent.WriteString(fmt.Sprintf("BlockInbound: %v\n", g.internalConfig.BlockInbound))
 	configContent.WriteString(fmt.Sprintf("DisableIPv6: %v\n", g.internalConfig.DisableIPv6))
+	configContent.WriteString(fmt.Sprintf("SyncMessageVersion: %v\n", g.internalConfig.SyncMessageVersion))
 
 	if g.internalConfig.DisableNotifications != nil {
 		configContent.WriteString(fmt.Sprintf("DisableNotifications: %v\n", *g.internalConfig.DisableNotifications))

client/internal/debug/debug_test.go

@@ -887,6 +887,8 @@ func TestAddConfig_AllFieldsCovered(t *testing.T) {
 		ClientCertKeyPath:             "/tmp/key",
 		LazyConnection:                "on",
 		MTU:                           1280,
+		DisableIPv6:                   true,
+		SyncMessageVersion:            func(v int) *int { return &v }(1),
 	}
 
 	for _, anonymize := range []bool{false, true} {

client/internal/dns/local/local.go

@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"net"
 	"net/netip"
+	"os"
 	"slices"
 	"strings"
 	"sync"
@@ -36,7 +37,43 @@ type resolver interface {
 // record is left alone (it points at something outside our mesh, e.g.
 // a non-peer upstream).
 type PeerConnectivity interface {
-	IsConnectedByIP(ip string) (known, connected bool)
+	IsConnectedByIP(ip netip.Addr) (known, connected bool)
+}
+
+// PeerActivator wakes lazy-connection peers on demand. The local resolver calls
+// it with the tunnel IPs an answer points at, so a peer that is idle (lazily
+// disconnected) starts connecting at DNS-resolution time rather than racing the
+// client's first request packet. nil disables warm-up.
+type PeerActivator interface {
+	// ActivatePeersByIP triggers wake-up for the peer(s) owning addrs and blocks
+	// until one is connected or ctx (a short per-query budget) expires. It is a
+	// fast no-op for unknown or already-connected addresses.
+	ActivatePeersByIP(ctx context.Context, addrs []netip.Addr)
+}
+
+const (
+	defaultLazyWarmupTimeout = 2 * time.Second
+	envLazyWarmupTimeout     = "NB_DNS_LAZY_WARMUP_TIMEOUT"
+)
+
+// lazyWarmupTimeoutFromEnv returns the per-query budget for waking a
+// lazy-connection peer a DNS answer points at. Tunable via
+// NB_DNS_LAZY_WARMUP_TIMEOUT (a Go duration). Parsed once at construction time.
+func lazyWarmupTimeoutFromEnv() time.Duration {
+	v := os.Getenv(envLazyWarmupTimeout)
+	if v == "" {
+		return defaultLazyWarmupTimeout
+	}
+	d, err := time.ParseDuration(v)
+	if err != nil {
+		log.Warnf("invalid %s value %q, using default %s: %v", envLazyWarmupTimeout, v, defaultLazyWarmupTimeout, err)
+		return defaultLazyWarmupTimeout
+	}
+	if d <= 0 {
+		log.Warnf("non-positive %s value %q, using default %s", envLazyWarmupTimeout, v, defaultLazyWarmupTimeout)
+		return defaultLazyWarmupTimeout
+	}
+	return d
 }
 
 type Resolver struct {
@@ -51,6 +88,12 @@ type Resolver struct {
 	// filter and preserves the legacy "return whatever is registered"
 	// behaviour for callers that never wire a status source.
 	peerConn PeerConnectivity
+	// peerActivator, when non-nil, is called at resolution time to warm the
+	// lazy connection to the peer(s) an answer points at. nil disables warm-up.
+	peerActivator PeerActivator
+	// warmupTimeout is the per-query budget for the lazy-connection warm-up
+	// wait, resolved from the environment once at construction time.
+	warmupTimeout time.Duration
 
 	ctx    context.Context
 	cancel context.CancelFunc
@@ -59,11 +102,12 @@ type Resolver struct {
 func NewResolver() *Resolver {
 	ctx, cancel := context.WithCancel(context.Background())
 	return &Resolver{
-		records: make(map[dns.Question][]dns.RR),
-		domains: make(map[domain.Domain]struct{}),
-		zones:   make(map[domain.Domain]bool),
-		ctx:     ctx,
-		cancel:  cancel,
+		records:       make(map[dns.Question][]dns.RR),
+		domains:       make(map[domain.Domain]struct{}),
+		zones:         make(map[domain.Domain]bool),
+		warmupTimeout: lazyWarmupTimeoutFromEnv(),
+		ctx:           ctx,
+		cancel:        cancel,
 	}
 }
 
@@ -76,6 +120,14 @@ func (d *Resolver) SetPeerConnectivity(p PeerConnectivity) {
 	d.peerConn = p
 }
 
+// SetPeerActivator wires the DNS-time lazy-connection warm-up. Pass nil to
+// disable. Safe to call multiple times; the latest value wins.
+func (d *Resolver) SetPeerActivator(a PeerActivator) {
+	d.mu.Lock()
+	defer d.mu.Unlock()
+	d.peerActivator = a
+}
+
 func (d *Resolver) MatchSubdomains() bool {
 	return true
 }
@@ -122,6 +174,9 @@ func (d *Resolver) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 	replyMessage.RecursionAvailable = true
 
 	result := d.lookupRecords(logger, question)
+	// Warm before filtering: activation flips a lazily-idle target to connected,
+	// which then lets it survive the disconnected-peer filter below.
+	d.warmLazyPeers(question, result.records)
 	result.records = d.filterDisconnectedPeerAnswers(logger, question, result.records)
 	replyMessage.Authoritative = !result.hasExternalData
 	replyMessage.Answer = result.records
@@ -495,8 +550,8 @@ func (d *Resolver) filterDisconnectedPeerAnswers(logger *log.Entry, question dns
 	kept := make([]dns.RR, 0, len(records))
 	var dropped int
 	for _, rr := range records {
-		ip := extractRecordIP(rr)
-		if ip == "" {
+		ip, ok := extractRecordAddr(rr)
+		if !ok {
 			kept = append(kept, rr)
 			continue
 		}
@@ -518,22 +573,57 @@ func (d *Resolver) filterDisconnectedPeerAnswers(logger *log.Entry, question dns
 	return kept
 }
 
-// extractRecordIP returns the dotted-decimal / colon-hex IP carried by
-// an A or AAAA record, or "" for any other record type.
-func extractRecordIP(rr dns.RR) string {
+// warmLazyPeers triggers lazy-connection wake-up for the peers a resolved
+// answer points at and waits briefly for one to connect, so the caller's first
+// request doesn't race the connection establishment. Warm-up is scoped to
+// match-only (non-authoritative) zones — the synthesized private-service zones
+// and user-created zones whose records point at specific peers. The account's
+// peer zone is authoritative, so plain peer-name lookups never trigger warm-up;
+// otherwise resolving any peer's name would wake its idle connection, defeating
+// laziness mesh-wide. No-op when no activator is wired (lazy connections
+// disabled) or the answer carries no peer IPs.
+func (d *Resolver) warmLazyPeers(question dns.Question, records []dns.RR) {
+	if len(records) < 2 {
+		return
+	}
+	d.mu.RLock()
+	activator := d.peerActivator
+	var nonAuth, found bool
+	if activator != nil {
+		nonAuth, found = d.findZone(question.Name)
+	}
+	d.mu.RUnlock()
+	if activator == nil || !found || !nonAuth {
+		return
+	}
+
+	var addrs []netip.Addr
+	for _, rr := range records {
+		if addr, ok := extractRecordAddr(rr); ok {
+			addrs = append(addrs, addr)
+		}
+	}
+	if len(addrs) == 0 {
+		return
+	}
+
+	ctx, cancel := context.WithTimeout(d.ctx, d.warmupTimeout)
+	defer cancel()
+	activator.ActivatePeersByIP(ctx, addrs)
+}
+
+// extractRecordAddr returns the IP address carried by an A or AAAA record.
+// ok is false for any other record type or a record with no address.
+func extractRecordAddr(rr dns.RR) (netip.Addr, bool) {
 	switch r := rr.(type) {
 	case *dns.A:
-		if r.A == nil {
-			return ""
-		}
-		return r.A.String()
+		addr, ok := netip.AddrFromSlice(r.A)
+		return addr.Unmap(), ok
 	case *dns.AAAA:
-		if r.AAAA == nil {
-			return ""
-		}
-		return r.AAAA.String()
+		addr, ok := netip.AddrFromSlice(r.AAAA)
+		return addr.Unmap(), ok
 	}
-	return ""
+	return netip.Addr{}, false
 }
 
 // Update replaces all zones and their records

client/internal/dns/local/local_test.go

@@ -37,8 +37,8 @@ type mockPeerConnectivity struct {
 	byIP map[string]struct{ known, connected bool }
 }
 
-func (m mockPeerConnectivity) IsConnectedByIP(ip string) (known, connected bool) {
-	v, ok := m.byIP[ip]
+func (m mockPeerConnectivity) IsConnectedByIP(ip netip.Addr) (known, connected bool) {
+	v, ok := m.byIP[ip.String()]
 	if !ok {
 		return false, false
 	}

client/internal/dns/local/warmup_test.go

@@ -0,0 +1,204 @@
+package local
+
+import (
+	"context"
+	"net"
+	"net/netip"
+	"sync"
+	"testing"
+	"time"
+
+	"github.com/miekg/dns"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"github.com/netbirdio/netbird/client/internal/dns/test"
+	nbdns "github.com/netbirdio/netbird/dns"
+)
+
+// recordingActivator records the addresses it was asked to warm and returns
+// immediately, so ServeDNS is not blocked by the test.
+type recordingActivator struct {
+	mu     sync.Mutex
+	called bool
+	addrs  []netip.Addr
+}
+
+func (r *recordingActivator) ActivatePeersByIP(_ context.Context, addrs []netip.Addr) {
+	r.mu.Lock()
+	defer r.mu.Unlock()
+	r.called = true
+	r.addrs = append(r.addrs, addrs...)
+}
+
+func serveA(t *testing.T, resolver *Resolver, name string) *dns.Msg {
+	t.Helper()
+	var resp *dns.Msg
+	w := &test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}
+	resolver.ServeDNS(w, new(dns.Msg).SetQuestion(name, dns.TypeA))
+	return resp
+}
+
+// serviceZone registers rec in a match-only (non-authoritative) zone, the shape
+// the synthesized private-service zones arrive in.
+func serviceZone(t *testing.T, resolver *Resolver, zone string, records ...nbdns.SimpleRecord) {
+	t.Helper()
+	resolver.Update([]nbdns.CustomZone{{
+		Domain:           zone,
+		Records:          records,
+		NonAuthoritative: true,
+	}})
+}
+
+func TestLocalResolver_WarmsLazyPeerOnResolve(t *testing.T) {
+	// Warm-up fires only for multi-record answers (the HA / round-robin shape of
+	// the synthesized private-service zones), so register two peer targets.
+	const name = "svc.proxy.netbird.cloud."
+	recs := []nbdns.SimpleRecord{
+		{Name: name, Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.7"},
+		{Name: name, Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.8"},
+	}
+	resolver := NewResolver()
+	serviceZone(t, resolver, "proxy.netbird.cloud", recs...)
+
+	act := &recordingActivator{}
+	resolver.SetPeerActivator(act)
+
+	resp := serveA(t, resolver, name)
+	require.NotNil(t, resp, "resolver must answer")
+	require.NotEmpty(t, resp.Answer, "answer must carry the A records")
+
+	act.mu.Lock()
+	defer act.mu.Unlock()
+	assert.True(t, act.called, "activator must be invoked for a multi-record service-zone answer")
+	assert.Contains(t, act.addrs, netip.MustParseAddr("100.64.0.7"), "activator must receive the first peer IP")
+	assert.Contains(t, act.addrs, netip.MustParseAddr("100.64.0.8"), "activator must receive the second peer IP")
+}
+
+func TestLocalResolver_NoWarmupForSingleRecord(t *testing.T) {
+	// A single-record answer does not trigger warm-up; the resolver only warms
+	// multi-record answers.
+	rec := nbdns.SimpleRecord{Name: "svc.proxy.netbird.cloud.", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.7"}
+	resolver := NewResolver()
+	serviceZone(t, resolver, "proxy.netbird.cloud", rec)
+
+	act := &recordingActivator{}
+	resolver.SetPeerActivator(act)
+
+	resp := serveA(t, resolver, rec.Name)
+	require.NotNil(t, resp, "resolver must answer")
+	require.NotEmpty(t, resp.Answer, "answer must carry the A record")
+
+	act.mu.Lock()
+	defer act.mu.Unlock()
+	assert.False(t, act.called, "activator must not be invoked for a single-record answer")
+}
+
+func TestLocalResolver_NoActivatorNoWarmup(t *testing.T) {
+	// With no activator wired the resolver behaves exactly as before.
+	rec := nbdns.SimpleRecord{Name: "svc.proxy.netbird.cloud.", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.7"}
+	resolver := NewResolver()
+	serviceZone(t, resolver, "proxy.netbird.cloud", rec)
+
+	resp := serveA(t, resolver, rec.Name)
+	require.NotNil(t, resp, "resolver must still answer without an activator")
+	require.NotEmpty(t, resp.Answer, "answer must carry the A record")
+}
+
+func TestLocalResolver_NoWarmupForMissingRecord(t *testing.T) {
+	// A query that resolves to nothing must not invoke the activator (no IPs).
+	resolver := NewResolver()
+	serviceZone(t, resolver, "proxy.netbird.cloud",
+		nbdns.SimpleRecord{Name: "svc.proxy.netbird.cloud.", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.7"})
+
+	act := &recordingActivator{}
+	resolver.SetPeerActivator(act)
+
+	serveA(t, resolver, "absent.proxy.netbird.cloud.")
+
+	act.mu.Lock()
+	defer act.mu.Unlock()
+	assert.False(t, act.called, "activator must not be invoked when there is no answer")
+}
+
+func TestLocalResolver_NoWarmupInAuthoritativeZone(t *testing.T) {
+	// The account's peer zone is authoritative; resolving a peer's name there
+	// must not wake its lazy connection — warm-up is scoped to match-only
+	// (non-authoritative) zones such as the synthesized private-service zones.
+	// Use a multi-record answer so the authoritative-zone scoping is the only
+	// reason warm-up is skipped, not the single-record guard.
+	const name = "peer.netbird.cloud."
+	recs := []nbdns.SimpleRecord{
+		{Name: name, Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.9"},
+		{Name: name, Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "100.64.0.10"},
+	}
+	resolver := NewResolver()
+	resolver.Update([]nbdns.CustomZone{{
+		Domain:  "netbird.cloud",
+		Records: recs,
+	}})
+
+	act := &recordingActivator{}
+	resolver.SetPeerActivator(act)
+
+	resp := serveA(t, resolver, name)
+	require.NotNil(t, resp, "resolver must answer")
+	require.NotEmpty(t, resp.Answer, "answer must carry the A records")
+
+	act.mu.Lock()
+	defer act.mu.Unlock()
+	assert.False(t, act.called, "activator must not be invoked for authoritative-zone answers")
+}
+
+func TestLazyWarmupTimeoutFromEnv(t *testing.T) {
+	tests := []struct {
+		name   string
+		value  string
+		envSet bool
+		want   time.Duration
+	}{
+		{name: "unset uses default", want: defaultLazyWarmupTimeout},
+		{name: "valid overrides", value: "5s", envSet: true, want: 5 * time.Second},
+		{name: "invalid falls back", value: "not-a-duration", envSet: true, want: defaultLazyWarmupTimeout},
+		{name: "negative falls back", value: "-1s", envSet: true, want: defaultLazyWarmupTimeout},
+		{name: "zero falls back", value: "0s", envSet: true, want: defaultLazyWarmupTimeout},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if tt.envSet {
+				t.Setenv(envLazyWarmupTimeout, tt.value)
+			}
+			assert.Equal(t, tt.want, lazyWarmupTimeoutFromEnv())
+			assert.Equal(t, tt.want, NewResolver().warmupTimeout, "constructor must resolve the timeout once")
+		})
+	}
+}
+
+func TestExtractRecordAddr(t *testing.T) {
+	t.Run("A record yields unmapped v4", func(t *testing.T) {
+		// net.ParseIP returns the 16-byte v4-in-v6 form, the same shape
+		// miekg/dns stores after parsing an A record; the extracted address
+		// must compare equal to a plain v4 netip.Addr.
+		addr, ok := extractRecordAddr(&dns.A{A: net.ParseIP("100.64.0.7")})
+		require.True(t, ok)
+		assert.True(t, addr.Is4())
+		assert.Equal(t, netip.MustParseAddr("100.64.0.7"), addr)
+	})
+
+	t.Run("AAAA record yields v6", func(t *testing.T) {
+		addr, ok := extractRecordAddr(&dns.AAAA{AAAA: net.ParseIP("fd00::1")})
+		require.True(t, ok)
+		assert.Equal(t, netip.MustParseAddr("fd00::1"), addr)
+	})
+
+	t.Run("A record without address", func(t *testing.T) {
+		_, ok := extractRecordAddr(&dns.A{})
+		assert.False(t, ok)
+	})
+
+	t.Run("non-address record", func(t *testing.T) {
+		_, ok := extractRecordAddr(&dns.CNAME{Target: "target.netbird.cloud."})
+		assert.False(t, ok)
+	})
+}

client/internal/dns/mock_server.go

@@ -8,6 +8,7 @@ import (
 	"github.com/miekg/dns"
 
 	dnsconfig "github.com/netbirdio/netbird/client/internal/dns/config"
+	"github.com/netbirdio/netbird/client/internal/dns/local"
 	nbdns "github.com/netbirdio/netbird/dns"
 	"github.com/netbirdio/netbird/route"
 	"github.com/netbirdio/netbird/shared/management/domain"
@@ -92,6 +93,11 @@ func (m *MockServer) SetFirewall(Firewall) {
 	// Mock implementation - no-op
 }
 
+// SetPeerActivator mock implementation of SetPeerActivator from Server interface
+func (m *MockServer) SetPeerActivator(local.PeerActivator) {
+	// Mock implementation - no-op
+}
+
 // BeginBatch mock implementation of BeginBatch from Server interface
 func (m *MockServer) BeginBatch() {
 	// Mock implementation - no-op

client/internal/dns/server.go

@@ -82,6 +82,7 @@ type Server interface {
 	PopulateManagementDomain(mgmtURL *url.URL) error
 	SetRouteSources(selected, active func() route.HAMap)
 	SetFirewall(Firewall)
+	SetPeerActivator(local.PeerActivator)
 }
 
 type nsGroupsByDomain struct {
@@ -491,6 +492,13 @@ func (s *DefaultServer) SetFirewall(fw Firewall) {
 	}
 }
 
+// SetPeerActivator wires the DNS-time lazy-connection warm-up on the local
+// resolver. Injected after the connection manager exists (it does not at
+// DNS-server construction time). Pass nil to disable.
+func (s *DefaultServer) SetPeerActivator(a local.PeerActivator) {
+	s.localResolver.SetPeerActivator(a)
+}
+
 // Stop stops the server
 func (s *DefaultServer) Stop() {
 	s.ctxCancel()
@@ -1435,11 +1443,11 @@ type localPeerConnectivity struct {
 
