use API

(We now do AsyncStart + waitEstablishedOrDone, so everything is managed inside the
supervisor)

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

@@ -1,68 +0,0 @@
-name: Agent Network E2E
-
-on:
-  # Nightly at 03:00 UTC, plus on demand from the Actions tab.
-  schedule:
-    - cron: "0 3 * * *"
-  workflow_dispatch:
-
-concurrency:
-  group: ${{ github.workflow }}-${{ github.ref }}
-  cancel-in-progress: true
-
-jobs:
-  e2e:
-    name: Agent Network E2E
-    runs-on: ubuntu-latest
-    timeout-minutes: 45
-    steps:
-      - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
-        with:
-          persist-credentials: false
-
-      - name: Install Go
-        uses: actions/setup-go@4a3601121dd01d1626a1e23e37211e3254c1c06c # v6.4.0
-        with:
-          go-version-file: "go.mod"
-
-      # Container-driver builder so the harness can build the combined/proxy/
-      # client images from source with a local layer cache.
-      - name: Set up Buildx
-        uses: docker/setup-buildx-action@d7f5e7f509e45cec5c76c4d5afdd7de93d0b3df5 # v4.1.0
-
-      # Persist the Docker layer cache across runs. This caches the base, apt,
-      # and go-mod-download layers; the Go compile still re-runs, as BuildKit
-      # mount caches cannot be exported to the GitHub cache.
-      - name: Cache Docker layers
-        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
-        with:
-          path: /tmp/.buildx-cache
-          key: ${{ runner.os }}-anet-e2e-buildx-${{ hashFiles('go.sum', 'combined/Dockerfile.multistage', 'proxy/Dockerfile.multistage', 'e2e/harness/Dockerfile.client') }}
-          restore-keys: |
-            ${{ runner.os }}-anet-e2e-buildx-
-
-      - name: Run agent-network e2e
-        env:
-          # Build the images from source (this branch's code) with the shared
-          # local layer cache.
-          NB_E2E_BUILDX_CACHE: /tmp/.buildx-cache
-          # Provider credentials. Each provider scenario skips if its
-          # token (and URL, for gateways) is unset, so partial coverage is fine.
-          OPENAI_TOKEN: ${{ secrets.E2E_OPENAI_TOKEN }}
-          ANTHROPIC_TOKEN: ${{ secrets.E2E_ANTHROPIC_TOKEN }}
-          VERCEL_URL: ${{ secrets.E2E_VERCEL_URL }}
-          VERCEL_TOKEN: ${{ secrets.E2E_VERCEL_TOKEN }}
-          OPENROUTER_URL: ${{ secrets.E2E_OPENROUTER_URL }}
-          OPENROUTER_TOKEN: ${{ secrets.E2E_OPENROUTER_TOKEN }}
-          CLOUDFLARE_URL: ${{ secrets.E2E_CLOUDFLARE_URL }}
-          CLOUDFLARE_TOKEN: ${{ secrets.E2E_CLOUDFLARE_TOKEN }}
-          AWS_BEARER_TOKEN_BEDROCK: ${{ secrets.E2E_AWS_BEARER_TOKEN_BEDROCK }}
-          AWS_REGION: ${{ secrets.E2E_AWS_REGION }}
-          # Vertex (Anthropic-on-Vertex): SA + project required; region defaults
-          # to "global", model to a pinned claude snapshot.
-          GOOGLE_VERTEX_SA_BASE64: ${{ secrets.E2E_GOOGLE_VERTEX_SA_BASE64 }}
-          GOOGLE_VERTEX_PROJECT: ${{ secrets.E2E_GOOGLE_VERTEX_PROJECT }}
-          GOOGLE_VERTEX_REGION: ${{ secrets.E2E_GOOGLE_VERTEX_REGION }}
-          GOOGLE_VERTEX_MODEL: ${{ secrets.E2E_GOOGLE_VERTEX_MODEL }}
-        run: go test -tags e2e -timeout 40m -v ./e2e/...