 // IsConnectedByIP looks the IP up in the peerstore and surfaces both
 // the known and connected bits. Used by Resolver.filterDisconnectedPeerAnswers.
-func (l localPeerConnectivity) IsConnectedByIP(ip string) (known, connected bool) {
+func (l localPeerConnectivity) IsConnectedByIP(ip netip.Addr) (known, connected bool) {
 	if l.status == nil {
 		return false, false
 	}
-	state, ok := l.status.PeerStateByIP(ip)
+	state, ok := l.status.PeerStateByIP(ip.String())
 	if !ok {
 		return false, false
 	}

client/internal/dns_peer_activator.go

@@ -0,0 +1,76 @@
+package internal
+
+import (
+	"context"
+	"net/netip"
+	"time"
+
+	"github.com/netbirdio/netbird/client/internal/peer"
+	"github.com/netbirdio/netbird/client/internal/peerstore"
+)
+
+const dnsActivationPollInterval = 50 * time.Millisecond
+
+// dnsPeerActivator wakes lazy-connection peers from the DNS resolution path. It
+// implements dns/local.PeerActivator. DNS queries run on their own goroutines,
+// so it only touches state that is safe for concurrent use — ConnMgr.ActivatePeer,
+// peerstore.Store and peer.Status — and never takes the engine's syncMsgMux,
+// keeping DNS resolution from contending with network-map processing.
+type dnsPeerActivator struct {
+	connMgr   *ConnMgr
+	peerStore *peerstore.Store
+	status    *peer.Status
+	// ctx is the engine's long-lived context. The connection dial is tied to it
+	// (not the per-query DNS wait budget) so a handshake that outlasts the wait
+	// still completes in the background rather than being cancelled at the deadline.
+	ctx context.Context
+}
+
+// ActivatePeersByIP triggers wake-up for the peer(s) owning addrs and waits
+// until one is connected or ctx (the per-query DNS wait budget) expires.
+// Activation itself is tied to the engine's long-lived context so the dial
+// survives a wait that times out. Unknown or already-connected addresses are
+// skipped, so the steady-state (warm) path adds no latency.
+func (a *dnsPeerActivator) ActivatePeersByIP(ctx context.Context, addrs []netip.Addr) {
+	if a == nil || a.connMgr == nil {
+		return
+	}
+
+	var pending []string
+	for _, addr := range addrs {
+		ip := addr.String()
+		st, ok := a.status.PeerStateByIP(ip)
+		if !ok || st.ConnStatus == peer.StatusConnected {
+			continue
+		}
+		conn, ok := a.peerStore.PeerConn(st.PubKey)
+		if !ok {
+			continue
+		}
+		a.connMgr.ActivatePeer(a.ctx, conn)
+		pending = append(pending, ip)
+	}
+
+	if len(pending) == 0 {
+		return
+	}
+	a.waitConnected(ctx, pending)
+}
+
+// waitConnected blocks until any of ips reports a connected peer or ctx expires.
+func (a *dnsPeerActivator) waitConnected(ctx context.Context, ips []string) {
+	ticker := time.NewTicker(dnsActivationPollInterval)
+	defer ticker.Stop()
+	for {
+		for _, ip := range ips {
+			if st, ok := a.status.PeerStateByIP(ip); ok && st.ConnStatus == peer.StatusConnected {
+				return
+			}
+		}
+		select {
+		case <-ctx.Done():
+			return
+		case <-ticker.C:
+		}
+	}
+}

client/internal/dns_peer_activator_test.go

@@ -0,0 +1,129 @@
+package internal
+
+import (
+	"context"
+	"net/netip"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/netbirdio/netbird/client/internal/peer"
+	"github.com/netbirdio/netbird/client/internal/peerstore"
+)
+
+func newTestPeerConn(t *testing.T, key string) *peer.Conn {
+	t.Helper()
+	conn, err := peer.NewConn(peer.ConnConfig{
+		Key:      key,
+		LocalKey: "local",
+		WgConfig: peer.WgConfig{
+			AllowedIps: []netip.Prefix{netip.MustParsePrefix("100.64.0.1/32")},
+		},
+	}, peer.ServiceDependencies{})
+	require.NoError(t, err)
+	return conn
+}
+
+func newTestDNSPeerActivator(t *testing.T) (*dnsPeerActivator, *peer.Status, *peerstore.Store) {
+	t.Helper()
+	status := peer.NewRecorder("https://mgm")
+	store := peerstore.NewConnStore()
+	// ConnMgr without Start: the lazy manager is nil, so ActivatePeer is a
+	// no-op — these tests exercise the activator's skip/wait logic.
+	connMgr := NewConnMgr(&EngineConfig{}, status, store, nil)
+	return &dnsPeerActivator{
+		connMgr:   connMgr,
+		peerStore: store,
+		status:    status,
+		ctx:       context.Background(),
+	}, status, store
+}
+
+func TestDNSPeerActivator_NilSafe(t *testing.T) {
+	var a *dnsPeerActivator
+	a.ActivatePeersByIP(context.Background(), []netip.Addr{netip.MustParseAddr("100.64.0.1")})
+}
+
+// TestDNSPeerActivator_SkipsUnknownAndConnectedPeers verifies the steady-state
+// (warm) path adds no latency: already-connected and unknown addresses never
+// enter the wait loop.
+func TestDNSPeerActivator_SkipsUnknownAndConnectedPeers(t *testing.T) {
+	a, status, store := newTestDNSPeerActivator(t)
+
+	require.NoError(t, status.AddPeer("peerA", "a.netbird.cloud", "100.64.0.1", "fd00::1"))
+	require.NoError(t, status.UpdatePeerState(peer.State{PubKey: "peerA", ConnStatus: peer.StatusConnected}))
+	store.AddPeerConn("peerA", newTestPeerConn(t, "peerA"))
+
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	start := time.Now()
+	a.ActivatePeersByIP(ctx, []netip.Addr{
+		netip.MustParseAddr("100.64.0.1"),  // known, connected -> skipped
+		netip.MustParseAddr("fd00::1"),     // known via IPv6, connected -> skipped
+		netip.MustParseAddr("100.64.0.99"), // unknown -> skipped
+	})
+	require.Less(t, time.Since(start), time.Second, "no pending peer must mean no wait")
+}
+
+// TestDNSPeerActivator_WaitsForPendingPeerToConnect verifies the wait loop
+// returns as soon as a pending peer reports connected, well before the
+// per-query budget expires.
+func TestDNSPeerActivator_WaitsForPendingPeerToConnect(t *testing.T) {
+	a, status, store := newTestDNSPeerActivator(t)
+
+	require.NoError(t, status.AddPeer("peerA", "a.netbird.cloud", "100.64.0.1", ""))
+	store.AddPeerConn("peerA", newTestPeerConn(t, "peerA"))
+
+	go func() {
+		time.Sleep(150 * time.Millisecond)
+		_ = status.UpdatePeerState(peer.State{PubKey: "peerA", ConnStatus: peer.StatusConnected})
+	}()
+
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	start := time.Now()
+	a.ActivatePeersByIP(ctx, []netip.Addr{netip.MustParseAddr("100.64.0.1")})
+	elapsed := time.Since(start)
+
+	require.GreaterOrEqual(t, elapsed, 100*time.Millisecond, "must wait for the pending peer")
+	require.Less(t, elapsed, 5*time.Second, "must return on connect, not at the deadline")
+}
+
+// TestDNSPeerActivator_ReturnsAtBudgetWhenPeerStaysIdle verifies a peer that
+// never connects releases the DNS response at the per-query budget instead of
+// blocking it indefinitely.
+func TestDNSPeerActivator_ReturnsAtBudgetWhenPeerStaysIdle(t *testing.T) {
+	a, status, store := newTestDNSPeerActivator(t)
+
+	require.NoError(t, status.AddPeer("peerA", "a.netbird.cloud", "100.64.0.1", ""))
+	store.AddPeerConn("peerA", newTestPeerConn(t, "peerA"))
+
+	ctx, cancel := context.WithTimeout(context.Background(), 300*time.Millisecond)
+	defer cancel()
+
+	start := time.Now()
+	a.ActivatePeersByIP(ctx, []netip.Addr{netip.MustParseAddr("100.64.0.1")})
+	elapsed := time.Since(start)
+
+	require.GreaterOrEqual(t, elapsed, 250*time.Millisecond, "must wait out the budget for a pending peer")
+	require.Less(t, elapsed, 5*time.Second, "must not block past the budget")
+}
+
+// TestDNSPeerActivator_NoWaitWithoutPeerConn verifies a known-but-idle peer
+// with no connection object in the store is not waited on: there is nothing to
+// activate, so waiting could only ever time out.
+func TestDNSPeerActivator_NoWaitWithoutPeerConn(t *testing.T) {
+	a, status, _ := newTestDNSPeerActivator(t)
+
+	require.NoError(t, status.AddPeer("peerA", "a.netbird.cloud", "100.64.0.1", ""))
+
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
+	defer cancel()
+
+	start := time.Now()
+	a.ActivatePeersByIP(ctx, []netip.Addr{netip.MustParseAddr("100.64.0.1")})
+	require.Less(t, time.Since(start), time.Second, "peer without a conn must not be waited on")
+}

client/internal/ebpf/ebpf/src/dns_fwd.c

@@ -52,11 +52,14 @@ int xdp_dns_fwd(struct iphdr  *ip, struct udphdr *udp) {
 
     if (udp->dest == GENERAL_DNS_PORT && ip->daddr == dns_ip) {
         udp->dest = dns_port;
+        // Clear the now-stale checksum; zero means "not computed" for IPv4.
+        udp->check = 0;
         return XDP_PASS;
     }
 
     if (udp->source == dns_port && ip->saddr == dns_ip) {
         udp->source = GENERAL_DNS_PORT;
+        udp->check = 0;
         return XDP_PASS;
     }
 

client/internal/ebpf/ebpf/src/wg_proxy.c

@@ -50,5 +50,11 @@ int xdp_wg_proxy(struct iphdr  *ip, struct udphdr *udp) {
     __be16 new_dst_port = htons(proxy_port);
     udp->dest = new_dst_port;
     udp->source = new_src_port;
+
+    // The ports are covered by the UDP checksum. This is an IPv4 loopback hop
+    // and the payload is already integrity-protected, so clear the checksum (a
+    // zero UDP checksum means "not computed" for IPv4) rather than leave a
+    // stale value the kernel would drop as UDP_CSUM.
+    udp->check = 0;
 	return XDP_PASS;
 }

client/internal/engine.go

@@ -64,7 +64,10 @@ import (
 	"github.com/netbirdio/netbird/route"
 	mgm "github.com/netbirdio/netbird/shared/management/client"
 	"github.com/netbirdio/netbird/shared/management/domain"
+	sharedgrpc "github.com/netbirdio/netbird/shared/management/grpc"
+	nbnetworkmap "github.com/netbirdio/netbird/shared/management/networkmap"
 	mgmProto "github.com/netbirdio/netbird/shared/management/proto"
+	types "github.com/netbirdio/netbird/shared/management/types"
 	"github.com/netbirdio/netbird/shared/netiputil"
 	auth "github.com/netbirdio/netbird/shared/relay/auth/hmac"
 	relayClient "github.com/netbirdio/netbird/shared/relay/client"
@@ -147,6 +150,7 @@ type EngineConfig struct {
 	BlockLANAccess      bool
 	BlockInbound        bool
 	DisableIPv6         bool
+	SyncMessageVersion  *int
 
 	// LazyConnection is the MDM-sourced lazy-connection override; StateUnset defers to
 	// the env var and management feature flag.
@@ -220,6 +224,13 @@ type Engine struct {
 	// networkSerial is the latest CurrentSerial (state ID) of the network sent by the Management service
 	networkSerial uint64
 
+	// latestComponents is the most-recent NetworkMapComponents decoded from
+	// a NetworkMapEnvelope (capability=3 peers only). Held alongside the
+	// NetworkMap that Calculate() produced from it so future incremental
+	// updates have a base to apply changes against. nil for legacy-format
+	// peers. Guarded by syncMsgMux.
+	latestComponents *types.NetworkMapComponents
+
 	networkMonitor *networkmonitor.NetworkMonitor
 
 	sshServer sshServer
@@ -652,8 +663,24 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	iceCfg := e.createICEConfig()
 
 	e.connMgr = NewConnMgr(e.config, e.statusRecorder, e.peerStore, wgIface)
+	e.connMgr.SetRoutedIPsReconciler(func(peerKey string) error {
+		if e.routeManager == nil {
+			return nil
+		}
+		return e.routeManager.ReconcilePeerAllowedIPs(peerKey)
+	})
 	e.connMgr.Start(e.ctx)
 
+	// Wire DNS-time lazy-connection warm-up now that the connection manager
+	// exists (it does not at DNS-server construction time). A DNS answer that
+	// points at an idle peer then wakes it before the client's first request.
+	e.dnsServer.SetPeerActivator(&dnsPeerActivator{
+		connMgr:   e.connMgr,
+		peerStore: e.peerStore,
+		status:    e.statusRecorder,
+		ctx:       e.ctx,
+	})
+
 	e.srWatcher = guard.NewSRWatcher(e.signal, e.relayManager, e.mobileDep.IFaceDiscover, iceCfg)
 	e.srWatcher.Start(peer.IsForceRelayed())
 
@@ -963,8 +990,12 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 
 	e.ApplySessionDeadline(update.GetSessionExpiresAt())
 
-	if update.NetworkMap != nil && update.NetworkMap.PeerConfig != nil {
-		e.handleAutoUpdateVersion(update.NetworkMap.PeerConfig.AutoUpdate)
+	// Envelope sync responses carry PeerConfig at the top level; legacy
+	// NetworkMap syncs carry it under NetworkMap.PeerConfig.
+	if pc := update.GetPeerConfig(); pc != nil {
+		e.handleAutoUpdateVersion(pc.GetAutoUpdate())
+	} else if nm := update.GetNetworkMap(); nm != nil && nm.GetPeerConfig() != nil {
+		e.handleAutoUpdateVersion(nm.GetPeerConfig().GetAutoUpdate())
 	}
 
 	done := e.phase("netbird_config")
@@ -974,12 +1005,47 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 		return err
 	}
 
+	// Decode the network map from either the components envelope or the
+	// legacy proto.NetworkMap before the posture-check gating below, so the
+	// "is there a network map" decision covers both wire shapes.
+	var (
+		nm         *mgmProto.NetworkMap
+		components *types.NetworkMapComponents
+	)
+	if version := update.GetVersion(); version == int32(sharedgrpc.ComponentNetworkMap) {
+		// Components-format peer: decode the envelope back to typed
+		// components, run Calculate() locally, and convert to the wire
+		// NetworkMap shape the rest of the engine consumes. Components are
+		// retained so future incremental updates can apply deltas instead
+		// of doing a full reconstruction.
+		envelope := update.GetNetworkMapEnvelope()
+		if envelope == nil {
+			return fmt.Errorf("received a SyncReponse indicating use of components network map, but components are missing")
+		}
+
+		localKey := e.config.WgPrivateKey.PublicKey().String()
+		dnsName := ""
+		if pc := update.GetPeerConfig(); pc != nil {
+			// PeerConfig.Fqdn = "<dns_label>.<dns_domain>" — extract the
+			// shared domain by stripping the peer's own label prefix. Falls
+			// back to empty if the FQDN doesn't have the expected shape.
+			dnsName = extractDNSDomainFromFQDN(pc.GetFqdn())
+		}
+		result, err := nbnetworkmap.EnvelopeToNetworkMap(e.ctx, envelope, localKey, dnsName)
+		if err != nil {
+			return fmt.Errorf("decode network map envelope: %w", err)
+		}
+		nm = result.NetworkMap
+		components = result.Components
+	} else {
+		nm = update.GetNetworkMap()
+	}
+
 	// Posture checks are bound to the network map presence:
 	//   NetworkMap != nil, checks present -> apply the received checks
 	//   NetworkMap != nil, checks nil     -> posture checks were removed, clear them
 	//   NetworkMap == nil                 -> config-only update (e.g. relay token rotation),
 	//                                        leave the previously applied checks untouched
-	nm := update.GetNetworkMap()
 	if nm == nil {
 		return nil
 	}
@@ -992,6 +1058,14 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 	}
 
 	done = e.phase("persist")
+	// Only retain the components view when the server sent the envelope
+	// path. A legacy proto.NetworkMap means components == nil; writing it
+	// here would clobber a previously-cached snapshot, breaking the
+	// incremental-delta base on a future envelope sync.
+	if components != nil {
+		e.latestComponents = components
+	}
+
 	e.persistSyncResponse(update)
 	done()
 
@@ -1005,6 +1079,19 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 	return nil
 }
 
+// extractDNSDomainFromFQDN returns the trailing dotted domain part of the
+// receiving peer's FQDN — the same value the management server fills as
+// dnsName when it builds the legacy NetworkMap. "peer42.netbird.cloud" →
+// "netbird.cloud". An empty string is returned for unrecognized formats.
+func extractDNSDomainFromFQDN(fqdn string) string {
+	for i := 0; i < len(fqdn); i++ {
+		if fqdn[i] == '.' && i+1 < len(fqdn) {
+			return fqdn[i+1:]
+		}
+	}
+	return ""
+}
+
 // updateNetbirdConfig applies the management-provided NetBird configuration:
 // STUN/TURN and relay servers, flow logging and DNS settings. A nil config is a no-op,
 // which is the case for sync updates carrying only a network map.
@@ -1164,6 +1251,7 @@ func (e *Engine) applyInfoFlags(info *system.Info) {
 		e.config.BlockLANAccess,
 		e.config.BlockInbound,
 		e.config.DisableIPv6,
+		e.config.SyncMessageVersion,
 		e.config.EnableSSHRoot,
 		e.config.EnableSSHSFTP,
 		e.config.EnableSSHLocalPortForwarding,
@@ -2032,6 +2120,7 @@ func (e *Engine) readInitialSettings() ([]*route.Route, *nbdns.Config, bool, err
 		e.config.BlockLANAccess,
 		e.config.BlockInbound,
 		e.config.DisableIPv6,
+		e.config.SyncMessageVersion,
 		e.config.EnableSSHRoot,
 		e.config.EnableSSHSFTP,
 		e.config.EnableSSHLocalPortForwarding,

client/internal/engine_session_deadline_test.go

@@ -75,4 +75,14 @@ func TestApplySessionDeadline_ThreeState(t *testing.T) {
 		require.True(t, e.statusRecorder.GetSessionExpiresAt().IsZero(),
 			"invalid timestamp must clear the deadline")
 	})
+
+	t.Run("recently expired timestamp stays visible as expired", func(t *testing.T) {
+		e := newEngine()
+		expired := time.Now().Add(-5 * time.Minute).UTC().Truncate(time.Second)
+
+		e.ApplySessionDeadline(timestamppb.New(expired))
+
+		require.True(t, e.statusRecorder.GetSessionExpiresAt().Equal(expired),
+			"recently-expired deadline must stay on the recorder so consumers render it as expired")
+	})
 }

client/internal/ipcauth/creds_stub.go

@@ -1,13 +0,0 @@
-//go:build !linux && !darwin && !freebsd && !windows
-
-package ipcauth
-
-import "google.golang.org/grpc/credentials"
-
-// NewTransportCredentials returns nil on platforms without a peer-identity
-// primitive. The daemon falls back to insecure credentials and skips per-RPC
-// authorization (logging a warning), preserving pre-hardening behavior until
-// the transport gains an identity primitive.
-func NewTransportCredentials() credentials.TransportCredentials {
-	return nil
-}

client/internal/ipcauth/creds_unix.go

@@ -1,48 +0,0 @@
-//go:build linux || darwin || freebsd
-
-package ipcauth
-
-import (
-	"context"
-	"net"
-
-	"google.golang.org/grpc/credentials"
-)
-
-// NewTransportCredentials returns gRPC transport credentials that extract the
-// caller's kernel-authenticated identity from a Unix-socket connection and
-// expose it via IdentityFromContext. It is non-nil on platforms with a
-// peer-credential primitive.
-func NewTransportCredentials() credentials.TransportCredentials {
-	return unixCreds{}
-}
-
-// unixCreds implements credentials.TransportCredentials over a Unix socket.
-// The server side reads SO_PEERCRED/LOCAL_PEERCRED during the handshake; the
-// client side is a no-op (the kernel supplies the peer identity to the server
-// without any client cooperation).
-type unixCreds struct{}
-
-func (unixCreds) ClientHandshake(_ context.Context, _ string, conn net.Conn) (net.Conn, credentials.AuthInfo, error) {
-	return conn, AuthInfo{}, nil
-}
-
-// ServerHandshake extracts the peer identity and fails closed if it cannot be read.
-func (unixCreds) ServerHandshake(conn net.Conn) (net.Conn, credentials.AuthInfo, error) {
-	id, err := PeerIdentity(conn)
-	if err != nil {
-		return nil, nil, err
-	}
-	return conn, AuthInfo{
-		CommonAuthInfo: credentials.CommonAuthInfo{SecurityLevel: credentials.NoSecurity},
-		Identity:       id,
-	}, nil
-}
-
-func (unixCreds) Info() credentials.ProtocolInfo {
-	return credentials.ProtocolInfo{SecurityProtocol: "netbird-ipc-peercred"}
-}
-
-func (unixCreds) Clone() credentials.TransportCredentials { return unixCreds{} }
-
-func (unixCreds) OverrideServerName(string) error { return nil }

client/internal/ipcauth/creds_windows.go

@@ -1,142 +0,0 @@
-//go:build windows
-
-package ipcauth
-
-import (
-	"context"
-	"fmt"
-	"net"
-	"runtime"
-
-	"golang.org/x/sys/windows"
-	"google.golang.org/grpc/credentials"
-)
-
-var (
-	modadvapi32                    = windows.NewLazySystemDLL("advapi32.dll")
-	procImpersonateNamedPipeClient = modadvapi32.NewProc("ImpersonateNamedPipeClient")
-	procRevertToSelf               = modadvapi32.NewProc("RevertToSelf")
-)
-
-// Windows group-SID attribute flags (winnt.h): a group only counts toward
-// membership when it is enabled and not marked use-for-deny-only.
-const (
-	seGroupEnabled        = 0x00000004
-	seGroupUseForDenyOnly = 0x00000010
-)
-
-// DefaultPipeSDDL restricts the daemon control pipe to LocalSystem (SY), the
-// Administrators group (BA), and interactive logon users (IU). It deliberately
-// excludes Authenticated Users / Everyone so remote or arbitrary service
-// principals cannot connect. This is the Layer-1 channel gate; the interceptor
-// (Layer 2) further restricts by per-profile ownership.
-func DefaultPipeSDDL() string {
-	return "D:P(A;;GA;;;SY)(A;;GA;;;BA)(A;;GA;;;IU)"
-}
-
-// NewTransportCredentials returns gRPC transport credentials that derive the
-// caller's identity from the named-pipe client token, following Microsoft's
-// "Verifying Client Access with ACLs" pattern: ImpersonateNamedPipeClient ->
-// OpenThreadToken -> RevertToSelf. Per threat-model M-NOIMP, impersonation is
-// used only to read the client token for identity, never to perform privileged work.
-func NewTransportCredentials() credentials.TransportCredentials {
-	return winpipeCreds{}
-}
-
-type winpipeCreds struct{}
-
-func (winpipeCreds) ClientHandshake(_ context.Context, _ string, conn net.Conn) (net.Conn, credentials.AuthInfo, error) {
-	return conn, AuthInfo{}, nil
-}
-
-// ServerHandshake extracts the connecting client's identity from the pipe token.
-// Fails closed if the handle or token cannot be read.
-func (winpipeCreds) ServerHandshake(conn net.Conn) (net.Conn, credentials.AuthInfo, error) {
-	// go-winio's pipe connection embeds *win32File, which exposes Fd().
-	fdConn, ok := conn.(interface{ Fd() uintptr })
-	if !ok {
-		return nil, nil, fmt.Errorf("connection %T does not expose a pipe handle", conn)
-	}
-	handle := windows.Handle(fdConn.Fd())
-
-	id, err := pipeClientIdentity(handle)
-	if err != nil {
-		return nil, nil, err
-	}
-	return conn, AuthInfo{
-		CommonAuthInfo: credentials.CommonAuthInfo{SecurityLevel: credentials.NoSecurity},
-		Identity:       id,
-	}, nil
-}
-
-func (winpipeCreds) Info() credentials.ProtocolInfo {
-	return credentials.ProtocolInfo{SecurityProtocol: "netbird-ipc-peercred"}
-}
-
-func (winpipeCreds) Clone() credentials.TransportCredentials { return winpipeCreds{} }
-
-func (winpipeCreds) OverrideServerName(string) error { return nil }
-
-// pipeClientIdentity reads the connecting client's user SID and enabled group
-// SIDs from the named-pipe handle. The impersonation window is kept as small as
-// possible and pinned to the OS thread (impersonation is thread-local).
-func pipeClientIdentity(handle windows.Handle) (Identity, error) {
-	var pid uint32
-	hasPID := windows.GetNamedPipeClientProcessId(handle, &pid) == nil
-
-	runtime.LockOSThread()
-	defer runtime.UnlockOSThread()
-
-	if err := impersonateNamedPipeClient(handle); err != nil {
-		return Identity{}, fmt.Errorf("impersonate named pipe client: %w", err)
-	}
-	defer func() { _ = revertToSelf() }()
-
-	// openAsSelf=true: the token is opened using the daemon's process context
-	// (LocalSystem), not the impersonated client's, so the open always succeeds.
-	var token windows.Token
-	if err := windows.OpenThreadToken(windows.CurrentThread(), windows.TOKEN_QUERY, true, &token); err != nil {
-		return Identity{}, fmt.Errorf("open thread token: %w", err)
-	}
-	defer token.Close()
-
-	tu, err := token.GetTokenUser()
-	if err != nil {
-		return Identity{}, fmt.Errorf("get token user: %w", err)
-	}
-
-	tg, err := token.GetTokenGroups()
-	if err != nil {
-		return Identity{}, fmt.Errorf("get token groups: %w", err)
-	}
-	var groups []string
-	for _, g := range tg.AllGroups() {
-		if g.Attributes&seGroupEnabled == 0 || g.Attributes&seGroupUseForDenyOnly != 0 {
-			continue
-		}
-		groups = append(groups, g.Sid.String())
-	}
-
-	return Identity{
-		SID:    tu.User.Sid.String(),
-		Groups: groups,
-		PID:    int32(pid),
-		HasPID: hasPID,
-	}, nil
-}
-
-func impersonateNamedPipeClient(h windows.Handle) error {
-	r, _, e := procImpersonateNamedPipeClient.Call(uintptr(h))
-	if r == 0 {
-		return e
-	}
-	return nil
-}
-
-func revertToSelf() error {
-	r, _, e := procRevertToSelf.Call()
-	if r == 0 {
-		return e
-	}
-	return nil
-}

client/internal/ipcauth/identity.go

@@ -1,93 +0,0 @@
-// Package ipcauth provides kernel-authenticated caller identity for the daemon's
-// local IPC (gRPC) channel and the transport credentials that populate it.
-//
-// It is the identity foundation shared by two layers of the local-IPC hardening:
-//   - the socket-permission layer (Layer 1, client/cmd), which reads the peer
-//     identity to gate who may connect and to run trust-on-first-use; and
-//   - the per-RPC authorization interceptor (Layer 2), which reads the same
-//     identity from the gRPC context to enforce per-profile ownership.
-//
-// On Unix the identity is read from the kernel via SO_PEERCRED (Linux) or
-// LOCAL_PEERCRED (Darwin/FreeBSD). On Windows it is derived from the named-pipe
-// client token. Platforms without a peer-identity primitive get no credentials
-// and therefore no enforcement (the daemon logs a warning and stays open,
-// preserving today's behavior until the transport is hardened).
-package ipcauth
-
-import (
-	"context"
-	"fmt"
-
-	"google.golang.org/grpc/credentials"
-	"google.golang.org/grpc/peer"
-)
-
-// Identity is the kernel-authenticated identity of a local IPC caller.
-//
-// The zero value is not a valid identity; callers obtain one via
-// IdentityFromContext (which reports presence) or PeerIdentity.
-type Identity struct {
-	// UID and GID are the caller's Unix user ID and primary group ID.
-	// Zero on Windows, where SID is authoritative instead.
-	UID uint32
-	GID uint32
-
-	// PID is the caller's process ID, for audit only. HasPID is false when the
-	// platform cannot supply it (e.g. Darwin/FreeBSD xucred carries no PID).
-	PID    int32
-	HasPID bool
-
-	// SID is the caller's Windows security identifier (empty on Unix).
-	SID string
-
-	// Groups holds the caller's Windows group SIDs, captured from the client
-	// token at handshake time (empty on Unix, where supplementary group
-	// membership is resolved on demand via NSS/getent by the authorizer).
-	Groups []string
-}
-
-// IsWindows reports whether this identity is a Windows principal (SID-based)
-// rather than a Unix uid/gid principal.
-func (i Identity) IsWindows() bool {
-	return i.SID != ""
-}
-
-// String renders the identity for audit logs.
-func (i Identity) String() string {
-	if i.IsWindows() {
-		if i.HasPID {
-			return fmt.Sprintf("sid=%s pid=%d", i.SID, i.PID)
-		}
-		return fmt.Sprintf("sid=%s", i.SID)
-	}
-	if i.HasPID {
-		return fmt.Sprintf("uid=%d gid=%d pid=%d", i.UID, i.GID, i.PID)
-	}
-	return fmt.Sprintf("uid=%d gid=%d", i.UID, i.GID)
-}
-
-// AuthInfo carries the peer Identity as a gRPC credentials.AuthInfo so the
-// interceptor can retrieve it from the request context via IdentityFromContext.
-type AuthInfo struct {
-	credentials.CommonAuthInfo
-	Identity Identity
-}
-
-// AuthType identifies the authentication scheme.
-func (AuthInfo) AuthType() string { return "netbird-ipc-peercred" }
-
-// IdentityFromContext extracts the caller's kernel-authenticated identity from
-// the gRPC peer context. The second return value is false when no IPC transport
-// credentials were negotiated (e.g. an unsupported platform, or a caller that
-// did not come through the daemon socket) — callers MUST fail closed in that case.
-func IdentityFromContext(ctx context.Context) (Identity, bool) {
-	p, ok := peer.FromContext(ctx)
-	if !ok {
-		return Identity{}, false
-	}
-	info, ok := p.AuthInfo.(AuthInfo)
-	if !ok {
-		return Identity{}, false
-	}
-	return info.Identity, true
-}

client/internal/ipcauth/identity_test.go

@@ -1,47 +0,0 @@
-package ipcauth
-
-import (
-	"context"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-	"google.golang.org/grpc/credentials"
-	"google.golang.org/grpc/peer"
-)
-
-func TestIdentityFromContext_NoPeer(t *testing.T) {
-	_, ok := IdentityFromContext(context.Background())
-	assert.False(t, ok, "bare context must report no identity (fail closed)")
-}
-
-func TestIdentityFromContext_WrongAuthInfo(t *testing.T) {
-	ctx := peer.NewContext(context.Background(), &peer.Peer{})
-	_, ok := IdentityFromContext(ctx)
-	assert.False(t, ok, "peer without our AuthInfo must report no identity")
-}
-
-func TestIdentityFromContext_Present(t *testing.T) {
-	want := Identity{UID: 1000, GID: 1000, PID: 4242, HasPID: true}
-	ctx := peer.NewContext(context.Background(), &peer.Peer{
-		AuthInfo: AuthInfo{
-			CommonAuthInfo: credentials.CommonAuthInfo{SecurityLevel: credentials.NoSecurity},
-			Identity:       want,
-		},
-	})
-
-	got, ok := IdentityFromContext(ctx)
-	assert.True(t, ok)
-	assert.Equal(t, want, got)
-}
-
-func TestIdentity_String(t *testing.T) {
-	assert.Equal(t, "uid=1000 gid=1000 pid=42", Identity{UID: 1000, GID: 1000, PID: 42, HasPID: true}.String())
-	assert.Equal(t, "uid=1000 gid=1000", Identity{UID: 1000, GID: 1000}.String())
-	assert.Equal(t, "sid=S-1-5-21-1 pid=42", Identity{SID: "S-1-5-21-1", PID: 42, HasPID: true}.String())
-	assert.Equal(t, "sid=S-1-5-21-1", Identity{SID: "S-1-5-21-1"}.String())
-}
-
-func TestIdentity_IsWindows(t *testing.T) {
-	assert.True(t, Identity{SID: "S-1-5-18"}.IsWindows())
-	assert.False(t, Identity{UID: 0}.IsWindows())
-}

client/internal/ipcauth/peercred_bsd.go

@@ -1,43 +0,0 @@
-//go:build darwin || freebsd
-
-package ipcauth
-
-import (
-	"fmt"
-	"net"
-
-	"golang.org/x/sys/unix"
-)
-
-// PeerIdentity reads the kernel-authenticated identity of the process on the
-// other end of a Unix socket connection via LOCAL_PEERCRED (xucred). xucred
-// carries the uid and group list but no pid, so audit on these platforms is
-// uid/gid-based (HasPID is false); PID via LOCAL_PEERPID is a possible follow-up.
-func PeerIdentity(c net.Conn) (Identity, error) {
-	uc, ok := c.(*net.UnixConn)
-	if !ok {
-		return Identity{}, fmt.Errorf("connection is not a unix socket: %T", c)
-	}
-	raw, err := uc.SyscallConn()
-	if err != nil {
-		return Identity{}, fmt.Errorf("raw conn: %w", err)
-	}
-
-	var cred *unix.Xucred
-	var credErr error
-	if err := raw.Control(func(fd uintptr) {
-		cred, credErr = unix.GetsockoptXucred(int(fd), unix.SOL_LOCAL, unix.LOCAL_PEERCRED)
-	}); err != nil {
-		return Identity{}, fmt.Errorf("getsockopt control: %w", err)
-	}
-	if credErr != nil {
-		return Identity{}, fmt.Errorf("LOCAL_PEERCRED: %w", credErr)
-	}
-
-	id := Identity{UID: cred.Uid}
-	// Groups[0] is the effective (primary) GID; guard against an empty list.
-	if cred.Ngroups > 0 {
-		id.GID = cred.Groups[0]
-	}
-	return id, nil
-}

client/internal/ipcauth/peercred_linux.go

@@ -1,43 +0,0 @@
-//go:build linux
-
-package ipcauth
-
-import (
-	"fmt"
-	"net"
-
-	"golang.org/x/sys/unix"
-)
-
-// PeerIdentity reads the kernel-authenticated identity of the process on the
-// other end of a Unix socket connection via SO_PEERCRED. The credentials are
-// captured by the kernel at connect() time and cannot be spoofed or changed for
-// the life of the connection.
-func PeerIdentity(c net.Conn) (Identity, error) {
-	uc, ok := c.(*net.UnixConn)
-	if !ok {
-		return Identity{}, fmt.Errorf("connection is not a unix socket: %T", c)
-	}
-	raw, err := uc.SyscallConn()
-	if err != nil {
-		return Identity{}, fmt.Errorf("raw conn: %w", err)
-	}
-
-	var cred *unix.Ucred
-	var credErr error
-	if err := raw.Control(func(fd uintptr) {
-		cred, credErr = unix.GetsockoptUcred(int(fd), unix.SOL_SOCKET, unix.SO_PEERCRED)
-	}); err != nil {
-		return Identity{}, fmt.Errorf("getsockopt control: %w", err)
-	}
-	if credErr != nil {
-		return Identity{}, fmt.Errorf("SO_PEERCRED: %w", credErr)
-	}
-
-	return Identity{
-		UID:    cred.Uid,
-		GID:    cred.Gid,
-		PID:    cred.Pid,
-		HasPID: true,
-	}, nil
-}

client/internal/ipcauth/peercred_stub.go

@@ -1,16 +0,0 @@
-//go:build !linux && !darwin && !freebsd
-
-package ipcauth
-
-import (
-	"fmt"
-	"net"
-	"runtime"
-)
-
-// PeerIdentity is unimplemented on platforms without a Unix-socket peer-credential
-// primitive. Windows derives identity from the named-pipe client token instead
-// (see the Windows transport credentials), so it never calls this.
-func PeerIdentity(net.Conn) (Identity, error) {
-	return Identity{}, fmt.Errorf("peer credential check not supported on %s", runtime.GOOS)
-}

client/internal/ipcauth/peercred_unix_test.go

@@ -1,113 +0,0 @@
-//go:build linux || darwin || freebsd
-
-package ipcauth
-
-import (
-	"net"
-	"os"
-	"path/filepath"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-// TestPeerIdentity_MatchesCurrentProcess connects to a real Unix socket and
-// verifies the extracted UID/GID match the running process (both ends are us).
-func TestPeerIdentity_MatchesCurrentProcess(t *testing.T) {
-	sock := filepath.Join(t.TempDir(), "peer.sock")
-	ln, err := net.Listen("unix", sock)
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = ln.Close() })
-
-	type result struct {
-		id  Identity
-		err error
-	}
-	done := make(chan result, 1)
-	go func() {
-		c, aerr := ln.Accept()
-		if aerr != nil {
-			done <- result{err: aerr}
-			return
-		}
-		defer func() { _ = c.Close() }()
-		id, ierr := PeerIdentity(c)
-		done <- result{id: id, err: ierr}
-	}()
-
-	client, err := net.Dial("unix", sock)
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = client.Close() })
-
-	res := <-done
-	require.NoError(t, res.err)
-	assert.Equal(t, uint32(os.Getuid()), res.id.UID, "UID should match current process")
-	assert.Equal(t, uint32(os.Getgid()), res.id.GID, "primary GID should match current process")
-}
-
-// TestPeerIdentity_NonUnixConn rejects non-Unix connections (fail closed).
-func TestPeerIdentity_NonUnixConn(t *testing.T) {
-	ln, err := net.Listen("tcp", "127.0.0.1:0")
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = ln.Close() })
-
-	done := make(chan error, 1)
-	go func() {
-		c, aerr := ln.Accept()
-		if aerr != nil {
-			done <- aerr
-			return
-		}
-		defer func() { _ = c.Close() }()
-		_, ierr := PeerIdentity(c)
-		done <- ierr
-	}()
-
-	client, err := net.Dial("tcp", ln.Addr().String())
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = client.Close() })
-
-	assert.Error(t, <-done, "PeerIdentity must reject a non-Unix connection")
-}
-
-// TestUnixCreds_ServerHandshake exercises the transport-credentials path end to end.
-func TestUnixCreds_ServerHandshake(t *testing.T) {
-	creds := NewTransportCredentials()
-	require.NotNil(t, creds)
-
-	sock := filepath.Join(t.TempDir(), "hs.sock")
-	ln, err := net.Listen("unix", sock)
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = ln.Close() })
-
-	type result struct {
-		info interface{ AuthType() string }
-		err  error
-	}
-	done := make(chan result, 1)
-	go func() {
-		c, aerr := ln.Accept()
-		if aerr != nil {
-			done <- result{err: aerr}
-			return
-		}
-		_, ai, herr := creds.ServerHandshake(c)
-		if herr != nil {
-			done <- result{err: herr}
-			return
-		}
-		done <- result{info: ai}
-	}()
-
-	client, err := net.Dial("unix", sock)
-	require.NoError(t, err)
-	t.Cleanup(func() { _ = client.Close() })
-
-	res := <-done
-	require.NoError(t, res.err)
-	ai, ok := res.info.(AuthInfo)
-	require.True(t, ok, "expected ipcauth.AuthInfo, got %T", res.info)
-	assert.Equal(t, uint32(os.Getuid()), ai.Identity.UID)
-	assert.Equal(t, "netbird-ipc-peercred", ai.AuthType())
-}

client/internal/lazyconn/manager/manager.go

@@ -29,6 +29,11 @@ type managedPeer struct {
 
 type Config struct {
 	InactivityThreshold *time.Duration
+	// ReconcileAllowedIPs re-applies a peer's routed allowed IPs after its wake endpoint is
+	// armed. The activity listener creates the wake peer with the overlay /32 only; without the
+	// routed prefixes WireGuard would not steer subnet-bound traffic to the wake endpoint, so an
+	// idle routing peer could never be woken by that traffic. Optional; nil disables the reconcile.
+	ReconcileAllowedIPs func(peerKey string) error
 }
 
 // Manager manages lazy connections
@@ -56,6 +61,9 @@ type Manager struct {
 	peerToHAGroups map[string][]route.HAUniqueID // peer ID -> HA groups they belong to
 	haGroupToPeers map[route.HAUniqueID][]string // HA group -> peer IDs in the group
 	routesMu       sync.RWMutex
+
+	// reconcileAllowedIPs re-applies a peer's routed allowed IPs after its wake endpoint is armed.
+	reconcileAllowedIPs func(peerKey string) error
 }
 
 // NewManager creates a new lazy connection manager
@@ -73,6 +81,7 @@ func NewManager(config Config, engineCtx context.Context, peerStore *peerstore.S
 		activityManager:      activity.NewManager(wgIface),
 		peerToHAGroups:       make(map[string][]route.HAUniqueID),
 		haGroupToPeers:       make(map[route.HAUniqueID][]string),
+		reconcileAllowedIPs:  config.ReconcileAllowedIPs,
 	}
 
 	if wgIface.IsUserspaceBind() {
@@ -201,7 +210,7 @@ func (m *Manager) AddPeer(peerCfg lazyconn.PeerConfig) (bool, error) {
 		return false, nil
 	}
 
-	if err := m.activityManager.MonitorPeerActivity(peerCfg); err != nil {
+	if err := m.armActivityListener(peerCfg); err != nil {
 		return false, err
 	}
 
@@ -288,7 +297,7 @@ func (m *Manager) DeactivatePeer(peerID peerid.ConnID) {
 
 	m.inactivityManager.RemovePeer(mp.peerCfg.PublicKey)
 
-	if err := m.activityManager.MonitorPeerActivity(*mp.peerCfg); err != nil {
+	if err := m.armActivityListener(*mp.peerCfg); err != nil {
 		mp.peerCfg.Log.Errorf("failed to create activity monitor: %v", err)
 		return
 	}
@@ -465,6 +474,31 @@ func (m *Manager) close() {
 }
 
 // shouldDeferIdleForHA checks if peer should stay connected due to HA group requirements
+// armRoutedAllowedIPs re-applies the peer's routed allowed IPs onto its freshly armed wake
+// endpoint. The activity listener creates the wake peer with the overlay /32 only, so without
+// this the routed prefixes would be missing and traffic to a routed subnet could not wake the
+// idle routing peer. It is a no-op when no reconciler is configured.
+// armActivityListener (re)arms the peer's wake endpoint via the activity manager and then
+// re-applies its routed allowed IPs, so traffic to a routed subnet can wake an idle routing
+// peer. The routed prefixes must be re-applied after the wake endpoint exists because the
+// listener creates it with the overlay /32 only.
+func (m *Manager) armActivityListener(peerCfg lazyconn.PeerConfig) error {
+	if err := m.activityManager.MonitorPeerActivity(peerCfg); err != nil {
+		return err
+	}
+	m.armRoutedAllowedIPs(&peerCfg)
+	return nil
+}
+
+func (m *Manager) armRoutedAllowedIPs(peerCfg *lazyconn.PeerConfig) {
+	if m.reconcileAllowedIPs == nil {
+		return
+	}
+	if err := m.reconcileAllowedIPs(peerCfg.PublicKey); err != nil {
+		peerCfg.Log.Errorf("failed to reconcile routed allowed IPs on wake endpoint: %v", err)
+	}
+}
+
 func (m *Manager) shouldDeferIdleForHA(inactivePeers map[string]struct{}, peerID string) bool {
 	m.routesMu.RLock()
 	defer m.routesMu.RUnlock()
@@ -577,7 +611,7 @@ func (m *Manager) onPeerInactivityTimedOut(peerIDs map[string]struct{}) {
 
 		mp.peerCfg.Log.Infof("start activity monitor")
 
-		if err := m.activityManager.MonitorPeerActivity(*mp.peerCfg); err != nil {
+		if err := m.armActivityListener(*mp.peerCfg); err != nil {
 			mp.peerCfg.Log.Errorf("failed to create activity monitor: %v", err)
 			continue
 		}

client/internal/peer/status.go

@@ -813,19 +813,14 @@ func (d *Status) SetSessionExpiresAt(deadline time.Time) {
 }
 
 // GetSessionExpiresAt returns the most recently recorded SSO session deadline,
-// or the zero value when no deadline is tracked. A deadline that has already
-// slipped into the past reports as "none": once the session has expired it is
-// no longer a meaningful countdown, and the sessionwatch.Watcher does not
-// arm a timer at the deadline itself to clear it (only the two pre-expiry
-// warnings). Without this guard the UI would keep painting a stale
-// "expires in …" against a moment that has passed until the next login,
-// extend, or teardown rewrote the value.
+// or the zero value when no deadline is tracked. A deadline in the past is
+// returned as-is: it means the session has expired, and consumers (tray row,
+// CLI status) render it as "expired" rather than hiding it — masking it as
+// "none" would blank the UI at the exact moment it should say the session
+// ended.
 func (d *Status) GetSessionExpiresAt() time.Time {
 	d.mux.Lock()
 	defer d.mux.Unlock()
-	if !d.sessionExpiresAt.IsZero() && d.sessionExpiresAt.Before(time.Now()) {
-		return time.Time{}
-	}
 	return d.sessionExpiresAt
 }
 

client/internal/profilemanager/config.go

@@ -96,6 +96,7 @@ type ConfigInput struct {
 	BlockLANAccess      *bool
 	BlockInbound        *bool
 	DisableIPv6         *bool
+	SyncMessageVersion  *int
 
 	DisableNotifications *bool
 
@@ -137,6 +138,7 @@ type Config struct {
 	BlockLANAccess      bool
 	BlockInbound        bool
 	DisableIPv6         bool
+	SyncMessageVersion  *int
 
 	DisableNotifications *bool
 
@@ -587,6 +589,12 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
+	if input.SyncMessageVersion != nil && *input.SyncMessageVersion != *config.SyncMessageVersion {
+		log.Infof("setting SyncMessageVersion to %v", *input.SyncMessageVersion)
+		*config.SyncMessageVersion = *input.SyncMessageVersion
+		updated = true
+	}
+
 	if input.DisableNotifications != nil && (config.DisableNotifications == nil || *input.DisableNotifications != *config.DisableNotifications) {
 		if *input.DisableNotifications {
 			log.Infof("disabling notifications")

client/internal/routemanager/dynamic/route.go

@@ -185,7 +185,7 @@ func (r *Route) startResolver(ctx context.Context) {
 }
 
 func (r *Route) update(ctx context.Context) error {
-	resolved, err := r.resolveDomains()
+	resolved, err := r.resolveDomains(ctx)
 	if err != nil {
 		if len(resolved) == 0 {
 			return fmt.Errorf("resolve domains: %w", err)
@@ -199,9 +199,9 @@ func (r *Route) update(ctx context.Context) error {
 	return nil
 }
 
-func (r *Route) resolveDomains() (domainMap, error) {
+func (r *Route) resolveDomains(ctx context.Context) (domainMap, error) {
 	results := make(chan resolveResult)
-	go r.resolve(results)
+	go r.resolve(ctx, results)
 
 	resolved := domainMap{}
 	var merr *multierror.Error
@@ -217,7 +217,7 @@ func (r *Route) resolveDomains() (domainMap, error) {
 	return resolved, nberrors.FormatErrorOrNil(merr)
 }
 
-func (r *Route) resolve(results chan resolveResult) {
+func (r *Route) resolve(ctx context.Context, results chan resolveResult) {
 	var wg sync.WaitGroup
 
 	for _, d := range r.route.Domains {
@@ -225,10 +225,10 @@ func (r *Route) resolve(results chan resolveResult) {
 		go func(domain domain.Domain) {
 			defer wg.Done()
 
-			ips, err := r.getIPsFromResolver(domain)
+			ips, err := r.getIPsFromResolver(ctx, domain)
 			if err != nil {
 				log.Tracef("Failed to resolve domain %s with private resolver: %v", domain.SafeString(), err)
-				ips, err = net.LookupIP(domain.PunycodeString())
+				ips, err = lookupHostIPs(ctx, domain)
 				if err != nil {
 					results <- resolveResult{domain: domain, err: fmt.Errorf("resolve d %s: %w", domain.SafeString(), err)}
 					return
@@ -364,6 +364,20 @@ func determinePrefixChanges(oldPrefixes, newPrefixes []netip.Prefix) (toAdd, toR
 	return
 }
 