.github/workflows/check-license-dependencies.yml

@@ -20,7 +20,7 @@ jobs:
 
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
@@ -59,12 +59,12 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-      - uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Set up Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: true

.github/workflows/git-town.yml

@@ -15,7 +15,7 @@ jobs:
       pull-requests: write
 
     steps:
-      - uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - uses: git-town/action@3d8b878379abb1ee393fb49865a28b4a6c2cd3b0 # v1.2.1

.github/workflows/golang-test-darwin.yml

@@ -16,18 +16,18 @@ jobs:
     runs-on: macos-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
 
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: ~/go/pkg/mod
           key: macos-gotest-${{ hashFiles('**/go.sum') }}
@@ -45,10 +45,10 @@ jobs:
         run: git --no-pager diff --exit-code
 
       - name: Test
-        run: NETBIRD_STORE_ENGINE=${{ matrix.store }} CI=true go test -coverprofile=coverage.txt -tags 'devcert privileged' -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE' -timeout 5m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/testutil/privileged)
+        run: NETBIRD_STORE_ENGINE=${{ matrix.store }} CI=true go test -coverprofile=coverage.txt -tags=devcert -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE' -timeout 5m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined)
 
       - name: Upload coverage reports to Codecov
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird

.github/workflows/golang-test-freebsd.yml

@@ -16,7 +16,7 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
@@ -28,7 +28,7 @@ jobs:
         id: test
         env:
           GO_VERSION: ${{ steps.goversion.outputs.version }}
-        uses: vmactions/freebsd-vm@b84ab5559b5a1bb4b8ee2737d2506a16e1737636 # v1.4.8
+        uses: vmactions/freebsd-vm@d1e65811565151536c0c894fff74f06351ed26e6 # v1.4.5
         with:
           usesh: true
           copyback: false
@@ -48,14 +48,14 @@ jobs:
             export PATH=$PATH:/usr/local/go/bin:$HOME/go/bin
             time go build -o netbird client/main.go
             # check all component except management, since we do not support management server on freebsd
-            time go test -tags privileged -timeout 1m -failfast ./base62/...
+            time go test -timeout 1m -failfast ./base62/...
             # NOTE: without -p1 `client/internal/dns` will fail because of `listen udp4 :33100: bind: address already in use`
-            time go test -tags privileged -timeout 8m -failfast -v -p 1 ./client/...
-            time go test -tags privileged -timeout 1m -failfast ./dns/...
-            time go test -tags privileged -timeout 1m -failfast ./encryption/...
-            time go test -tags privileged -timeout 1m -failfast ./formatter/...
-            time go test -tags privileged -timeout 1m -failfast ./client/iface/...
-            time go test -tags privileged -timeout 1m -failfast ./route/...
-            time go test -tags privileged -timeout 1m -failfast ./sharedsock/...
-            time go test -tags privileged -timeout 1m -failfast ./util/...
-            time go test -tags privileged -timeout 1m -failfast ./version/...
+            time go test -timeout 8m -failfast -v -p 1 ./client/...
+            time go test -timeout 1m -failfast ./dns/...
+            time go test -timeout 1m -failfast ./encryption/...
+            time go test -timeout 1m -failfast ./formatter/...
+            time go test -timeout 1m -failfast ./client/iface/...
+            time go test -timeout 1m -failfast ./route/...
+            time go test -timeout 1m -failfast ./sharedsock/...
+            time go test -timeout 1m -failfast ./util/...
+            time go test -timeout 1m -failfast ./version/...

.github/workflows/golang-test-linux.yml

@@ -18,7 +18,7 @@ jobs:
       management: ${{ steps.filter.outputs.management }}
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
@@ -30,7 +30,7 @@ jobs:
               - 'management/**'
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -41,7 +41,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         id: cache
         with:
           path: |
@@ -119,12 +119,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -135,7 +135,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -158,11 +158,11 @@ jobs:
         run: git --no-pager diff --exit-code
 