+// lookupHostIPs resolves d via the system resolver, honoring ctx cancellation.
+func lookupHostIPs(ctx context.Context, d domain.Domain) ([]net.IP, error) {
+	addrs, err := net.DefaultResolver.LookupIPAddr(ctx, d.PunycodeString())
+	if err != nil {
+		return nil, err
+	}
+
+	ips := make([]net.IP, 0, len(addrs))
+	for _, addr := range addrs {
+		ips = append(ips, addr.IP)
+	}
+	return ips, nil
+}
+
 func combinePrefixes(oldPrefixes, removedPrefixes, addedPrefixes []netip.Prefix) []netip.Prefix {
 	prefixSet := make(map[netip.Prefix]struct{})
 	for _, prefix := range oldPrefixes {

client/internal/routemanager/dynamic/route_generic.go

@@ -3,11 +3,12 @@
 package dynamic
 
 import (
+	"context"
 	"net"
 
 	"github.com/netbirdio/netbird/shared/management/domain"
 )
 
-func (r *Route) getIPsFromResolver(domain domain.Domain) ([]net.IP, error) {
-	return net.LookupIP(domain.PunycodeString())
+func (r *Route) getIPsFromResolver(ctx context.Context, domain domain.Domain) ([]net.IP, error) {
+	return lookupHostIPs(ctx, domain)
 }

client/internal/routemanager/dynamic/route_ios.go

@@ -3,6 +3,7 @@
 package dynamic
 
 import (
+	"context"
 	"fmt"
 	"net"
 	"time"
@@ -16,7 +17,7 @@ import (
 
 const dialTimeout = 10 * time.Second
 
-func (r *Route) getIPsFromResolver(domain domain.Domain) ([]net.IP, error) {
+func (r *Route) getIPsFromResolver(ctx context.Context, domain domain.Domain) ([]net.IP, error) {
 	privateClient, err := nbdns.GetClientPrivate(r.wgInterface, r.resolverAddr.Addr(), dialTimeout)
 	if err != nil {
 		return nil, fmt.Errorf("error while creating private client: %s", err)
@@ -32,7 +33,7 @@ func (r *Route) getIPsFromResolver(domain domain.Domain) ([]net.IP, error) {
 		msg := new(dns.Msg)
 		msg.SetQuestion(fqdn, qtype)
 
-		response, _, err := nbdns.ExchangeWithFallback(nil, privateClient, msg, r.resolverAddr.String())
+		response, _, err := nbdns.ExchangeWithFallback(ctx, privateClient, msg, r.resolverAddr.String())
 		if err != nil {
 			if queryErr == nil {
 				queryErr = fmt.Errorf("DNS query for %s (type %d) after %s: %w", domain.SafeString(), qtype, time.Since(startTime), err)

client/internal/routemanager/manager.go

@@ -61,6 +61,7 @@ type Manager interface {
 	InitialRouteRange() []string
 	SetFirewall(firewall.Manager) error
 	SetDNSForwarderPort(port uint16)
+	ReconcilePeerAllowedIPs(peerKey string) error
 	Stop(stateManager *statemanager.Manager)
 }
 
@@ -232,6 +233,30 @@ func (m *DefaultManager) setupRefCounters(useNoop bool) {
 	)
 }
 
+// ReconcilePeerAllowedIPs re-applies every routed allowed IP currently tracked for the peer
+// onto the WireGuard device. The allowed-IP refcounter only calls its AddFunc (which pushes to
+// the device) on a prefix's 0->1 transition, so a peer whose device entry was rebuilt without a
+// matching refcounter change — e.g. a lazy connection cycling through idle->wake, which recreates
+// the WireGuard peer with the overlay /32 only — ends up missing routed prefixes the refcounter
+// still considers installed, and nothing retries. Calling this when the peer's WireGuard entry is
+// (re)created restores convergence. It is add-only and idempotent: AddAllowedIP is update-only, so
+// prefixes are re-added to an existing peer and an absent peer is left untouched.
+func (m *DefaultManager) ReconcilePeerAllowedIPs(peerKey string) error {
+	if m.allowedIPsRefCounter == nil {
+		return nil
+	}
+
+	return m.allowedIPsRefCounter.ReapplyMatching(
+		func(out string) bool { return out == peerKey },
+		func(prefix netip.Prefix) error {
+			if err := m.wgInterface.AddAllowedIP(peerKey, prefix); err != nil {
+				return fmt.Errorf("add allowed IP %s for peer %s: %w", prefix, peerKey, err)
+			}
+			return nil
+		},
+	)
+}
+
 // Init sets up the routing
 func (m *DefaultManager) Init() error {
 	m.routeSelector = m.initSelector()

client/internal/routemanager/mock.go

@@ -112,6 +112,11 @@ func (m *MockManager) SetFirewall(firewall.Manager) error {
 func (m *MockManager) SetDNSForwarderPort(port uint16) {
 }
 
+// ReconcilePeerAllowedIPs mock implementation of ReconcilePeerAllowedIPs from Manager interface
+func (m *MockManager) ReconcilePeerAllowedIPs(peerKey string) error {
+	return nil
+}
+
 // Stop mock implementation of Stop from Manager interface
 func (m *MockManager) Stop(stateManager *statemanager.Manager) {
 	if m.StopFunc != nil {

client/internal/routemanager/reconcile_test.go

@@ -0,0 +1,90 @@
+//go:build !windows
+
+package routemanager
+
+import (
+	"net"
+	"net/netip"
+	"sync"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+	"golang.zx2c4.com/wireguard/tun/netstack"
+
+	"github.com/netbirdio/netbird/client/iface/device"
+	"github.com/netbirdio/netbird/client/iface/wgaddr"
+	"github.com/netbirdio/netbird/client/internal/routemanager/refcounter"
+)
+
+// reconcileWGMock is a minimal iface.WGIface that only records AddAllowedIP calls; every other
+// method is an inert stub because ReconcilePeerAllowedIPs exercises none of them.
+type reconcileWGMock struct {
+	mu   sync.Mutex
+	adds map[string][]netip.Prefix
+}
+
+func (m *reconcileWGMock) AddAllowedIP(peerKey string, allowedIP netip.Prefix) error {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	if m.adds == nil {
+		m.adds = map[string][]netip.Prefix{}
+	}
+	m.adds[peerKey] = append(m.adds[peerKey], allowedIP)
+	return nil
+}
+
+func (m *reconcileWGMock) added(peerKey string) []netip.Prefix {
+	m.mu.Lock()
+	defer m.mu.Unlock()
+	return m.adds[peerKey]
+}
+
+func (m *reconcileWGMock) RemoveAllowedIP(string, netip.Prefix) error { return nil }
+func (m *reconcileWGMock) Name() string                               { return "utun-test" }
+func (m *reconcileWGMock) Address() wgaddr.Address                    { return wgaddr.Address{} }
+func (m *reconcileWGMock) ToInterface() *net.Interface                { return nil }
+func (m *reconcileWGMock) IsUserspaceBind() bool                      { return false }
+func (m *reconcileWGMock) GetFilter() device.PacketFilter             { return nil }
+func (m *reconcileWGMock) GetDevice() *device.FilteredDevice          { return nil }
+func (m *reconcileWGMock) GetNet() *netstack.Net                      { return nil }
+
+// TestReconcilePeerAllowedIPs verifies the declarative reconcile re-applies every routed prefix
+// tracked for the peer (self-heal, independent of refcount level) and stays scoped to that peer.
+func TestReconcilePeerAllowedIPs(t *testing.T) {
+	wg := &reconcileWGMock{}
+	m := &DefaultManager{wgInterface: wg}
+	m.allowedIPsRefCounter = refcounter.New[netip.Prefix, string, string](
+		func(_ netip.Prefix, peerKey string) (string, error) { return peerKey, nil },
+		func(netip.Prefix, string) error { return nil },
+	)
+
+	peerA1 := netip.MustParsePrefix("10.0.0.0/24")
+	peerA2 := netip.MustParsePrefix("10.1.0.0/24")
+	peerB1 := netip.MustParsePrefix("10.2.0.0/24")
+
+	for prefix, peer := range map[netip.Prefix]string{peerA1: "peerA", peerA2: "peerA", peerB1: "peerB"} {
+		_, err := m.allowedIPsRefCounter.Increment(prefix, peer)
+		require.NoError(t, err)
+	}
+	// Extra reference: reconcile must still re-apply the prefix even though its refcount never
+	// hit 0 again (the exact case the plain incremental path skips).
+	_, err := m.allowedIPsRefCounter.Increment(peerA1, "peerA")
+	require.NoError(t, err)
+
+	require.NoError(t, m.ReconcilePeerAllowedIPs("peerA"))
+
+	assert.ElementsMatch(t, []netip.Prefix{peerA1, peerA2}, wg.added("peerA"),
+		"reconcile must re-apply all routed prefixes of the peer")
+	assert.Empty(t, wg.added("peerB"), "reconcile must not touch another peer's prefixes")
+}
+
+// TestReconcilePeerAllowedIPsNoCounter verifies reconcile is a safe no-op before the refcounter is
+// set up.
+func TestReconcilePeerAllowedIPsNoCounter(t *testing.T) {
+	wg := &reconcileWGMock{}
+	m := &DefaultManager{wgInterface: wg}
+
+	require.NoError(t, m.ReconcilePeerAllowedIPs("peerA"))
+	assert.Empty(t, wg.added("peerA"))
+}

client/internal/routemanager/refcounter/refcounter.go

@@ -94,6 +94,26 @@ func (rm *Counter[Key, I, O]) Get(key Key) (Ref[O], bool) {
 	return ref, ok
 }
 
+// ReapplyMatching calls apply for every key whose stored Out satisfies pred, holding the
+// counter lock for the whole pass. Running apply under the lock keeps it atomic with respect
+// to Increment/Decrement: a prefix dropped to zero is removed from the map (and had its
+// RemoveFunc called) before this pass observes it, so a stale key can never be re-applied.
+// pred and apply are invoked under the lock, so they must not call back into the counter.
+func (rm *Counter[Key, I, O]) ReapplyMatching(pred func(out O) bool, apply func(key Key) error) error {
+	rm.mu.Lock()
+	defer rm.mu.Unlock()
+
+	var merr *multierror.Error
+	for key, ref := range rm.refCountMap {
+		if pred(ref.Out) {
+			if err := apply(key); err != nil {
+				merr = multierror.Append(merr, err)
+			}
+		}
+	}
+	return nberrors.FormatErrorOrNil(merr)
+}
+
 // Increment increments the reference count for the given key.
 // If this is the first reference to the key, the AddFunc is called.
 func (rm *Counter[Key, I, O]) Increment(key Key, in I) (Ref[O], error) {

client/internal/routemanager/refcounter/refcounter_test.go

@@ -0,0 +1,47 @@
+package refcounter
+
+import (
+	"net/netip"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+// TestReapplyMatching verifies ReapplyMatching invokes apply for exactly the keys whose stored
+// Out satisfies the predicate (no duplicates for multiply-referenced keys) — the primitive
+// ReconcilePeerAllowedIPs relies on to re-apply a single peer's routed prefixes.
+func TestReapplyMatching(t *testing.T) {
+	rc := New[netip.Prefix, string, string](
+		func(_ netip.Prefix, peerKey string) (string, error) { return peerKey, nil },
+		func(netip.Prefix, string) error { return nil },
+	)
+
+	peerA1 := netip.MustParsePrefix("10.0.0.0/24")
+	peerA2 := netip.MustParsePrefix("10.1.0.0/24")
+	peerB1 := netip.MustParsePrefix("10.2.0.0/24")
+
+	for prefix, peer := range map[netip.Prefix]string{peerA1: "peerA", peerA2: "peerA", peerB1: "peerB"} {
+		_, err := rc.Increment(prefix, peer)
+		require.NoError(t, err)
+	}
+	// a second reference must not make the key applied twice
+	_, err := rc.Increment(peerA1, "peerA")
+	require.NoError(t, err)
+
+	var applied []netip.Prefix
+	err = rc.ReapplyMatching(
+		func(out string) bool { return out == "peerA" },
+		func(key netip.Prefix) error { applied = append(applied, key); return nil },
+	)
+	require.NoError(t, err)
+	assert.ElementsMatch(t, []netip.Prefix{peerA1, peerA2}, applied)
+
+	var none []netip.Prefix
+	err = rc.ReapplyMatching(
+		func(out string) bool { return out == "missing" },
+		func(key netip.Prefix) error { none = append(none, key); return nil },
+	)
+	require.NoError(t, err)
+	assert.Empty(t, none)
+}

client/internal/shell/getent_cgo_unix.go

@@ -1,29 +0,0 @@
-//go:build cgo && !osusergo && !windows
-
-package shell
-
-import "os/user"
-
-// LookupWithGetent with CGO delegates directly to os/user.Lookup.
-// When CGO is enabled, os/user uses libc (getpwnam_r) which goes through
-// the NSS stack natively. If it fails, the user truly doesn't exist and
-// getent would also fail.
-func LookupWithGetent(username string) (*user.User, error) {
-	return user.Lookup(username)
-}
-
-// CurrentUserWithGetent with CGO delegates directly to os/user.Current.
-func CurrentUserWithGetent() (*user.User, error) {
-	return user.Current()
-}
-
-// LookupGroupWithGetent returns the resolved group from either a gid or groupname.
-func LookupGroupWithGetent(name string) (*user.Group, error) {
-	return user.LookupGroup(name)
-}
-
-// GroupIdsWithFallback with CGO delegates directly to user.GroupIds.
-// libc's getgrouplist handles NSS groups natively.
-func GroupIdsWithFallback(u *user.User) ([]string, error) {
-	return u.GroupIds()
-}

client/internal/shell/getent_windows.go

@@ -1,31 +0,0 @@
-//go:build windows
-
-package shell
-
-import "os/user"
-
-// LookupWithGetent on Windows just delegates to os/user.Lookup.
-// Windows does not use NSS/getent; its user lookup works without CGO.
-func LookupWithGetent(username string) (*user.User, error) {
-	return user.Lookup(username)
-}
-
-// CurrentUserWithGetent on Windows just delegates to os/user.Current.
-func CurrentUserWithGetent() (*user.User, error) {
-	return user.Current()
-}
-
-// LookupGroupWithGetent on Windows just delegates to os/user.LookupGroup.
-func LookupGroupWithGetent(name string) (*user.Group, error) {
-	return user.LookupGroup(name)
-}
-
-// GetShellFromGetent is a no-op on Windows; shell resolution uses PowerShell detection.
-func GetShellFromGetent(_ string) string {
-	return ""
-}
-
-// GroupIdsWithFallback on Windows just delegates to u.GroupIds().
-func GroupIdsWithFallback(u *user.User) ([]string, error) {
-	return u.GroupIds()
-}

client/ios/NetBirdSDK/client.go

@@ -233,6 +233,9 @@ func (c *Client) DebugBundle(anonymize bool) (string, error) {
 		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
 			}

client/server/server.go

@@ -1081,7 +1081,10 @@ func (s *Server) Down(ctx context.Context, _ *proto.DownRequest) (*proto.DownRes
 
 	if err := s.cleanupConnection(); err != nil {
 		s.mutex.Unlock()
-		// todo review to update the status in case any type of error
+		if errors.Is(err, ErrServiceNotUp) {
+			log.Debugf("Down called while service not up: %v", err)
+			return nil, err
+		}
 		log.Errorf("failed to shut down properly: %v", err)
 		return nil, err
 	}
@@ -1154,7 +1157,7 @@ func (s *Server) cleanupConnection() error {
 	// making the run loop the sole owner of engine shutdown.
 	if engine != nil {
 		if err := engine.Stop(); err != nil {
-			return err
+			log.Errorf("failed to stop engine during cleanup: %v", err)
 		}
 	}
 

client/ssh/server/command_execution_unix.go

@@ -20,8 +20,6 @@ import (
 	"github.com/creack/pty"
 	"github.com/gliderlabs/ssh"
 	log "github.com/sirupsen/logrus"
-
-	shellutil "github.com/netbirdio/netbird/client/internal/shell"
 )
 
 // ptyManager manages Pty file operations with thread safety
@@ -148,10 +146,10 @@ func (s *Server) createShellCommand(ctx context.Context, shell string, args []st
 
 // prepareCommandEnv prepares environment variables for command execution on Unix
 func (s *Server) prepareCommandEnv(_ *log.Entry, localUser *user.User, session ssh.Session) []string {
-	env := shellutil.PrepareUserEnv(localUser, shellutil.GetUserShell(localUser.Uid))
+	env := prepareUserEnv(localUser, getUserShell(localUser.Uid))
 	env = append(env, prepareSSHEnv(session)...)
 	for _, v := range session.Environ() {
-		if shellutil.AcceptEnv(v) {
+		if acceptEnv(v) {
 			env = append(env, v)
 		}
 	}

client/ssh/server/command_execution_windows.go

@@ -15,7 +15,6 @@ import (
 	"golang.org/x/sys/windows"
 	"golang.org/x/sys/windows/registry"
 
-	shellutil "github.com/netbirdio/netbird/client/internal/shell"
 	"github.com/netbirdio/netbird/client/ssh/server/winpty"
 )
 
@@ -248,10 +247,10 @@ func (s *Server) prepareCommandEnv(logger *log.Entry, localUser *user.User, sess
 	userEnv, err := s.getUserEnvironment(logger, username, domain)
 	if err != nil {
 		log.Debugf("failed to get user environment for %s\\%s, using fallback: %v", domain, username, err)
-		env := shellutil.PrepareUserEnv(localUser, shellutil.GetUserShell(localUser.Uid))
+		env := prepareUserEnv(localUser, getUserShell(localUser.Uid))
 		env = append(env, prepareSSHEnv(session)...)
 		for _, v := range session.Environ() {
-			if shellutil.AcceptEnv(v) {
+			if acceptEnv(v) {
 				env = append(env, v)
 			}
 		}
@@ -261,7 +260,7 @@ func (s *Server) prepareCommandEnv(logger *log.Entry, localUser *user.User, sess
 	env := userEnv
 	env = append(env, prepareSSHEnv(session)...)
 	for _, v := range session.Environ() {
-		if shellutil.AcceptEnv(v) {
+		if acceptEnv(v) {
 			env = append(env, v)
 		}
 	}
@@ -274,7 +273,7 @@ func (s *Server) handlePtyLogin(logger *log.Entry, session ssh.Session, privileg
 		return false
 	}
 
-	shell := shellutil.GetUserShell(privilegeResult.User.Uid)
+	shell := getUserShell(privilegeResult.User.Uid)
 	logger.Infof("starting interactive shell: %s", shell)
 
 	s.executeCommandWithPty(logger, session, nil, privilegeResult, ptyReq, nil)
@@ -385,7 +384,7 @@ func (s *Server) executeCommandWithPty(logger *log.Entry, session ssh.Session, _
 	}
 
 	username, domain := s.parseUsername(localUser.Username)
-	shell := shellutil.GetUserShell(localUser.Uid)
+	shell := getUserShell(localUser.Uid)
 
 	req := PtyExecutionRequest{
 		Shell:    shell,

client/ssh/server/getent_cgo_unix.go

@@ -0,0 +1,24 @@
+//go:build cgo && !osusergo && !windows
+
+package server
+
+import "os/user"
+
+// lookupWithGetent with CGO delegates directly to os/user.Lookup.
+// When CGO is enabled, os/user uses libc (getpwnam_r) which goes through
+// the NSS stack natively. If it fails, the user truly doesn't exist and
+// getent would also fail.
+func lookupWithGetent(username string) (*user.User, error) {
+	return user.Lookup(username)
+}
+
+// currentUserWithGetent with CGO delegates directly to os/user.Current.
+func currentUserWithGetent() (*user.User, error) {
+	return user.Current()
+}
+
+// groupIdsWithFallback with CGO delegates directly to user.GroupIds.
+// libc's getgrouplist handles NSS groups natively.
+func groupIdsWithFallback(u *user.User) ([]string, error) {
+	return u.GroupIds()
+}

client/ssh/server/getent_nocgo_unix.go

@@ -1,6 +1,6 @@
 //go:build (!cgo || osusergo) && !windows
 
-package shell
+package server
 
 import (
 	"os"
@@ -10,10 +10,10 @@ import (
 	log "github.com/sirupsen/logrus"
 )
 
-// LookupWithGetent looks up a user by name, falling back to getent if os/user fails.
+// lookupWithGetent looks up a user by name, falling back to getent if os/user fails.
 // Without CGO, os/user only reads /etc/passwd and misses NSS-provided users.
 // getent goes through the host's NSS stack.
-func LookupWithGetent(username string) (*user.User, error) {
+func lookupWithGetent(username string) (*user.User, error) {
 	u, err := user.Lookup(username)
 	if err == nil {
 		return u, nil
@@ -22,7 +22,7 @@ func LookupWithGetent(username string) (*user.User, error) {
 	stdErr := err
 	log.Debugf("os/user.Lookup(%q) failed, trying getent: %v", username, err)
 
-	u, _, getentErr := runGetentPasswd(username)
+	u, _, getentErr := runGetent(username)
 	if getentErr != nil {
 		log.Debugf("getent fallback for %q also failed: %v", username, getentErr)
 		return nil, stdErr
@@ -31,26 +31,8 @@ func LookupWithGetent(username string) (*user.User, error) {
 	return u, nil
 }
 
-// LookupGroupWithGetent returns the resolved group from either a gid or groupname,
-// falling back to getent if os/user fails (NSS groups under nocgo).
-func LookupGroupWithGetent(name string) (*user.Group, error) {
-	g, err := user.LookupGroup(name)
-	if err == nil {
-		return g, nil
-	}
-
-	stdErr := err
-	log.Debugf("os/user.LookupGroup(%q) failed, trying getent: %v", name, err)
-	g, getentErr := runGetentGroup(name)
-	if getentErr != nil {
-		log.Debugf("getent fallback for %q also failed: %v", name, getentErr)
-		return nil, stdErr
-	}
-	return g, nil
-}
-
-// CurrentUserWithGetent gets the current user, falling back to getent if os/user fails.
-func CurrentUserWithGetent() (*user.User, error) {
+// currentUserWithGetent gets the current user, falling back to getent if os/user fails.
+func currentUserWithGetent() (*user.User, error) {
 	u, err := user.Current()
 	if err == nil {
 		return u, nil
@@ -60,7 +42,7 @@ func CurrentUserWithGetent() (*user.User, error) {
 	uid := strconv.Itoa(os.Getuid())
 	log.Debugf("os/user.Current() failed, trying getent with UID %s: %v", uid, err)
 
-	u, _, getentErr := runGetentPasswd(uid)
+	u, _, getentErr := runGetent(uid)
 	if getentErr != nil {
 		return nil, stdErr
 	}
@@ -68,14 +50,14 @@ func CurrentUserWithGetent() (*user.User, error) {
 	return u, nil
 }
 
-// GroupIdsWithFallback gets group IDs for a user via the id command first,
+// groupIdsWithFallback gets group IDs for a user via the id command first,
 // falling back to user.GroupIds().
-// NOTE: unlike LookupWithGetent/CurrentUserWithGetent which try stdlib first,
+// NOTE: unlike lookupWithGetent/currentUserWithGetent which try stdlib first,
 // this intentionally tries `id -G` first because without CGO, user.GroupIds()
 // only reads /etc/group and silently returns incomplete results for NSS users
 // (no error, just missing groups). The id command goes through NSS and returns
 // the full set.
-func GroupIdsWithFallback(u *user.User) ([]string, error) {
+func groupIdsWithFallback(u *user.User) ([]string, error) {
 	ids, err := runIdGroups(u.Username)
 	if err == nil {
 		return ids, nil

client/ssh/server/getent_test.go

@@ -1,4 +1,4 @@
-package shell
+package server
 
 import (
 	"os/user"
@@ -15,7 +15,7 @@ func TestLookupWithGetent_CurrentUser(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
-	u, err := LookupWithGetent(current.Username)
+	u, err := lookupWithGetent(current.Username)
 	require.NoError(t, err)
 	assert.Equal(t, current.Username, u.Username)
 	assert.Equal(t, current.Uid, u.Uid)
@@ -23,7 +23,7 @@ func TestLookupWithGetent_CurrentUser(t *testing.T) {
 }
 
 func TestLookupWithGetent_NonexistentUser(t *testing.T) {
-	_, err := LookupWithGetent("nonexistent_user_xyzzy_12345")
+	_, err := lookupWithGetent("nonexistent_user_xyzzy_12345")
 	require.Error(t, err, "should fail for nonexistent user")
 }
 
@@ -31,7 +31,7 @@ func TestCurrentUserWithGetent(t *testing.T) {
 	stdUser, err := user.Current()
 	require.NoError(t, err)
 
-	u, err := CurrentUserWithGetent()
+	u, err := currentUserWithGetent()
 	require.NoError(t, err)
 	assert.Equal(t, stdUser.Uid, u.Uid)
 	assert.Equal(t, stdUser.Username, u.Username)
@@ -41,7 +41,7 @@ func TestGroupIdsWithFallback_CurrentUser(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
-	groups, err := GroupIdsWithFallback(current)
+	groups, err := groupIdsWithFallback(current)
 	require.NoError(t, err)
 	require.NotEmpty(t, groups, "current user should have at least one group")
 
@@ -56,7 +56,7 @@ func TestGroupIdsWithFallback_CurrentUser(t *testing.T) {
 func TestGetShellFromGetent_CurrentUser(t *testing.T) {
 	if runtime.GOOS == "windows" {
 		// Windows stub always returns empty, which is correct
-		shell := GetShellFromGetent("1000")
+		shell := getShellFromGetent("1000")
 		assert.Empty(t, shell, "Windows stub should return empty")
 		return
 	}
@@ -65,9 +65,9 @@ func TestGetShellFromGetent_CurrentUser(t *testing.T) {
 	require.NoError(t, err)
 
 	// getent may not be available on all systems (e.g., macOS without Homebrew getent)
-	shell := GetShellFromGetent(current.Uid)
+	shell := getShellFromGetent(current.Uid)
 	if shell == "" {
-		t.Log("GetShellFromGetent returned empty, getent may not be available")
+		t.Log("getShellFromGetent returned empty, getent may not be available")
 		return
 	}
 	assert.True(t, shell[0] == '/', "shell should be an absolute path, got %q", shell)
@@ -78,7 +78,7 @@ func TestLookupWithGetent_RootUser(t *testing.T) {
 		t.Skip("no root user on Windows")
 	}
 
-	u, err := LookupWithGetent("root")
+	u, err := lookupWithGetent("root")
 	if err != nil {
 		t.Skip("root user not available on this system")
 	}
@@ -86,25 +86,25 @@ func TestLookupWithGetent_RootUser(t *testing.T) {
 }
 
 // TestIntegration_FullLookupChain exercises the complete user lookup chain
-// against the real system, testing that all wrappers (LookupWithGetent,
-// CurrentUserWithGetent, GroupIdsWithFallback, GetShellFromGetent) produce
+// against the real system, testing that all wrappers (lookupWithGetent,
+// currentUserWithGetent, groupIdsWithFallback, getShellFromGetent) produce
 // consistent and correct results when composed together.
 func TestIntegration_FullLookupChain(t *testing.T) {
-	// Step 1: CurrentUserWithGetent must resolve the running user.
-	current, err := CurrentUserWithGetent()
-	require.NoError(t, err, "CurrentUserWithGetent must resolve the running user")
+	// Step 1: currentUserWithGetent must resolve the running user.
+	current, err := currentUserWithGetent()
+	require.NoError(t, err, "currentUserWithGetent must resolve the running user")
 	require.NotEmpty(t, current.Uid)
 	require.NotEmpty(t, current.Username)
 
-	// Step 2: LookupWithGetent by the same username must return matching identity.
-	byName, err := LookupWithGetent(current.Username)
+	// Step 2: lookupWithGetent by the same username must return matching identity.
+	byName, err := lookupWithGetent(current.Username)
 	require.NoError(t, err)
 	assert.Equal(t, current.Uid, byName.Uid, "lookup by name should return same UID")
 	assert.Equal(t, current.Gid, byName.Gid, "lookup by name should return same GID")
 	assert.Equal(t, current.HomeDir, byName.HomeDir, "lookup by name should return same home")
 
-	// Step 3: GroupIdsWithFallback must return at least the primary GID.
-	groups, err := GroupIdsWithFallback(current)
+	// Step 3: groupIdsWithFallback must return at least the primary GID.
+	groups, err := groupIdsWithFallback(current)
 	require.NoError(t, err)
 	require.NotEmpty(t, groups, "user must have at least one group")
 
@@ -120,10 +120,10 @@ func TestIntegration_FullLookupChain(t *testing.T) {
 	}
 	assert.True(t, foundPrimary, "primary GID %s should appear in supplementary groups", current.Gid)
 
-	// Step 4: GetShellFromGetent should either return a valid shell path or empty
+	// Step 4: getShellFromGetent should either return a valid shell path or empty
 	// (empty is OK when getent is not available, e.g. macOS without Homebrew getent).
 	if runtime.GOOS != "windows" {
-		shell := GetShellFromGetent(current.Uid)
+		shell := getShellFromGetent(current.Uid)
 		if shell != "" {
 			assert.True(t, shell[0] == '/', "shell should be an absolute path, got %q", shell)
 		}
@@ -131,17 +131,17 @@ func TestIntegration_FullLookupChain(t *testing.T) {
 }
 
 // TestIntegration_LookupAndGroupsConsistency verifies that a user resolved via
-// LookupWithGetent can have their groups resolved via GroupIdsWithFallback,
+// lookupWithGetent can have their groups resolved via groupIdsWithFallback,
 // testing the handoff between the two functions as used by the SSH server.
 func TestIntegration_LookupAndGroupsConsistency(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
 	// Simulate the SSH server flow: lookup user, then get their groups.
-	resolved, err := LookupWithGetent(current.Username)
+	resolved, err := lookupWithGetent(current.Username)
 	require.NoError(t, err)
 
-	groups, err := GroupIdsWithFallback(resolved)
+	groups, err := groupIdsWithFallback(resolved)
 	require.NoError(t, err)
 	require.NotEmpty(t, groups, "resolved user must have groups")
 
@@ -156,7 +156,7 @@ func TestIntegration_LookupAndGroupsConsistency(t *testing.T) {
 }
 
 // TestIntegration_ShellLookupChain tests the full shell resolution chain
-// (getShellFromPasswd -> GetShellFromGetent -> $SHELL -> default) on Unix.
+// (getShellFromPasswd -> getShellFromGetent -> $SHELL -> default) on Unix.
 func TestIntegration_ShellLookupChain(t *testing.T) {
 	if runtime.GOOS == "windows" {
 		t.Skip("Unix shell lookup not applicable on Windows")
@@ -165,8 +165,8 @@ func TestIntegration_ShellLookupChain(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
-	// GetUserShell is the top-level function used by the SSH server.
-	shell := GetUserShell(current.Uid)
-	require.NotEmpty(t, shell, "GetUserShell must always return a shell")
+	// getUserShell is the top-level function used by the SSH server.
+	shell := getUserShell(current.Uid)
+	require.NotEmpty(t, shell, "getUserShell must always return a shell")
 	assert.True(t, shell[0] == '/', "shell should be an absolute path, got %q", shell)
 }

client/ssh/server/getent_unix.go

@@ -1,6 +1,6 @@
 //go:build !windows
 
-package shell
+package server
 
 import (
 	"context"
@@ -14,26 +14,19 @@ import (
 
 const getentTimeout = 5 * time.Second
 
-// GetShellFromGetent gets a user's login shell via getent by UID.
+// getShellFromGetent gets a user's login shell via getent by UID.
 // This is needed even with CGO because getShellFromPasswd reads /etc/passwd
 // directly and won't find NSS-provided users there.
-func GetShellFromGetent(userID string) string {
-	_, shell, err := runGetentPasswd(userID)
+func getShellFromGetent(userID string) string {
+	_, shell, err := runGetent(userID)
 	if err != nil {
 		return ""
 	}
 	return shell
 }
 
-// GetUserFromGetent returns the resolved user from either a uid or username,
-// going through the host's NSS stack.
-func GetUserFromGetent(query string) (*user.User, error) {
-	u, _, err := runGetentPasswd(query)
-	return u, err
-}
-
-// runGetentPasswd executes `getent passwd <query>` and returns the user and login shell.
-func runGetentPasswd(query string) (*user.User, string, error) {
+// runGetent executes `getent passwd <query>` and returns the user and login shell.
+func runGetent(query string) (*user.User, string, error) {
 	if !validateGetentInput(query) {
 		return nil, "", fmt.Errorf("invalid getent input: %q", query)
 	}
@@ -49,24 +42,7 @@ func runGetentPasswd(query string) (*user.User, string, error) {
 	return parseGetentPasswd(string(out))
 }
 
-// runGetentGroup executes `getent group <query>` and returns the group.
-func runGetentGroup(query string) (*user.Group, error) {
-	if !validateGetentInput(query) {
-		return nil, fmt.Errorf("invalid getent input: %q", query)
-	}
-
-	ctx, cancel := context.WithTimeout(context.Background(), getentTimeout)
-	defer cancel()
-
-	out, err := exec.CommandContext(ctx, "getent", "group", query).Output()
-	if err != nil {
-		return nil, fmt.Errorf("getent group %s: %w", query, err)
-	}
-
-	return parseGetentGroup(string(out))
-}
-
-// parseGetentPasswd parses getent passwd output: "name:x:uid:gid:gecos:home:shell".
+// parseGetentPasswd parses getent passwd output: "name:x:uid:gid:gecos:home:shell"
 func parseGetentPasswd(output string) (*user.User, string, error) {
 	fields := strings.SplitN(strings.TrimSpace(output), ":", 8)
 	if len(fields) < 6 {
@@ -91,20 +67,6 @@ func parseGetentPasswd(output string) (*user.User, string, error) {
 	}, shell, nil
 }
 
-// parseGetentGroup parses getent group output: "group:x:gid:members".
-func parseGetentGroup(output string) (*user.Group, error) {
-	fields := strings.SplitN(strings.TrimSpace(output), ":", 8)
-	if len(fields) < 4 {
-		return nil, fmt.Errorf("unexpected getent output (need 4+ fields): %q", output)
-	}
-
-	if fields[0] == "" || fields[2] == "" {
-		return nil, fmt.Errorf("missing required fields in getent output: %q", output)
-	}
-
-	return &user.Group{Gid: fields[2], Name: fields[0]}, nil
-}
-
 // validateGetentInput checks that the input is safe to pass to getent or id.
 // Allows POSIX usernames, numeric UIDs, and common NSS extensions
 // (@ for Kerberos, $ for Samba, + for NIS compat). A leading hyphen is

client/ssh/server/getent_unix_test.go

@@ -1,6 +1,6 @@
 //go:build !windows
 
-package shell
+package server
 
 import (
 	"os/exec"
@@ -198,7 +198,7 @@ func TestRunGetent_RootUser(t *testing.T) {
 		t.Skip("getent not available on this system")
 	}
 
-	u, shell, err := runGetentPasswd("root")
+	u, shell, err := runGetent("root")
 	require.NoError(t, err)
 	assert.Equal(t, "root", u.Username)
 	assert.Equal(t, "0", u.Uid)
@@ -211,7 +211,7 @@ func TestRunGetent_ByUID(t *testing.T) {
 		t.Skip("getent not available on this system")
 	}
 
-	u, _, err := runGetentPasswd("0")
+	u, _, err := runGetent("0")
 	require.NoError(t, err)
 	assert.Equal(t, "root", u.Username)
 	assert.Equal(t, "0", u.Uid)
@@ -222,15 +222,15 @@ func TestRunGetent_NonexistentUser(t *testing.T) {
 		t.Skip("getent not available on this system")
 	}
 
-	_, _, err := runGetentPasswd("nonexistent_user_xyzzy_12345")
+	_, _, err := runGetent("nonexistent_user_xyzzy_12345")
 	assert.Error(t, err)
 }
 
 func TestRunGetent_InvalidInput(t *testing.T) {
-	_, _, err := runGetentPasswd("")
+	_, _, err := runGetent("")
 	assert.Error(t, err)
 
-	_, _, err = runGetentPasswd("user\x00name")
+	_, _, err = runGetent("user\x00name")
 	assert.Error(t, err)
 }
 
@@ -239,7 +239,7 @@ func TestRunGetent_NotAvailable(t *testing.T) {
 		t.Skip("getent is available, can't test missing case")
 	}
 
-	_, _, err := runGetentPasswd("root")
+	_, _, err := runGetent("root")
 	assert.Error(t, err, "should fail when getent is not installed")
 }
 
@@ -286,7 +286,7 @@ func TestGetentResultsMatchStdlib(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
-	getentUser, _, err := runGetentPasswd(current.Username)
+	getentUser, _, err := runGetent(current.Username)
 	require.NoError(t, err)
 
 	assert.Equal(t, current.Username, getentUser.Username, "username should match")
@@ -303,7 +303,7 @@ func TestGetentResultsMatchStdlib_ByUID(t *testing.T) {
 	current, err := user.Current()
 	require.NoError(t, err)
 
-	getentUser, _, err := runGetentPasswd(current.Uid)
+	getentUser, _, err := runGetent(current.Uid)
 	require.NoError(t, err)
 
 	assert.Equal(t, current.Username, getentUser.Username, "username should match when looked up by UID")
@@ -359,7 +359,7 @@ func TestGetShellFromPasswd_CurrentUser(t *testing.T) {
 	assert.True(t, shell[0] == '/', "shell should be an absolute path, got %q", shell)
 
 	if _, err := exec.LookPath("getent"); err == nil {
-		_, getentShell, getentErr := runGetentPasswd(current.Uid)
+		_, getentShell, getentErr := runGetent(current.Uid)
 		if getentErr == nil && getentShell != "" {
 			assert.Equal(t, getentShell, shell, "shell from /etc/passwd should match getent")
 		}
@@ -403,7 +403,7 @@ func TestGetShellFromPasswd_MatchesGetentForKnownUsers(t *testing.T) {
 			continue
 		}
 
-		_, getentShell, err := runGetentPasswd(uid)
+		_, getentShell, err := runGetent(uid)
 		if err != nil {
 			continue
 		}

client/ssh/server/getent_windows.go

@@ -0,0 +1,26 @@
+//go:build windows
+
+package server
+
+import "os/user"
+
+// lookupWithGetent on Windows just delegates to os/user.Lookup.
+// Windows does not use NSS/getent; its user lookup works without CGO.
+func lookupWithGetent(username string) (*user.User, error) {
+	return user.Lookup(username)
+}
+
+// currentUserWithGetent on Windows just delegates to os/user.Current.
+func currentUserWithGetent() (*user.User, error) {
+	return user.Current()
+}
+
+// getShellFromGetent is a no-op on Windows; shell resolution uses PowerShell detection.
+func getShellFromGetent(_ string) string {
+	return ""
+}
+
+// groupIdsWithFallback on Windows just delegates to u.GroupIds().
+func groupIdsWithFallback(u *user.User) ([]string, error) {
+	return u.GroupIds()
+}

client/ssh/server/shell.go

@@ -1,14 +1,17 @@
-package shell
+package server
 
 import (
 	"bufio"
 	"fmt"
+	"net"
 	"os"
 	"os/exec"
 	"os/user"
 	"runtime"
+	"strconv"
 	"strings"
 
+	"github.com/gliderlabs/ssh"
 	log "github.com/sirupsen/logrus"
 )
 
@@ -19,9 +22,9 @@ const (
 	powershellExe = "powershell.exe"
 )
 
-// GetUserShell returns the appropriate shell for the given user ID.
-// Handles all platform-specific logic and fallbacks consistently.
-func GetUserShell(userID string) string {
+// getUserShell returns the appropriate shell for the given user ID
+// Handles all platform-specific logic and fallbacks consistently
+func getUserShell(userID string) string {
 	switch runtime.GOOS {
 	case "windows":
 		return getWindowsUserShell()
@@ -53,7 +56,7 @@ func getUnixUserShell(userID string) string {
 		return shell
 	}
 
-	if shell := GetShellFromGetent(userID); shell != "" {
+	if shell := getShellFromGetent(userID); shell != "" {
 		return shell
 	}
 
@@ -64,7 +67,7 @@ func getUnixUserShell(userID string) string {
 	return defaultUnixShell
 }
 
-// getShellFromPasswd reads the shell from /etc/passwd for the given user ID.
+// getShellFromPasswd reads the shell from /etc/passwd for the given user ID
 func getShellFromPasswd(userID string) string {
 	file, err := os.Open("/etc/passwd")
 	if err != nil {
@@ -98,8 +101,8 @@ func getShellFromPasswd(userID string) string {
 	return ""
 }
 
-// PrepareUserEnv prepares environment variables for user execution.
-func PrepareUserEnv(user *user.User, shell string) []string {
+// prepareUserEnv prepares environment variables for user execution
+func prepareUserEnv(user *user.User, shell string) []string {
 	pathValue := "/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games"
 	if runtime.GOOS == "windows" {
 		pathValue = `C:\Windows\System32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0`
@@ -114,9 +117,9 @@ func PrepareUserEnv(user *user.User, shell string) []string {
 	}
 }
 
-// AcceptEnv checks if an environment variable from an SSH client should be accepted.
-// This is a whitelist of variables that SSH clients can send to the server.
-func AcceptEnv(envVar string) bool {
+// acceptEnv checks if environment variable from SSH client should be accepted
+// This is a whitelist of variables that SSH clients can send to the server
+func acceptEnv(envVar string) bool {
 	varName := envVar
 	if idx := strings.Index(envVar, "="); idx != -1 {
 		varName = envVar[:idx]
@@ -153,3 +156,29 @@ func AcceptEnv(envVar string) bool {
 
 	return false
 }
+
+// prepareSSHEnv prepares SSH protocol-specific environment variables
+// These variables provide information about the SSH connection itself
+func prepareSSHEnv(session ssh.Session) []string {
+	remoteAddr := session.RemoteAddr()
+	localAddr := session.LocalAddr()
+
+	remoteHost, remotePort, err := net.SplitHostPort(remoteAddr.String())
+	if err != nil {
+		remoteHost = remoteAddr.String()
+		remotePort = "0"
+	}
+
+	localHost, localPort, err := net.SplitHostPort(localAddr.String())
+	if err != nil {
+		localHost = localAddr.String()
+		localPort = strconv.Itoa(InternalSSHPort)
+	}
+
+	return []string{
+		// SSH_CLIENT format: "client_ip client_port server_port"
+		fmt.Sprintf("SSH_CLIENT=%s %s %s", remoteHost, remotePort, localPort),
+		// SSH_CONNECTION format: "client_ip client_port server_ip server_port"
+		fmt.Sprintf("SSH_CONNECTION=%s %s %s %s", remoteHost, remotePort, localHost, localPort),
+	}
+}

client/ssh/server/ssh_env.go

@@ -1,35 +0,0 @@
-package server
-
-import (
-	"fmt"
-	"net"
-	"strconv"
-
-	"github.com/gliderlabs/ssh"
-)
-
-// prepareSSHEnv prepares SSH protocol-specific environment variables
-// These variables provide information about the SSH connection itself
-func prepareSSHEnv(session ssh.Session) []string {
-	remoteAddr := session.RemoteAddr()
-	localAddr := session.LocalAddr()
-
-	remoteHost, remotePort, err := net.SplitHostPort(remoteAddr.String())
-	if err != nil {
-		remoteHost = remoteAddr.String()
-		remotePort = "0"
-	}
-
-	localHost, localPort, err := net.SplitHostPort(localAddr.String())
-	if err != nil {
-		localHost = localAddr.String()
-		localPort = strconv.Itoa(InternalSSHPort)
-	}
-
-	return []string{
-		// SSH_CLIENT format: "client_ip client_port server_port"
-		fmt.Sprintf("SSH_CLIENT=%s %s %s", remoteHost, remotePort, localPort),
-		// SSH_CONNECTION format: "client_ip client_port server_ip server_port"
-		fmt.Sprintf("SSH_CONNECTION=%s %s %s %s", remoteHost, remotePort, localHost, localPort),
-	}
-}

client/ssh/server/user_utils.go

@@ -9,8 +9,6 @@ import (
 	"strings"
 
 	log "github.com/sirupsen/logrus"
-
-	shellutil "github.com/netbirdio/netbird/client/internal/shell"
 )
 
 var (
@@ -25,8 +23,8 @@ func isPlatformUnix() bool {
 
 // Dependency injection variables for testing - allows mocking dynamic runtime checks
 var (
-	getCurrentUser         = shellutil.CurrentUserWithGetent
-	lookupUser             = shellutil.LookupWithGetent
+	getCurrentUser         = currentUserWithGetent
+	lookupUser             = lookupWithGetent
 	getCurrentOS           = func() string { return runtime.GOOS }
 	getIsProcessPrivileged = isCurrentProcessPrivileged
 

client/ssh/server/userswitching_unix.go

@@ -16,8 +16,6 @@ import (
 
 	"github.com/gliderlabs/ssh"
 	log "github.com/sirupsen/logrus"
-
-	shellutil "github.com/netbirdio/netbird/client/internal/shell"
 )
 