       - name: Test
-        run: CGO_ENABLED=1 GOARCH=${{ matrix.arch }} CI=true go test -coverprofile=coverage.txt -tags devcert -timeout 10m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined)
+        run: CGO_ENABLED=1 GOARCH=${{ matrix.arch }} CI=true go test -coverprofile=coverage.txt -tags devcert -exec 'sudo' -timeout 10m -p 1 $(go list ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined)
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird
@@ -175,12 +175,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -192,7 +192,7 @@ jobs:
           echo "modcache_dir=$(go env GOMODCACHE)" >> $GITHUB_OUTPUT
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         id: cache-restore
         with:
           path: |
@@ -229,7 +229,7 @@ jobs:
             sh -c ' \
               apk update; apk add --no-cache \
                 ca-certificates iptables ip6tables dbus dbus-dev libpcap-dev build-base; \
-              go test -buildvcs=false -tags "devcert privileged" -v -timeout 10m -p 1 $(go list -buildvcs=false ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /upload-server -e /client/testutil/privileged)
+              go test -buildvcs=false -tags devcert -v -timeout 10m -p 1 $(go list -buildvcs=false ./... | grep -v -e /management -e /signal -e /relay -e /proxy -e /combined -e /client/ui -e /upload-server)
             '
 
   test_relay:
@@ -246,12 +246,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -266,7 +266,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -290,7 +290,7 @@ jobs:
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird
@@ -306,12 +306,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -325,7 +325,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -347,7 +347,7 @@ jobs:
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird
@@ -363,12 +363,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -383,7 +383,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -407,7 +407,7 @@ jobs:
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird
@@ -424,12 +424,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -440,7 +440,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -484,7 +484,7 @@ jobs:
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird
@@ -529,12 +529,12 @@ jobs:
             prom/prometheus
 
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -545,7 +545,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -579,11 +579,10 @@ jobs:
           CGO_ENABLED=1 GOARCH=${{ matrix.arch }} \
           NETBIRD_STORE_ENGINE=${{ matrix.store }} \
           CI=true \
+          GIT_BRANCH=${{ github.ref_name }} \
           go test -tags devcert -run=^$ -bench=. \
           -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE,GIT_BRANCH,GITHUB_RUN_ID' \
           -timeout 20m ./management/... ./shared/management/... $(go list ./management/... ./shared/management/... | grep -v -e /management/server/http)
-        env:
-          GIT_BRANCH: ${{ github.ref_name }}
 
   api_benchmark:
     name: "Management / Benchmark (API)"
@@ -624,12 +623,12 @@ jobs:
             prom/prometheus
 
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -640,7 +639,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -674,13 +673,12 @@ jobs:
           CGO_ENABLED=1 GOARCH=${{ matrix.arch }} \
           NETBIRD_STORE_ENGINE=${{ matrix.store }} \
           CI=true \
+          GIT_BRANCH=${{ github.ref_name }} \
           go test -tags=benchmark \
             -run=^$ \
             -bench=. \
             -exec 'sudo --preserve-env=CI,NETBIRD_STORE_ENGINE,GIT_BRANCH,GITHUB_RUN_ID' \
             -timeout 20m ./management/server/http/...
-        env:
-          GIT_BRANCH: ${{ github.ref_name }}
 
   api_integration_test:
     name: "Management / Integration"
@@ -694,12 +692,12 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
@@ -710,7 +708,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -736,7 +734,7 @@ jobs:
 
       - name: Upload coverage reports to Codecov
         if: matrix.arch == 'amd64'
-        uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f #v7.0.0
+        uses: codecov/codecov-action@e79a6962e0d4c0c17b229090214935d2e33f8354 #v6.0.1
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
           slug: netbirdio/netbird

.github/workflows/golang-test-windows.yml

@@ -18,12 +18,12 @@ jobs:
     runs-on: windows-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         id: go
         with:
           go-version-file: "go.mod"
@@ -35,7 +35,7 @@ jobs:
           echo "modcache=$(go env GOMODCACHE)" >> $env:GITHUB_ENV
 
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ${{ env.cache }}
@@ -68,7 +68,7 @@ jobs:
         run: |
           $packages = go list ./... | Where-Object { $_ -notmatch '/management' } | Where-Object { $_ -notmatch '/relay' } | Where-Object { $_ -notmatch '/signal' } | Where-Object { $_ -notmatch '/proxy' } | Where-Object { $_ -notmatch '/combined' }
           $goExe = "C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe"
-          $cmd = "$goExe test -tags `"devcert privileged`" -timeout 10m -p 1 $($packages -join ' ') > test-out.txt 2>&1"
+          $cmd = "$goExe test -tags=devcert -timeout 10m -p 1 $($packages -join ' ') > test-out.txt 2>&1"
           Set-Content -Path "${{ github.workspace }}\run-tests.cmd" -Value $cmd
 