 // POSIX portable filename character set regex: [a-zA-Z0-9._-]
@@ -162,7 +160,7 @@ func (s *Server) parseUserCredentials(localUser *user.User) (uint32, uint32, []u
 // getSupplementaryGroups retrieves supplementary group IDs for a user.
 // Uses id/getent fallback for NSS users in CGO_ENABLED=0 builds.
 func (s *Server) getSupplementaryGroups(u *user.User) ([]uint32, error) {
-	groupIDStrings, err := shellutil.GroupIdsWithFallback(u)
+	groupIDStrings, err := groupIdsWithFallback(u)
 	if err != nil {
 		return nil, fmt.Errorf("get group IDs for user %s: %w", u.Username, err)
 	}
@@ -198,7 +196,7 @@ func (s *Server) createExecutorCommand(logger *log.Entry, session ssh.Session, l
 		GID:        gid,
 		Groups:     groups,
 		WorkingDir: localUser.HomeDir,
-		Shell:      shellutil.GetUserShell(localUser.Uid),
+		Shell:      getUserShell(localUser.Uid),
 		Command:    session.RawCommand(),
 		PTY:        hasPty,
 	}
@@ -230,7 +228,7 @@ func (s *Server) createPtyCommand(privilegeResult PrivilegeCheckResult, ptyReq s
 func (s *Server) createDirectPtyCommand(session ssh.Session, localUser *user.User, ptyReq ssh.Pty) *exec.Cmd {
 	log.Debugf("creating direct Pty command for user %s (no user switching needed)", localUser.Username)
 
-	shell := shellutil.GetUserShell(localUser.Uid)
+	shell := getUserShell(localUser.Uid)
 	args := s.getShellCommandArgs(shell, session.RawCommand())
 
 	cmd := s.createShellCommand(session.Context(), shell, args)
@@ -247,12 +245,12 @@ func (s *Server) preparePtyEnv(localUser *user.User, ptyReq ssh.Pty, session ssh
 		termType = "xterm-256color"
 	}
 
-	env := shellutil.PrepareUserEnv(localUser, shellutil.GetUserShell(localUser.Uid))
+	env := prepareUserEnv(localUser, getUserShell(localUser.Uid))
 	env = append(env, prepareSSHEnv(session)...)
 	env = append(env, fmt.Sprintf("TERM=%s", termType))
 
 	for _, v := range session.Environ() {
-		if shellutil.AcceptEnv(v) {
+		if acceptEnv(v) {
 			env = append(env, v)
 		}
 	}

client/ssh/server/userswitching_windows.go

@@ -13,8 +13,6 @@ import (
 	"github.com/gliderlabs/ssh"
 	log "github.com/sirupsen/logrus"
 	"golang.org/x/sys/windows"
-
-	shellutil "github.com/netbirdio/netbird/client/internal/shell"
 )
 
 // validateUsername validates Windows usernames according to SAM Account Name rules
@@ -106,7 +104,7 @@ func (s *Server) createExecutorCommand(logger *log.Entry, session ssh.Session, l
 func (s *Server) createUserSwitchCommand(logger *log.Entry, session ssh.Session, localUser *user.User) (*exec.Cmd, func(), error) {
 	username, domain := s.parseUsername(localUser.Username)
 
-	shell := shellutil.GetUserShell(localUser.Uid)
+	shell := getUserShell(localUser.Uid)
 
 	rawCmd := session.RawCommand()
 	var command string

client/system/info.go

@@ -79,13 +79,15 @@ type Info struct {
 	EnableSSHLocalPortForwarding  bool
 	EnableSSHRemotePortForwarding bool
 	DisableSSHAuth                bool
+
+	SyncMessageVersion *int
 }
 
 func (i *Info) SetFlags(
 	rosenpassEnabled, rosenpassPermissive bool,
 	serverSSHAllowed *bool,
 	disableClientRoutes, disableServerRoutes,
-	disableDNS, disableFirewall, blockLANAccess, blockInbound, disableIPv6 bool,
+	disableDNS, disableFirewall, blockLANAccess, blockInbound, disableIPv6 bool, syncMessageVersion *int,
 	enableSSHRoot, enableSSHSFTP, enableSSHLocalPortForwarding, enableSSHRemotePortForwarding *bool,
 	disableSSHAuth *bool,
 ) {
@@ -103,6 +105,8 @@ func (i *Info) SetFlags(
 	i.BlockInbound = blockInbound
 	i.DisableIPv6 = disableIPv6
 
+	i.SyncMessageVersion = syncMessageVersion
+
 	if enableSSHRoot != nil {
 		i.EnableSSHRoot = *enableSSHRoot
 	}

client/ui/autostart_default.go

@@ -51,7 +51,7 @@ func autostartDisabledByMDM(policy *mdm.Policy) bool {
 // netbirdFootprintExists reports whether the machine already carries NetBird
 // daemon config or state, meaning this is not a genuinely fresh install. It is
 // the update-safety gate for the autostart default: upgrading users always
-// have a footprint, so an update can never trigger a login-item write.
+// have a footprint, so an update can never trigger a autostart entry write.
 func netbirdFootprintExists() bool {
 	candidates := []string{
 		profilemanager.DefaultConfigPath,
@@ -69,9 +69,23 @@ func netbirdFootprintExists() bool {
 // applyAutostartDefault runs the one-time launch-on-login default for genuinely
 // fresh installs. The autostartInitialized marker is persisted before any
 // enable attempt so a crash mid-flow degrades to "never enabled" instead of
-// retrying login-item writes on every launch. A user's later disable in
+// retrying autostart entry writes on every launch. A user's later disable in
 // Settings is never overridden: the marker guarantees at-most-once, ever.
 func applyAutostartDefault(ctx context.Context, autostart *services.Autostart, prefs *preferences.Store, prefsFileExisted bool) {
+	mdmDisabled := autostartDisabledByMDM(mdm.LoadPolicy())
+
+	if mdmDisabled {
+		if enabled, err := autostart.IsEnabled(ctx); err != nil {
+			log.Warnf("MDM disableAutostart: read autostart state: %v", err)
+		} else if enabled {
+			if err := autostart.SetEnabled(ctx, false); err != nil {
+				log.Warnf("MDM disableAutostart: force off failed: %v", err)
+			} else {
+				log.Info("MDM disableAutostart enforced: autostart turned off")
+			}
+		}
+	}
+
 	priorFootprint := netbirdFootprintExists() || prefsFileExisted
 
 	if prefs.Get().AutostartInitialized {
@@ -84,7 +98,7 @@ func applyAutostartDefault(ctx context.Context, autostart *services.Autostart, p
 
 	state := autostartDefaultState{
 		supported:    autostart.Supported(ctx),
-		mdmDisabled:  autostartDisabledByMDM(mdm.LoadPolicy()),
+		mdmDisabled:  mdmDisabled,
 		priorInstall: priorFootprint,
 	}
 	enable, reason := shouldEnableAutostartDefault(state)

client/ui/frontend/src/components/empty-state/DaemonOutdatedOverlay.tsx

@@ -1,10 +1,13 @@
+import { useEffect, useState } from "react";
 import { useTranslation } from "react-i18next";
 import { AlertTriangleIcon, DownloadIcon } from "lucide-react";
 import { Browser } from "@wailsio/runtime";
+import { Version } from "@bindings/services";
 import { Button } from "@/components/buttons/Button";
 import { useStatus } from "@/contexts/StatusContext.tsx";
 
 const RELEASES_URL = "https://github.com/netbirdio/netbird/releases/latest";
+const RC_RELEASES_URL = "https://pkgs.netbird.io/releases/rc";
 
 function openUrl(url: string) {
     Browser.OpenURL(url).catch(() => globalThis.open(url, "_blank"));
@@ -12,7 +15,26 @@ function openUrl(url: string) {
 
 export const DaemonOutdatedOverlay = () => {
     const { t } = useTranslation();
-    const { isDaemonOutdated } = useStatus();
+    const { status, isDaemonOutdated } = useStatus();
+
+    const [guiVersion, setGuiVersion] = useState<string>("-");
+    const clientVersion = status?.daemonVersion ?? "—";
+
+    const isRc = /-rc/i.test(guiVersion) || /-rc/i.test(clientVersion);
+    const downloadUrl = isRc ? RC_RELEASES_URL : RELEASES_URL;
+
+    useEffect(() => {
+        if (!isDaemonOutdated) return;
+        let cancelled = false;
+        Version.GUI()
+            .then((v) => {
+                if (!cancelled) setGuiVersion(v);
+            })
+            .catch((err) => console.error("[DaemonOutdatedOverlay] GUI version error", err));
+        return () => {
+            cancelled = true;
+        };
+    }, [isDaemonOutdated]);
 
     if (!isDaemonOutdated) return null;
 
@@ -38,10 +60,37 @@ export const DaemonOutdatedOverlay = () => {
                     <p className={"text-sm text-nb-gray-300"}>{t("daemon.outdated.description")}</p>
                 </div>
 
+                <div className={"flex flex-col items-center gap-0.5 text-center"}>
+                    <p className={"text-sm font-semibold text-nb-gray-100"}>
+                        {clientVersion === "development" ? (
+                            <span>
+                                {t("settings.about.clientName")}{" "}
+                                <span className={"font-mono text-yellow-400"}>
+                                    {t("settings.about.development")}
+                                </span>
+                            </span>
+                        ) : (
+                            t("settings.about.client", { version: clientVersion })
+                        )}
+                    </p>
+                    <p className={"text-sm font-medium text-nb-gray-250"}>
+                        {guiVersion === "development" ? (
+                            <span>
+                                {t("settings.about.guiName")}{" "}
+                                <span className={"font-mono text-yellow-400"}>
+                                    {t("settings.about.development")}
+                                </span>
+                            </span>
+                        ) : (
+                            t("settings.about.gui", { version: guiVersion })
+                        )}
+                    </p>
+                </div>
+
                 <div className={"wails-no-draggable"}>
-                    <Button variant={"primary"} size={"xs"} onClick={() => openUrl(RELEASES_URL)}>
+                    <Button variant={"primary"} size={"xs"} onClick={() => openUrl(downloadUrl)}>
                         <DownloadIcon size={14} />
-                        {t("update.card.getInstaller")}
+                        {t("daemon.outdated.download")}
                     </Button>
                 </div>
             </div>

client/ui/frontend/src/contexts/ProfileContext.tsx

@@ -28,6 +28,7 @@ type ProfileContextValue = {
     loaded: boolean;
     refresh: () => Promise<void>;
     switchProfile: (id: string) => Promise<void>;
+    switchProfileNoConnect: (id: string) => Promise<void>;
     addProfile: (name: string) => Promise<string>;
     removeProfile: (id: string) => Promise<void>;
     renameProfile: (id: string, newName: string) => Promise<void>;
@@ -112,6 +113,16 @@ export const ProfileProvider = ({ children }: { children: ReactNode }) => {
         [username, refresh],
     );
 
+    // Manage-profiles variant: switches without connecting, so the user can
+    // still adjust the management URL before bringing the connection up.
+    const switchProfileNoConnect = useCallback(
+        async (id: string) => {
+            await ProfileSwitcher.SwitchActiveNoConnect({ profileName: id, username });
+            await refresh();
+        },
+        [username, refresh],
+    );
+
     // addProfile creates a profile by display name and returns the
     // daemon-generated ID, so the caller can immediately address it by ID.
     const addProfile = useCallback(
@@ -158,6 +169,7 @@ export const ProfileProvider = ({ children }: { children: ReactNode }) => {
             loaded,
             refresh,
             switchProfile,
+            switchProfileNoConnect,
             addProfile,
             removeProfile,
             renameProfile,
@@ -171,6 +183,7 @@ export const ProfileProvider = ({ children }: { children: ReactNode }) => {
             loaded,
             refresh,
             switchProfile,
+            switchProfileNoConnect,
             addProfile,
             removeProfile,
             renameProfile,

client/ui/frontend/src/modules/profiles/ProfilesTab.tsx

@@ -45,7 +45,7 @@ export function ProfilesTab() {
         activeProfileId,
         loaded,
         username,
-        switchProfile,
+        switchProfileNoConnect,
         addProfile,
         removeProfile,
         renameProfile,
@@ -100,7 +100,7 @@ export function ProfilesTab() {
             confirmLabel: t("profile.switch.confirm"),
         });
         if (!ok) return;
-        await guarded(i18next.t("profile.error.switchTitle"), () => switchProfile(id));
+        await guarded(i18next.t("profile.error.switchTitle"), () => switchProfileNoConnect(id));
     };
 
     const handleDeregister = async (id: string, name: string) => {
@@ -129,14 +129,13 @@ export function ProfilesTab() {
         await guarded(i18next.t("profile.error.createTitle"), async () => {
             const id = await addProfile(name);
             // SetConfig is keyed by the new profile's ID, so it writes the
-            // not-yet-active profile. Write before switching so any reconnect
-            // targets the right deployment.
+            // not-yet-active profile before the switch makes it current.
             if (!isNetbirdCloud(managementUrl)) {
                 await SettingsSvc.SetConfig(
                     new SetConfigParams({ profileName: id, username, managementUrl }),
                 );
             }
-            await switchProfile(id);
+            await switchProfileNoConnect(id);
         });
     };
 

client/ui/frontend/src/modules/session/SessionExpirationDialog.tsx

@@ -73,6 +73,13 @@ export default function SessionExpirationDialog() {
 
         let offCancel: (() => void) | undefined;
 
+        // Return the dialog to its interactive state and dismiss the browser popup
+        const resetDialog = () => {
+            offCancel?.();
+            WindowManager.CloseBrowserLogin().catch(console.error);
+            setBusy(false);
+        };
+
         try {
             const start = await Session.RequestExtend({ hint: "" });
             const uri = start.verificationUriComplete || start.verificationUri;
@@ -105,25 +112,22 @@ export default function SessionExpirationDialog() {
             if (outcome.kind === "cancel") {
                 waitPromise.cancel?.();
                 waitPromise.catch(() => {});
+                resetDialog();
                 return;
             }
 
             // Another surface owns this flow; keep the dialog open to retry.
             if (outcome.result.preempted) {
+                resetDialog();
                 return;
             }
-
-            // Close before the popup so the restore can't flash this window back.
-            WindowManager.CloseSessionExpiration().catch(console.error);
+            WindowManager.CloseRenewFlow().catch(console.error);
         } catch (e) {
+            resetDialog();
             await errorDialog({
                 Title: t("sessionExpiration.extendFailedTitle"),
                 Message: formatErrorMessage(e),
             });
-        } finally {
-            offCancel?.();
-            WindowManager.CloseBrowserLogin().catch(console.error);
-            setBusy(false);
         }
     }, [busy, t]);
 
@@ -139,12 +143,11 @@ export default function SessionExpirationDialog() {
             });
             WindowManager.CloseSessionExpiration().catch(console.error);
         } catch (e) {
+            setBusy(false);
             await errorDialog({
                 Title: t("sessionExpiration.logoutFailedTitle"),
                 Message: formatErrorMessage(e),
             });
-        } finally {
-            setBusy(false);
         }
     }, [busy, t]);
 

client/ui/frontend/src/modules/welcome/WelcomeStepTray.tsx

@@ -22,6 +22,9 @@ type WelcomeStepTrayProps = {
 export function WelcomeStepTray({ onContinue }: Readonly<WelcomeStepTrayProps>) {
     const { t } = useTranslation();
     const trayScreenshot = trayScreenshotForOS();
+    // macOS has no tray — the icon sits in the menu bar, so the copy says so.
+    const titleKey = isMacOS() ? "welcome.titleMac" : "welcome.title";
+    const descriptionKey = isMacOS() ? "welcome.descriptionMac" : "welcome.description";
 
     return (
         <>
@@ -36,9 +39,9 @@ export function WelcomeStepTray({ onContinue }: Readonly<WelcomeStepTrayProps>)
 
             <div className={"flex w-full flex-col gap-1"}>
                 <DialogHeading id={"nb-welcome-title"} align={"left"}>
-                    {t("welcome.title")}
+                    {t(titleKey)}
                 </DialogHeading>
-                <DialogDescription align={"left"}>{t("welcome.description")}</DialogDescription>
+                <DialogDescription align={"left"}>{t(descriptionKey)}</DialogDescription>
             </div>
 
             <DialogActions>

client/ui/i18n/locales/de/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Suchen Sie NetBird in der Taskleiste"
     },
+    "welcome.titleMac": {
+        "message": "Suchen Sie NetBird in der Menüleiste"
+    },
     "welcome.description": {
         "message": "NetBird läuft in Ihrer Taskleiste. Klicken Sie auf das Symbol, um sich zu verbinden, Profile zu wechseln oder die Einstellungen zu öffnen."
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird läuft in Ihrer Menüleiste. Klicken Sie auf das Symbol, um sich zu verbinden, Profile zu wechseln oder die Einstellungen zu öffnen."
+    },
     "welcome.continue": {
         "message": "Weiter"
     },
@@ -1293,10 +1299,13 @@
         "message": "Dokumentation"
     },
     "daemon.outdated.title": {
-        "message": "NetBird-Dienst ist veraltet"
+        "message": "NetBird Client ist veraltet"
     },
     "daemon.outdated.description": {
-        "message": "Aktualisieren Sie den NetBird-Dienst, um diese App zu verwenden."
+        "message": "Die neue GUI ist nicht mit Ihrem älteren Client kompatibel. Aktualisieren Sie Ihren Client, um die neue Anwendung zu verwenden."
+    },
+    "daemon.outdated.download": {
+        "message": "Neueste Version herunterladen"
     },
     "error.jwt_clock_skew": {
         "message": "Anmeldung fehlgeschlagen: Die Uhr dieses Geräts ist nicht mit dem Server synchron. Bitte synchronisieren Sie die Systemuhr und versuchen Sie es erneut."

client/ui/i18n/locales/en/common.json

@@ -1377,11 +1377,19 @@
     },
     "welcome.title": {
         "message": "Look for NetBird in your tray",
-        "description": "Heading on the first onboarding step, pointing the user to the tray icon. 'tray' = system tray / menu bar."
+        "description": "Heading on the first onboarding step, pointing the user to the tray icon. Shown on Windows and Linux; macOS uses welcome.titleMac."
+    },
+    "welcome.titleMac": {
+        "message": "Look for NetBird in your menu bar",
+        "description": "Heading on the first onboarding step on macOS, pointing the user to the menu bar icon. Use your language's Apple term for the macOS menu bar."
     },
     "welcome.description": {
         "message": "NetBird lives in your tray. Click the icon to connect, switch profiles, or open settings.",
-        "description": "Body of the first onboarding step explaining the tray icon."
+        "description": "Body of the first onboarding step explaining the tray icon. Shown on Windows and Linux; macOS uses welcome.descriptionMac."
+    },
+    "welcome.descriptionMac": {
+        "message": "NetBird lives in your menu bar. Click the icon to connect, switch profiles, or open settings.",
+        "description": "Body of the first onboarding step on macOS explaining the menu bar icon. Use your language's Apple term for the macOS menu bar."
     },
     "welcome.continue": {
         "message": "Continue",
@@ -1724,12 +1732,16 @@
         "description": "Documentation link on the daemon-unavailable overlay."
     },
     "daemon.outdated.title": {
-        "message": "NetBird Service Is Outdated",
-        "description": "Title of the overlay shown when the NetBird background service is too old to drive this UI."
+        "message": "NetBird Client Is Outdated",
+        "description": "Title of the overlay shown when the NetBird client (daemon) is too old to drive this UI."
     },
     "daemon.outdated.description": {
-        "message": "Update the NetBird service to use this app.",
-        "description": "Body of the daemon-outdated overlay telling the user to upgrade the service."
+        "message": "The new GUI isn't compatible with the older NetBird client. Update your client to use the new application.",
+        "description": "Body of the daemon-outdated overlay explaining that the GUI is newer than the client and the client must be updated."
+    },
+    "daemon.outdated.download": {
+        "message": "Download Latest",
+        "description": "Button on the daemon-outdated overlay that opens the download page for the latest release."
     },
     "error.jwt_clock_skew": {
         "message": "Sign-in failed: this device's clock is out of sync with the server. Please sync your system clock and try again.",

client/ui/i18n/locales/es/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Busque NetBird en su bandeja del sistema"
     },
+    "welcome.titleMac": {
+        "message": "Busque NetBird en su barra de menús"
+    },
     "welcome.description": {
         "message": "NetBird reside en su bandeja del sistema. Haga clic en el icono para conectarse, cambiar de perfil o abrir la configuración."
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird reside en su barra de menús. Haga clic en el icono para conectarse, cambiar de perfil o abrir la configuración."
+    },
     "welcome.continue": {
         "message": "Continuar"
     },
@@ -1293,10 +1299,13 @@
         "message": "Documentación"
     },
     "daemon.outdated.title": {
-        "message": "El servicio de NetBird está desactualizado"
+        "message": "NetBird Client está desactualizado"
     },
     "daemon.outdated.description": {
-        "message": "Actualice el servicio de NetBird para usar esta aplicación."
+        "message": "La nueva GUI no es compatible con su cliente anterior. Actualice su cliente para usar la nueva aplicación."
+    },
+    "daemon.outdated.download": {
+        "message": "Descargar la última versión"
     },
     "error.jwt_clock_skew": {
         "message": "Error al iniciar sesión: el reloj de este dispositivo no está sincronizado con el servidor. Sincronice el reloj del sistema e inténtelo de nuevo."

client/ui/i18n/locales/fr/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Cherchez NetBird dans votre barre d’état système"
     },
+    "welcome.titleMac": {
+        "message": "Cherchez NetBird dans votre barre des menus"
+    },
     "welcome.description": {
         "message": "NetBird se trouve dans votre barre d’état système. Cliquez sur l’icône pour vous connecter, changer de profil ou ouvrir les paramètres."
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird se trouve dans votre barre des menus. Cliquez sur l’icône pour vous connecter, changer de profil ou ouvrir les paramètres."
+    },
     "welcome.continue": {
         "message": "Continuer"
     },
@@ -1293,10 +1299,13 @@
         "message": "Documentation"
     },
     "daemon.outdated.title": {
-        "message": "Le service NetBird est obsolète"
+        "message": "Le Client NetBird est obsolète"
     },
     "daemon.outdated.description": {
-        "message": "Mettez à jour le service NetBird pour utiliser cette application."
+        "message": "La nouvelle GUI n'est pas compatible avec votre ancien client. Mettez à jour votre client pour utiliser la nouvelle application."
+    },
+    "daemon.outdated.download": {
+        "message": "Télécharger la dernière version"
     },
     "error.jwt_clock_skew": {
         "message": "Échec de la connexion : l’horloge de cet appareil n’est pas synchronisée avec le serveur. Veuillez synchroniser l’horloge de votre système et réessayer."

client/ui/i18n/locales/hu/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Keresse a NetBirdöt a tálcán"
     },
+    "welcome.titleMac": {
+        "message": "Keresse a NetBirdöt a menüsorban"
+    },
     "welcome.description": {
         "message": "A NetBird a tálcán fut. Kattintson az ikonra a csatlakozáshoz, profilváltáshoz vagy a beállítások megnyitásához."
     },
+    "welcome.descriptionMac": {
+        "message": "A NetBird a menüsorban fut. Kattintson az ikonra a csatlakozáshoz, profilváltáshoz vagy a beállítások megnyitásához."
+    },
     "welcome.continue": {
         "message": "Folytatás"
     },
@@ -1293,10 +1299,13 @@
         "message": "Dokumentáció"
     },
     "daemon.outdated.title": {
-        "message": "A NetBird szolgáltatás elavult"
+        "message": "A NetBird Kliens elavult"
     },
     "daemon.outdated.description": {
-        "message": "Frissítsd a NetBird szolgáltatást az alkalmazás használatához."
+        "message": "Az új GUI nem kompatibilis a régebbi klienseddel. Frissítsd a klienst az új alkalmazás használatához."
+    },
+    "daemon.outdated.download": {
+        "message": "Legújabb letöltése"
     },
     "error.jwt_clock_skew": {
         "message": "A bejelentkezés sikertelen: az eszköz órája eltér a szerverétől. Kérjük, szinkronizálja a rendszer óráját, majd próbálja újra."

client/ui/i18n/locales/it/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Cerchi NetBird nella tray"
     },
+    "welcome.titleMac": {
+        "message": "Cerchi NetBird nella barra dei menu"
+    },
     "welcome.description": {
         "message": "NetBird risiede nella tray. Clicchi sull'icona per connettersi, cambiare profilo o aprire le impostazioni."
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird risiede nella barra dei menu. Clicchi sull'icona per connettersi, cambiare profilo o aprire le impostazioni."
+    },
     "welcome.continue": {
         "message": "Continua"
     },
@@ -1293,10 +1299,13 @@
         "message": "Documentazione"
     },
     "daemon.outdated.title": {
-        "message": "Il servizio NetBird è obsoleto"
+        "message": "NetBird Client è obsoleto"
     },
     "daemon.outdated.description": {
-        "message": "Aggiorna il servizio NetBird per usare questa app."
+        "message": "La nuova GUI non è compatibile con il tuo client precedente. Aggiorna il client per usare la nuova applicazione."
+    },
+    "daemon.outdated.download": {
+        "message": "Scarica l'ultima versione"
     },
     "error.jwt_clock_skew": {
         "message": "Accesso non riuscito: l'orologio di questo dispositivo non è sincronizzato con il server. Sincronizzi l'orologio di sistema e riprovi."

client/ui/i18n/locales/ja/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "トレイの NetBird を確認してください"
     },
+    "welcome.titleMac": {
+        "message": "メニューバーの NetBird を確認してください"
+    },
     "welcome.description": {
         "message": "NetBird はトレイに常駐します。アイコンをクリックして、接続、プロファイルの切り替え、設定を開くことができます。"
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird はメニューバーに常駐します。アイコンをクリックして、接続、プロファイルの切り替え、設定を開くことができます。"
+    },
     "welcome.continue": {
         "message": "続ける"
     },

client/ui/i18n/locales/pt/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Procure o NetBird na sua bandeja"
     },
+    "welcome.titleMac": {
+        "message": "Procure o NetBird na sua barra de menus"
+    },
     "welcome.description": {
         "message": "O NetBird fica na sua bandeja. Clique no ícone para conectar, alternar perfis ou abrir as configurações."
     },
+    "welcome.descriptionMac": {
+        "message": "O NetBird fica na sua barra de menus. Clique no ícone para conectar, alternar perfis ou abrir as configurações."
+    },
     "welcome.continue": {
         "message": "Continuar"
     },
@@ -1293,10 +1299,13 @@
         "message": "Documentação"
     },
     "daemon.outdated.title": {
-        "message": "O serviço NetBird está desatualizado"
+        "message": "O NetBird Client está desatualizado"
     },
     "daemon.outdated.description": {
-        "message": "Atualize o serviço NetBird para usar este aplicativo."
+        "message": "A nova GUI não é compatível com o seu cliente mais antigo. Atualize o seu cliente para usar o novo aplicativo."
+    },
+    "daemon.outdated.download": {
+        "message": "Baixar a versão mais recente"
     },
     "error.jwt_clock_skew": {
         "message": "Falha no login: o relógio deste dispositivo está fora de sincronia com o servidor. Sincronize o relógio do sistema e tente novamente."

client/ui/i18n/locales/ru/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "Найдите NetBird в системном трее"
     },
+    "welcome.titleMac": {
+        "message": "Найдите NetBird в строке меню"
+    },
     "welcome.description": {
         "message": "NetBird находится в системном трее. Нажмите на значок, чтобы подключиться, переключить профиль или открыть настройки."
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird находится в строке меню. Нажмите на значок, чтобы подключиться, переключить профиль или открыть настройки."
+    },
     "welcome.continue": {
         "message": "Продолжить"
     },
@@ -1293,10 +1299,13 @@
         "message": "Документация"
     },
     "daemon.outdated.title": {
-        "message": "Служба NetBird устарела"
+        "message": "Клиент NetBird устарел"
     },
     "daemon.outdated.description": {
-        "message": "Обновите службу NetBird, чтобы использовать это приложение."
+        "message": "Новый GUI несовместим с вашим более старым клиентом. Обновите клиент, чтобы использовать новое приложение."
+    },
+    "daemon.outdated.download": {
+        "message": "Скачать последнюю версию"
     },
     "error.jwt_clock_skew": {
         "message": "Не удалось войти: часы этого устройства рассинхронизированы с сервером. Синхронизируйте системные часы и повторите попытку."

client/ui/i18n/locales/zh-CN/common.json

@@ -1034,9 +1034,15 @@
     "welcome.title": {
         "message": "在托盘中查找 NetBird"
     },
+    "welcome.titleMac": {
+        "message": "在菜单栏中查找 NetBird"
+    },
     "welcome.description": {
         "message": "NetBird 驻留在您的托盘中。点击图标即可连接、切换配置文件或打开设置。"
     },
+    "welcome.descriptionMac": {
+        "message": "NetBird 驻留在您的菜单栏中。点击图标即可连接、切换配置文件或打开设置。"
+    },
     "welcome.continue": {
         "message": "继续"
     },
@@ -1293,10 +1299,13 @@
         "message": "文档"
     },
     "daemon.outdated.title": {
-        "message": "NetBird 服务版本过旧"
+        "message": "NetBird 客户端版本过旧"
     },
     "daemon.outdated.description": {
-        "message": "请更新 NetBird 服务以使用此应用。"
+        "message": "新版 GUI 与您较旧的客户端不兼容。请更新客户端以使用新应用。"
+    },
+    "daemon.outdated.download": {
+        "message": "下载最新版本"
     },
     "error.jwt_clock_skew": {
         "message": "登录失败：此设备的时钟与服务器不同步。请同步您的系统时钟后重试。"

client/ui/services/autostart.go

@@ -10,7 +10,7 @@ import (
 	"github.com/wailsapp/wails/v3/pkg/application"
 )
 
-// Autostart facade over Wails' AutostartManager. The OS login-item registration
+// Autostart facade over Wails' AutostartManager. The OS autostart entry registration
 // is the single source of truth; nothing is mirrored to preferences.
 type Autostart struct {
 	mgr *application.AutostartManager

client/ui/services/profileswitcher.go

@@ -12,13 +12,15 @@ import (
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 )
 
-// ProfileSwitcher holds the reconnect policy shared by the tray and React
-// frontend so both flip profiles identically. The policy keys off prevStatus
-// from DaemonFeed.Get at SwitchActive entry:
+// ProfileSwitcher holds the switch policy shared by the tray and React
+// frontend so both flip profiles identically. SwitchActive (plain selection:
+// header dropdown, tray submenu) always connects after the switch;
+// SwitchActiveNoConnect (manage-profiles screen) never does, so the user can
+// still adjust the management URL before connecting. prevStatus from
+// DaemonFeed.Get at entry only decides the teardown:
 //
-//	Connected/Connecting → Switch + Down + Up; optimistic Connecting paint.
-//	NeedsLogin/LoginFailed/SessionExpired → Switch + Down; clear stale error for re-login.
-//	Idle → Switch only.
+//	Connected/Connecting/NeedsLogin/LoginFailed/SessionExpired → Down first.
+//	Idle → no Down.
 type ProfileSwitcher struct {
 	profiles   *Profiles
 	connection *Connection
@@ -29,29 +31,40 @@ func NewProfileSwitcher(profiles *Profiles, connection *Connection, feed *Daemon
 	return &ProfileSwitcher{profiles: profiles, connection: connection, feed: feed}
 }
 
-// SwitchActive switches to the named profile applying the reconnect policy.
+// SwitchActive switches to the named profile and always connects afterwards.
 func (s *ProfileSwitcher) SwitchActive(ctx context.Context, p ProfileRef) error {
+	return s.switchActive(ctx, p, true)
+}
+
+// SwitchActiveNoConnect switches to the named profile without connecting,
+// tearing down any existing connection first.
+func (s *ProfileSwitcher) SwitchActiveNoConnect(ctx context.Context, p ProfileRef) error {
+	return s.switchActive(ctx, p, false)
+}
+
+func (s *ProfileSwitcher) switchActive(ctx context.Context, p ProfileRef, connect bool) error {
 	prevStatus := ""
-	if st, err := s.feed.Get(ctx); err == nil {
-		prevStatus = st.Status
-	} else {
-		log.Warnf("profileswitcher: get status: %v", err)
+	if s.feed != nil {
+		if st, err := s.feed.Get(ctx); err == nil {
+			prevStatus = st.Status
+		} else {
+			log.Warnf("profileswitcher: get status: %v", err)
+		}
 	}
 
-	wasActive := strings.EqualFold(prevStatus, StatusConnected) ||
-		strings.EqualFold(prevStatus, StatusConnecting)
-	needsDown := wasActive ||
+	needsDown := strings.EqualFold(prevStatus, StatusConnected) ||
+		strings.EqualFold(prevStatus, StatusConnecting) ||
 		strings.EqualFold(prevStatus, StatusNeedsLogin) ||
 		strings.EqualFold(prevStatus, StatusLoginFailed) ||
 		strings.EqualFold(prevStatus, StatusSessionExpired)
 
-	log.Infof("profileswitcher: switch profile=%q prevStatus=%q wasActive=%v needsDown=%v",
-		p.ProfileName, prevStatus, wasActive, needsDown)
+	log.Infof("profileswitcher: switch profile=%q prevStatus=%q connect=%v needsDown=%v",
+		p.ProfileName, prevStatus, connect, needsDown)
 
-	// Optimistic Connecting paint only when wasActive: those prevStatuses emit
-	// stale Connected + transient Idle pushes during Down that must be
-	// suppressed until Up resumes the stream (see DaemonFeed suppression table).
-	if wasActive {
+	// Optimistic Connecting paint plus stale-push suppression during Down (see
+	// DaemonFeed suppression table); also arms the login-watch that pops
+	// browser-login when the new profile turns out to need SSO.
+	if connect && s.feed != nil {
 		s.feed.BeginProfileSwitch()
 	}
 
@@ -76,9 +89,9 @@ func (s *ProfileSwitcher) SwitchActive(ctx context.Context, p ProfileRef) error
 		}
 	}
 
-	if wasActive {
+	if connect {
 		if err := s.connection.Up(ctx, UpParams(p)); err != nil {
-			return fmt.Errorf("reconnect %q: %w", p.ProfileName, err)
+			return fmt.Errorf("connect %q: %w", p.ProfileName, err)
 		}
 	}
 

client/ui/services/windowmanager.go

@@ -185,37 +185,38 @@ func (s *WindowManager) OpenBrowserLogin(uri string) {
 			startURL = "/#/dialog/browser-login?uri=" + url.QueryEscape(uri)
 		}
 		s.hideOtherWindowsLocked("browser-login")
-		// Prefer the main window's screen (multi-monitor); falls back to OS-default centering.
-		var screen *application.Screen
-		if s.mainWindow != nil {
-			if sc, err := s.mainWindow.GetScreen(); err == nil {
-				screen = sc
-			}
-		}
 		opts := DialogWindowOptions("browser-login", s.title("window.title.signIn"), startURL, s.linuxIcon)
 		// Not always-on-top: it would obscure the browser tab the user logs in through.
 		opts.AlwaysOnTop = false
 		opts.InitialPosition = application.WindowCentered
-		opts.Screen = screen
+		// Open on the active (where users cursor is) display, like the session-expiration dialog.
+		opts.Screen = s.getScreenBasedOnCursorPosition()
 		s.browserLogin = s.app.Window.NewWithOptions(opts)
 		bl := s.browserLogin
-		// Red-X close means cancel: emit the event so startLogin() tears down the SSO wait.
 		bl.OnWindowEvent(events.Common.WindowClosing, func(_ *application.WindowEvent) {
-			s.app.Event.Emit(EventBrowserLoginCancel)
 			s.mu.Lock()
-			s.browserLogin = nil
-			s.restoreHiddenWindowsLocked()
+			// Only a live user red-X still has this registered; programmatic closers
+			// nil s.browserLogin first and clean up themselves. Guarding here stops a
+			// stale close event from wiping a replacement popup's state.
+			userClosed := s.browserLogin == bl
+			if userClosed {
+				s.browserLogin = nil
+				s.restoreHiddenWindowsLocked()
+			}
 			s.mu.Unlock()
+			if userClosed {
+				s.app.Event.Emit(EventBrowserLoginCancel)
+			}
 		})
-		s.centerWhenReady(s.browserLogin)
+		s.centerOnCursorScreen(s.browserLogin)
 		return
 	}
 	if uri != "" {
 		s.browserLogin.SetURL("/#/dialog/browser-login?uri=" + url.QueryEscape(uri))
 	}
+	s.centerOnCursorScreen(s.browserLogin)
 	s.browserLogin.Show()
 	s.browserLogin.Focus()
-	s.centerWhenReady(s.browserLogin)
 }
 
 // BrowserLoginWindow returns the live SSO popup, or nil. While non-nil it is the
@@ -238,6 +239,15 @@ func (s *WindowManager) CloseBrowserLogin() {
 	s.mu.Lock()
 	w := s.browserLogin
 	s.browserLogin = nil
+	// The WindowClosing hook no-ops on a programmatic close, so restore here —
+	// but only if a popup was actually open. The frontend calls this even when no
+	// popup was ever shown (e.g. resetDialog() after an early RequestExtend failure,
+	// or connection.ts's catch path), and hiddenForLogin is shared with
+	// OpenInstallProgress, so an unconditional restore could re-show windows a
+	// still-running install-progress is hiding.
+	if w != nil {
+		s.restoreHiddenWindowsLocked()
+	}
 	s.mu.Unlock()
 	if w != nil {
 		w.Close()
@@ -279,6 +289,35 @@ func (s *WindowManager) CloseSessionExpiration() {
 	}
 }
 
+// CloseRenewFlow tears down the SSO session-renewal UI in a single call: it
+// closes the browser-login popup and the session-expiration window together.
+func (s *WindowManager) CloseRenewFlow() {
+	s.mu.Lock()
+	bl := s.browserLogin
+	se := s.sessionExpiration
+	s.browserLogin = nil
+	s.sessionExpiration = nil
+	if se != nil {
+		kept := s.hiddenForLogin[:0]
+		for _, w := range s.hiddenForLogin {
+			if w != se {
+				kept = append(kept, w)
+			}
+		}
+		s.hiddenForLogin = kept
+	}
+	s.restoreHiddenWindowsLocked()
+	s.mu.Unlock()
+
+	// Close after unlock so the re-entrant handlers can take s.mu.
+	if bl != nil {
+		bl.Close()
+	}
+	if se != nil {
+		se.Close()
+	}
+}
+
 // OpenInstallProgress shows the install-progress window and hides the rest for the duration
 // (restored on close). It owns its own result polling since the daemon restarts mid-install.
 func (s *WindowManager) OpenInstallProgress(version string) {

client/ui/tray.go

@@ -30,6 +30,8 @@ const (
 
 	statusError = "Error"
 
+	quitDownTimeout = 5 * time.Second
+
 	urlGitHubRepo     = "https://github.com/netbirdio/netbird"
 	urlGitHubReleases = "https://github.com/netbirdio/netbird/releases/latest"
 	urlDocs           = "https://docs.netbird.io"
@@ -315,8 +317,7 @@ func (t *Tray) relayoutMenu() {
 		if sessionDeadline.IsZero() {
 			t.sessionExpiresItem.SetHidden(true)
 		} else {
-			remaining := t.formatSessionRemaining(time.Until(sessionDeadline))
-			t.sessionExpiresItem.SetLabel(t.loc.T("tray.session.expiresIn", "remaining", remaining))
+			t.sessionExpiresItem.SetLabel(t.sessionRowLabel(sessionDeadline))
 			t.sessionExpiresItem.SetHidden(false)
 		}
 	}
@@ -446,11 +447,28 @@ func (t *Tray) buildMenu() *application.Menu {
 	menu.AddSeparator()
 	menu.Add(t.loc.T("tray.menu.quit")).
 		SetAccelerator("CmdOrCtrl+Q").
-		OnClick(func(*application.Context) { t.app.Quit() })
+		OnClick(func(*application.Context) { t.handleQuit() })
 
 	return menu
 }
 
+func (t *Tray) handleQuit() {
+	t.profileMu.Lock()
+	if t.switchCancel != nil {
+		t.switchCancel()
+		t.switchCancel = nil
+	}
+	t.profileMu.Unlock()
+	t.svc.DaemonFeed.CancelProfileSwitch()
+
+	ctx, cancel := context.WithTimeout(context.Background(), quitDownTimeout)
+	defer cancel()
+	if err := t.svc.Connection.Down(ctx); err != nil {
+		log.Errorf("disconnect on quit: %v", err)
+	}
+	t.app.Quit()
+}
+
 // handleConnect receives the clicked item from the buildMenu closure —
 // t.upItem is menuMu-guarded and must not be read here.
 func (t *Tray) handleConnect(upItem *application.MenuItem) {

client/ui/tray_session.go

@@ -64,11 +64,42 @@ func (t *Tray) applySessionExpiry(deadline *time.Time, connected bool) bool {
 	return changed
 }
 
-// runSessionExpiryTicker recomputes the "Expires in …" row label every 30s. Runs until process exit.
+// runSessionExpiryTicker recomputes the "Expires in …" row label until process exit.
+// The interval scales with the remaining time: coarse when the deadline is far off,
+// down to 10s in the final two minutes so the label doesn't lag the ceiling-rounded
+// countdown near expiry. The cached deadline is re-read every iteration, so an extend
+// or reconnect that moves it is picked up on the next tick.
 func (t *Tray) runSessionExpiryTicker() {
-	tk := time.NewTicker(30 * time.Second)
-	for range tk.C {
+	tm := time.NewTimer(sessionRefreshInterval(t.sessionRemaining()))
+	defer tm.Stop()
+	for range tm.C {
 		t.refreshSessionExpiresLabel()
+		tm.Reset(sessionRefreshInterval(t.sessionRemaining()))
+	}
+}
+
+// sessionRemaining returns the time left on the cached SSO deadline, or 0 when unknown.
+func (t *Tray) sessionRemaining() time.Duration {
+	t.sessionMu.Lock()
+	deadline := t.sessionExpiresAt
+	t.sessionMu.Unlock()
+	if deadline.IsZero() {
+		return 0
+	}
+	return time.Until(deadline)
+}
+
+// sessionRefreshInterval picks how long to wait before the next label recompute.
+func sessionRefreshInterval(remaining time.Duration) time.Duration {
+	switch {
+	case remaining <= 0:
+		return 30 * time.Second
+	case remaining <= 2*time.Minute:
+		return 10 * time.Second
+	case remaining <= time.Hour:
+		return 30 * time.Second
+	default:
+		return time.Minute
 	}
 }
 
@@ -87,30 +118,39 @@ func (t *Tray) refreshSessionExpiresLabel() {
 	if deadline.IsZero() {
 		return
 	}
-	remaining := t.formatSessionRemaining(time.Until(deadline))
-	item.SetLabel(t.loc.T("tray.session.expiresIn", "remaining", remaining))
+	item.SetLabel(t.sessionRowLabel(deadline))
+}
+
+func (t *Tray) sessionRowLabel(deadline time.Time) string {
+	remaining := time.Until(deadline)
+	if remaining <= 0 {
+		return t.loc.T("tray.status.sessionExpired")
+	}
+	return t.loc.T("tray.session.expiresIn", "remaining", t.formatSessionRemaining(remaining))
 }
 
 // formatSessionRemaining renders d as a localised long-form string picking the largest non-zero unit.
+// Each unit is rounded up so the label never claims less time than actually remains, matching the
+// upper-bound sense of the sub-minute "less than a minute" fragment.
 // Singular/plural keys are split per language for proper translation.
 func (t *Tray) formatSessionRemaining(d time.Duration) string {
 	switch {
 	case d < time.Minute:
 		return t.loc.T("tray.session.unit.lessThanMinute")
-	case d < time.Hour:
-		m := int(d / time.Minute)
+	case d <= 59*time.Minute:
+		m := ceilDiv(d, time.Minute)
 		if m == 1 {
 			return t.loc.T("tray.session.unit.minute")
 		}
 		return t.loc.T("tray.session.unit.minutes", "count", strconv.Itoa(m))
-	case d < 24*time.Hour:
-		h := int((d + 30*time.Minute) / time.Hour)
+	case d <= 23*time.Hour:
+		h := ceilDiv(d, time.Hour)
 		if h == 1 {
 			return t.loc.T("tray.session.unit.hour")
 		}
 		return t.loc.T("tray.session.unit.hours", "count", strconv.Itoa(h))
 	default:
-		days := int((d + 12*time.Hour) / (24 * time.Hour))
+		days := ceilDiv(d, 24*time.Hour)
 		if days == 1 {
 			return t.loc.T("tray.session.unit.day")
 		}
@@ -118,6 +158,11 @@ func (t *Tray) formatSessionRemaining(d time.Duration) string {
 	}
 }
 
+// ceilDiv divides d by unit rounding up, assuming d > 0.
+func ceilDiv(d, unit time.Duration) int {
+	return int((d + unit - time.Nanosecond) / unit)
+}
+
 // registerSessionWarningCategory wires the OS notification category and response handler for the expiry warning.
 // Errors are swallowed since the worst case is a plain notification without buttons.
 func (t *Tray) registerSessionWarningCategory() {
@@ -252,11 +297,9 @@ func (t *Tray) openSessionExpiration() {
 }
 
 // openSessionExtendFlow opens the SessionExpiration window seeded with the cached deadline's remaining time,
-// for the "Expires in …" tray row. No-ops when the deadline is unknown or elapsed.
+// for the "Expires in …" tray row. Once the deadline has elapsed the row reads "Session expired" and the
+// click routes to the login flow instead. No-op when the deadline is unknown.
 func (t *Tray) openSessionExtendFlow() {
-	if t.svc.WindowManager == nil {
-		return
-	}
 	t.sessionMu.Lock()
 	deadline := t.sessionExpiresAt
 	t.sessionMu.Unlock()
@@ -265,6 +308,14 @@ func (t *Tray) openSessionExtendFlow() {
 	}
 	seconds := int(time.Until(deadline).Seconds())
 	if seconds <= 0 {
+		if t.window != nil {
+			t.window.SetURL("/#/login")
+			t.window.Show()
+			t.window.Focus()
+		}
+		return
+	}
+	if t.svc.WindowManager == nil {
 		return
 	}
 	t.svc.WindowManager.OpenSessionExpiration(seconds)

combined/cmd/admin.go

@@ -0,0 +1,151 @@
+package cmd
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/dexidp/dex/storage"
+	log "github.com/sirupsen/logrus"
+	"github.com/spf13/cobra"
+
+	"github.com/netbirdio/netbird/formatter/hook"
+	admincmd "github.com/netbirdio/netbird/management/cmd/admin"
+	tokencmd "github.com/netbirdio/netbird/management/cmd/token"
+	nbconfig "github.com/netbirdio/netbird/management/internals/server/config"
+	"github.com/netbirdio/netbird/management/server/activity"
+	activitystore "github.com/netbirdio/netbird/management/server/activity/store"
+	"github.com/netbirdio/netbird/management/server/store"
+	"github.com/netbirdio/netbird/management/server/types"
+	"github.com/netbirdio/netbird/util"
+)
+
+// newAdminCommands creates the admin command tree with combined-specific resource openers.
+func newAdminCommands() *cobra.Command {
+	return admincmd.NewCommands(admincmd.Openers{
+		Resources: withAdminResources,
+		Store:     withAdminStoreOnly,
+		IDP:       withAdminIDPOnly,
+	})
+}
+
+func newLegacyTokenCommand() *cobra.Command {
+	cmd := tokencmd.NewCommands(tokencmd.StoreOpener(withAdminStoreOnly))
+	cmd.Deprecated = "use 'admin token' instead"
+	return cmd
+}
+
+// withAdminResources loads the combined YAML config, initializes stores, and calls fn.
+func withAdminResources(cmd *cobra.Command, fn func(ctx context.Context, resources admincmd.Resources) error) error {
+	return withAdminConfig(cmd, func(ctx context.Context, cfg *CombinedConfig) error {
+		mgmtConfig, err := adminManagementConfig(cfg)
+		if err != nil {
+			return err
+		}
+
+		managementStore, err := openAdminStore(ctx, cfg)
+		if err != nil {
+			return err
+		}
+		defer admincmd.CloseStore(ctx, managementStore)
+
+		idpStorage, idpStorageFile, err := admincmd.OpenIDPStorage(mgmtConfig)
+		if err != nil {
+			return err
+		}
+		defer admincmd.CloseIDPStorage(idpStorage)
+
+		eventStore, esErr := openAdminEventStore(ctx, cfg, mgmtConfig)
+		if esErr != nil {
+			_, _ = fmt.Fprintf(cmd.ErrOrStderr(), "Warning: audit events will not be recorded: %v\n", esErr)
+		}
+		if eventStore != nil {
+			defer func() {
+				if err := eventStore.Close(ctx); err != nil {
+					log.Debugf("close activity event store: %v", err)
+				}
+			}()
+		}
+
+		return fn(ctx, admincmd.Resources{Store: managementStore, IDPStorage: idpStorage, IDPStorageFile: idpStorageFile, EventStore: eventStore})
+	})
+}
+
+// withAdminStoreOnly opens only the management store for admin subcommands that do not
+// need embedded IdP storage.
+func withAdminStoreOnly(cmd *cobra.Command, fn func(ctx context.Context, s store.Store) error) error {
+	return withAdminConfig(cmd, func(ctx context.Context, cfg *CombinedConfig) error {
+		managementStore, err := openAdminStore(ctx, cfg)
+		if err != nil {
+			return err
+		}
+		defer admincmd.CloseStore(ctx, managementStore)
+
+		return fn(ctx, managementStore)
+	})
+}
+
+func withAdminIDPOnly(cmd *cobra.Command, fn func(ctx context.Context, idpStorage storage.Storage, storageFile string) error) error {
+	return withAdminConfig(cmd, func(ctx context.Context, cfg *CombinedConfig) error {
+		mgmtConfig, err := adminManagementConfig(cfg)
+		if err != nil {
+			return err
+		}
+		idpStorage, idpStorageFile, err := admincmd.OpenIDPStorage(mgmtConfig)
+		if err != nil {
+			return err
+		}
+		defer admincmd.CloseIDPStorage(idpStorage)
+
+		return fn(ctx, idpStorage, idpStorageFile)
+	})
+}
+
+func withAdminConfig(cmd *cobra.Command, fn func(ctx context.Context, cfg *CombinedConfig) error) error {
+	if err := util.InitLog("error", "console"); err != nil {
+		return fmt.Errorf("init log: %w", err)
+	}
+
+	ctx := context.WithValue(cmd.Context(), hook.ExecutionContextKey, hook.SystemSource) //nolint:staticcheck
+
+	cfg, err := LoadConfig(configPath)
+	if err != nil {
+		return fmt.Errorf("load config: %w", err)
+	}
+	cfg.ApplyAdminDefaults()
+	applyServerStoreEnv(cfg.Server.Store)
+
+	return fn(ctx, cfg)
+}
+
+func adminManagementConfig(cfg *CombinedConfig) (*nbconfig.Config, error) {
+	mgmtConfig, err := cfg.ToManagementConfig()
+	if err != nil {
+		return nil, fmt.Errorf("create management config: %w", err)
+	}
+	return mgmtConfig, nil
+}
+
+func openAdminStore(ctx context.Context, cfg *CombinedConfig) (store.Store, error) {
+	managementStore, err := store.NewStore(ctx, types.Engine(cfg.Management.Store.Engine), cfg.Management.DataDir, nil, true)
+	if err != nil {
+		return nil, fmt.Errorf("create store: %w", err)
+	}
+	return managementStore, nil
+}
+
+func openAdminEventStore(ctx context.Context, cfg *CombinedConfig, config *nbconfig.Config) (activity.Store, error) {
+	if config.DataStoreEncryptionKey == "" {
+		return nil, fmt.Errorf("data store encryption key is not configured")
+	}
+	if err := applyActivityStoreEnv(cfg.Server.ActivityStore); err != nil {
+		return nil, fmt.Errorf("configure activity event store: %w", err)
+	}
+	eventStore, err := activitystore.NewSqlStore(ctx, config.Datadir, config.DataStoreEncryptionKey)
+	if err != nil {
+		return nil, fmt.Errorf("open activity event store: %w", err)
+	}
+	if eventStore == nil {
+		return nil, fmt.Errorf("open activity event store: returned nil store")
+	}
+	return eventStore, nil
+}

combined/cmd/admin_config_test.go

@@ -0,0 +1,47 @@
+package cmd
+
+import (
+	"context"
+	"os"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	nbconfig "github.com/netbirdio/netbird/management/internals/server/config"
+)
+
+func TestApplyAdminDefaultsCopiesServerStoreWithoutExposedAddress(t *testing.T) {
+	cfg := DefaultConfig()
+	cfg.Server.ExposedAddress = ""
+	cfg.Server.DataDir = "/srv/netbird"
+	cfg.Server.Store = StoreConfig{
+		Engine: "postgres",
+		DSN:    "postgres://user:pass@example.com/netbird",
+	}
+
+	cfg.ApplyAdminDefaults()
+
+	require.Equal(t, "/srv/netbird", cfg.Management.DataDir)
+	require.Equal(t, "postgres", cfg.Management.Store.Engine)
+	require.Equal(t, cfg.Server.Store.DSN, cfg.Management.Store.DSN)
+}
+
+func TestOpenAdminEventStoreMissingEncryptionKeyReturnsNilInterface(t *testing.T) {
+	eventStore, err := openAdminEventStore(context.Background(), &CombinedConfig{}, &nbconfig.Config{})
+	require.Error(t, err)
+	require.Contains(t, err.Error(), "encryption key")
+	require.Nil(t, eventStore)
+}
+
+func TestApplyServerStoreEnv(t *testing.T) {
+	t.Setenv("NB_STORE_ENGINE_POSTGRES_DSN", "")
+	t.Setenv("NB_STORE_ENGINE_MYSQL_DSN", "")
+	t.Setenv("NB_STORE_ENGINE_SQLITE_FILE", "")
+
+	applyServerStoreEnv(StoreConfig{Engine: "postgres", DSN: "postgres-dsn", File: "store.db"})
+	require.Equal(t, "postgres-dsn", os.Getenv("NB_STORE_ENGINE_POSTGRES_DSN"))
+	require.Equal(t, "store.db", os.Getenv("NB_STORE_ENGINE_SQLITE_FILE"))
+
+	applyServerStoreEnv(StoreConfig{Engine: "mysql", DSN: "mysql-dsn"})
+	require.Equal(t, "mysql-dsn", os.Getenv("NB_STORE_ENGINE_MYSQL_DSN"))
+}

combined/cmd/config.go

@@ -6,8 +6,7 @@ import (
 	"net"
 	"net/netip"
 	"os"
-	"path"
-	"path/filepath"
+	filePath "path/filepath"
 	"strings"
 	"time"
 
@@ -74,6 +73,9 @@ type ServerConfig struct {
 	ActivityStore           StoreConfig        `yaml:"activityStore"`
 	AuthStore               StoreConfig        `yaml:"authStore"`
 	ReverseProxy            ReverseProxyConfig `yaml:"reverseProxy"`
+
+	SupportedSyncMessageVersions           *int           `yaml:"supportedSyncMessageVersions,omitempty"`
+	PerAccountSupportedSyncMessageVersions map[string]int `yaml:"perAccountSupportedSyncMessageVersions,omitempty"`
 }
 
 // TLSConfig contains TLS/HTTPS settings
@@ -300,6 +302,19 @@ func (c *CombinedConfig) ApplySimplifiedDefaults() {
 	c.autoConfigureClientSettings(exposedProto, exposedHost, exposedHostPort, hasExternalStuns, hasExternalRelay, hasExternalSignal)
 }
 
+// ApplyAdminDefaults applies the management settings needed by admin commands even
+// when the full server config is invalid and ApplySimplifiedDefaults cannot run.
+func (c *CombinedConfig) ApplyAdminDefaults() {
+	if c.Management.DataDir == "" || c.Management.DataDir == "/var/lib/netbird/" {
+		c.Management.DataDir = c.Server.DataDir
+	}
+	if c.Management.Store.Engine == "" || c.Management.Store.Engine == "sqlite" {
+		if c.Server.Store.Engine != "" || c.Server.Store.File != "" || c.Server.Store.DSN != "" {
+			c.Management.Store = c.Server.Store
+		}
+	}
+}
+
 // applyRelayDefaults configures the relay service if no external relay is configured.
 func (c *CombinedConfig) applyRelayDefaults(exposedProto, exposedHostPort string, hasExternalRelay, hasExternalStuns bool) {
 	if hasExternalRelay {
@@ -577,11 +592,11 @@ func (c *CombinedConfig) buildEmbeddedIdPConfig(mgmt ManagementConfig) (*idp.Emb
 			return nil, fmt.Errorf("authStore.dsn is required when authStore.engine is postgres")
 		}
 	} else {
-		authStorageFile = path.Join(mgmt.DataDir, "idp.db")
+		authStorageFile = filePath.Join(mgmt.DataDir, "idp.db")
 		if c.Server.AuthStore.File != "" {
 			authStorageFile = c.Server.AuthStore.File
-			if !filepath.IsAbs(authStorageFile) {
-				authStorageFile = filepath.Join(mgmt.DataDir, authStorageFile)
+			if !filePath.IsAbs(authStorageFile) {
+				authStorageFile = filePath.Join(mgmt.DataDir, authStorageFile)
 			}
 		}
 	}
@@ -696,16 +711,18 @@ func (c *CombinedConfig) ToManagementConfig() (*nbconfig.Config, error) {
 	httpConfig.AuthCallbackURL = callbackURL + types.ProxyCallbackEndpointFull
 
 	return &nbconfig.Config{
-		Stuns:                  stuns,
-		Relay:                  relayConfig,
-		Signal:                 signalConfig,
-		Datadir:                mgmt.DataDir,
-		DataStoreEncryptionKey: mgmt.Store.EncryptionKey,
-		HttpConfig:             httpConfig,
-		StoreConfig:            storeConfig,
-		ReverseProxy:           reverseProxy,
-		DisableDefaultPolicy:   mgmt.DisableDefaultPolicy,
-		EmbeddedIdP:            embeddedIdP,
+		Stuns:                              stuns,
+		Relay:                              relayConfig,
+		Signal:                             signalConfig,
+		Datadir:                            mgmt.DataDir,
+		DataStoreEncryptionKey:             mgmt.Store.EncryptionKey,
+		HttpConfig:                         httpConfig,
+		StoreConfig:                        storeConfig,
+		ReverseProxy:                       reverseProxy,
+		DisableDefaultPolicy:               mgmt.DisableDefaultPolicy,
+		EmbeddedIdP:                        embeddedIdP,
+		HighestSupportedSyncMessageVersion: c.Server.SupportedSyncMessageVersions,
+		PerAccountHighestSupportedSyncMessageVersion: c.Server.PerAccountSupportedSyncMessageVersions,
 	}, nil
 }
 
@@ -729,7 +746,7 @@ func ApplyEmbeddedIdPConfig(ctx context.Context, cfg *nbconfig.Config, mgmtPort
 		cfg.EmbeddedIdP.Storage.Type = "sqlite3"
 	}
 	if cfg.EmbeddedIdP.Storage.Config.File == "" && cfg.Datadir != "" {
-		cfg.EmbeddedIdP.Storage.Config.File = path.Join(cfg.Datadir, "idp.db")
+		cfg.EmbeddedIdP.Storage.Config.File = filePath.Join(cfg.Datadir, "idp.db")
 	}
 
 	issuer := cfg.EmbeddedIdP.Issuer

combined/cmd/root.go

@@ -31,6 +31,7 @@ import (
 	relayServer "github.com/netbirdio/netbird/relay/server"
 	"github.com/netbirdio/netbird/relay/server/listener"
 	"github.com/netbirdio/netbird/relay/server/listener/ws"
+	syncgrpc "github.com/netbirdio/netbird/shared/management/grpc"
 	sharedMetrics "github.com/netbirdio/netbird/shared/metrics"
 	"github.com/netbirdio/netbird/shared/relay/auth"
 	"github.com/netbirdio/netbird/shared/signal/proto"
@@ -64,7 +65,8 @@ func init() {
 	rootCmd.PersistentFlags().StringVarP(&configPath, "config", "c", "", "path to YAML configuration file (required)")
 	_ = rootCmd.MarkPersistentFlagRequired("config")
 
-	rootCmd.AddCommand(newTokenCommands())
+	rootCmd.AddCommand(newAdminCommands())
+	rootCmd.AddCommand(newLegacyTokenCommand())
 }
 
 func RootCmd() *cobra.Command {
@@ -122,6 +124,37 @@ func execute(cmd *cobra.Command, _ []string) error {
 }
 
 // initializeConfig loads and validates the configuration, then initializes logging.
+func applyServerStoreEnv(storeConfig StoreConfig) {
+	if dsn := storeConfig.DSN; dsn != "" {
+		switch strings.ToLower(storeConfig.Engine) {
+		case "postgres":
+			os.Setenv("NB_STORE_ENGINE_POSTGRES_DSN", dsn)
+		case "mysql":
+			os.Setenv("NB_STORE_ENGINE_MYSQL_DSN", dsn)
+		}
+	}
+	if file := storeConfig.File; file != "" {
+		os.Setenv("NB_STORE_ENGINE_SQLITE_FILE", file)
+	}
+}
+
+func applyActivityStoreEnv(storeConfig StoreConfig) error {
+	if engine := storeConfig.Engine; engine != "" {
+		engineLower := strings.ToLower(engine)
+		if engineLower == "postgres" && storeConfig.DSN == "" {
+			return fmt.Errorf("activityStore.dsn is required when activityStore.engine is postgres")
+		}
+		os.Setenv("NB_ACTIVITY_EVENT_STORE_ENGINE", engineLower)
+		if dsn := storeConfig.DSN; dsn != "" {
+			os.Setenv("NB_ACTIVITY_EVENT_POSTGRES_DSN", dsn)
+		}
+	}
+	if file := storeConfig.File; file != "" {
+		os.Setenv("NB_ACTIVITY_EVENT_SQLITE_FILE", file)
+	}
+	return nil
+}
+
 func initializeConfig() error {
 	var err error
 	config, err = LoadConfig(configPath)
@@ -137,30 +170,10 @@ func initializeConfig() error {
 		return fmt.Errorf("failed to initialize log: %w", err)
 	}
 
-	if dsn := config.Server.Store.DSN; dsn != "" {
-		switch strings.ToLower(config.Server.Store.Engine) {
-		case "postgres":
-			os.Setenv("NB_STORE_ENGINE_POSTGRES_DSN", dsn)
-		case "mysql":
-			os.Setenv("NB_STORE_ENGINE_MYSQL_DSN", dsn)
-		}
-	}
-	if file := config.Server.Store.File; file != "" {
-		os.Setenv("NB_STORE_ENGINE_SQLITE_FILE", file)
-	}
+	applyServerStoreEnv(config.Server.Store)
 
-	if engine := config.Server.ActivityStore.Engine; engine != "" {
-		engineLower := strings.ToLower(engine)
-		if engineLower == "postgres" && config.Server.ActivityStore.DSN == "" {
-			return fmt.Errorf("activityStore.dsn is required when activityStore.engine is postgres")
-		}
-		os.Setenv("NB_ACTIVITY_EVENT_STORE_ENGINE", engineLower)
-		if dsn := config.Server.ActivityStore.DSN; dsn != "" {
-			os.Setenv("NB_ACTIVITY_EVENT_POSTGRES_DSN", dsn)
-		}
-	}
-	if file := config.Server.ActivityStore.File; file != "" {
-		os.Setenv("NB_ACTIVITY_EVENT_SQLITE_FILE", file)
+	if err := applyActivityStoreEnv(config.Server.ActivityStore); err != nil {
+		return err
 	}
 
 	log.Infof("Starting combined NetBird server")
@@ -505,6 +518,16 @@ func createManagementServer(cfg *CombinedConfig, mgmtConfig *nbconfig.Config) (m
 	}
 	mgmtPort, _ := strconv.Atoi(portStr)
 
+	if err := syncgrpc.ValidateSyncMessageVersion(mgmtConfig.HighestSupportedSyncMessageVersion); err != nil {
+		return nil, err
+	}
+
+	for accountId, version := range mgmtConfig.PerAccountHighestSupportedSyncMessageVersion {
+		if err := syncgrpc.ValidateSyncMessageVersion(&version); err != nil {
+			return nil, fmt.Errorf("unrecognized sync message version in perAccountSupportedSyncMessageVersions for account %s %w", accountId, err)
+		}
+	}
+
 	mgmtSrv := newServer(
 		&mgmtServer.Config{
 			NbConfig:                mgmtConfig,

combined/cmd/token.go

@@ -1,63 +0,0 @@
-package cmd
-
-import (
-	"context"
-	"fmt"
-	"os"
-	"strings"
-
-	log "github.com/sirupsen/logrus"
-	"github.com/spf13/cobra"
-
-	"github.com/netbirdio/netbird/formatter/hook"
-	tokencmd "github.com/netbirdio/netbird/management/cmd/token"
-	"github.com/netbirdio/netbird/management/server/store"
-	"github.com/netbirdio/netbird/management/server/types"
-	"github.com/netbirdio/netbird/util"
-)
-
-// newTokenCommands creates the token command tree with combined-specific store opener.
-func newTokenCommands() *cobra.Command {
-	return tokencmd.NewCommands(withTokenStore)
-}
-
-// withTokenStore loads the combined YAML config, initializes the store, and calls fn.
-func withTokenStore(cmd *cobra.Command, fn func(ctx context.Context, s store.Store) error) error {
-	if err := util.InitLog("error", "console"); err != nil {
-		return fmt.Errorf("init log: %w", err)
-	}
-
-	ctx := context.WithValue(cmd.Context(), hook.ExecutionContextKey, hook.SystemSource) //nolint:staticcheck
-
-	cfg, err := LoadConfig(configPath)
-	if err != nil {
-		return fmt.Errorf("load config: %w", err)
-	}
-
-	if dsn := cfg.Server.Store.DSN; dsn != "" {
-		switch strings.ToLower(cfg.Server.Store.Engine) {
-		case "postgres":
-			os.Setenv("NB_STORE_ENGINE_POSTGRES_DSN", dsn)
-		case "mysql":
-			os.Setenv("NB_STORE_ENGINE_MYSQL_DSN", dsn)
-		}
-	}
-	if file := cfg.Server.Store.File; file != "" {
-		os.Setenv("NB_STORE_ENGINE_SQLITE_FILE", file)
-	}
-
-	datadir := cfg.Management.DataDir
-	engine := types.Engine(cfg.Management.Store.Engine)
-
-	s, err := store.NewStore(ctx, engine, datadir, nil, true)
-	if err != nil {
-		return fmt.Errorf("create store: %w", err)
-	}
-	defer func() {
-		if err := s.Close(ctx); err != nil {
-			log.Debugf("close store: %v", err)
-		}
-	}()
-
-	return fn(ctx, s)
-}