       - name: test

.github/workflows/golangci-lint.yml

@@ -15,13 +15,13 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: codespell
         uses: codespell-project/actions-codespell@8f01853be192eb0f849a5c7d721450e7a467c579 # v2.2
         with:
-          ignore_words_list: erro,clienta,hastable,iif,groupd,testin,groupe,cros,ans,deriver,te,userA,ede,additionals,flate,recordin,unparseable
+          ignore_words_list: erro,clienta,hastable,iif,groupd,testin,groupe,cros,ans,deriver,te,userA,ede,additionals
           skip: go.mod,go.sum,**/proxy/web/**
   golangci:
     strategy:
@@ -40,7 +40,7 @@ jobs:
     timeout-minutes: 15
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Check for duplicate constants
@@ -48,7 +48,7 @@ jobs:
         run: |
           ! awk '/const \(/,/)/{print $0}' management/server/activity/codes.go | grep -o '= [0-9]*' | sort | uniq -d | grep .
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false

.github/workflows/install-script-test.yml

@@ -22,7 +22,7 @@ jobs:
     runs-on: ${{ matrix.os }}
     steps:
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 

.github/workflows/mobile-build-validation.yml

@@ -16,11 +16,11 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout repository
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
       - name: Setup Android SDK
@@ -28,13 +28,13 @@ jobs:
         with:
           cmdline-tools-version: 8512546
       - name: Setup Java
-        uses: actions/setup-java@1bcf9fb12cf4aa7d266a90ae39939e61372fe520
+        uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654
         with:
           java-version: "11"
           distribution: "adopt"
       - name: NDK Cache
         id: ndk-cache
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: /usr/local/lib/android/sdk/ndk
           key: ndk-cache-23.1.7779620
@@ -54,11 +54,11 @@ jobs:
     runs-on: macos-latest
     steps:
       - name: Checkout repository
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
       - name: install gomobile

.github/workflows/release.yml

@@ -27,7 +27,7 @@ jobs:
     runs-on: ubuntu-22.04
     steps:
       - name: Checkout
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
@@ -64,7 +64,7 @@ jobs:
         if: steps.check_diff.outputs.diff_exists == 'true'
         env:
           GO_VERSION: ${{ steps.goversion.outputs.version }}
-        uses: vmactions/freebsd-vm@b84ab5559b5a1bb4b8ee2737d2506a16e1737636 # v1.4.8
+        uses: vmactions/freebsd-vm@d1e65811565151536c0c894fff74f06351ed26e6 # v1.4.5
         with:
           usesh: true
           copyback: false
@@ -135,7 +135,7 @@ jobs:
       ghcr_images: ${{ steps.tag_and_push_images.outputs.images_markdown }}
     steps:
       - name: Checkout
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           fetch-depth: 0 # It is required for GoReleaser to work properly
           persist-credentials: false
@@ -166,12 +166,12 @@ jobs:
           fi
 
       - name: Set up Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
       - name: Cache Go modules
-        uses: actions/cache/restore@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache/restore@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ~/go/pkg/mod
@@ -186,9 +186,9 @@ jobs:
       - name: check git status
         run: git --no-pager diff --exit-code
       - name: Set up QEMU
-        uses: docker/setup-qemu-action@06116385d9baf250c9f4dcb4858b16962ea869c3 #v4.1.0
+        uses: docker/setup-qemu-action@ce360397dd3f832beb865e1373c09c0e9f86d70a #v4.0.0
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@d7f5e7f509e45cec5c76c4d5afdd7de93d0b3df5 #v4.1.0
+        uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd #v4.0.0
       - name: Login to Docker hub
         if: github.event_name != 'pull_request'
         uses: docker/login-action@650006c6eb7dba73a995cc03b0b2d7f5ca915bee # v4.2.0
@@ -221,7 +221,7 @@ jobs:
         run: goversioninfo -arm -64 -icon client/ui/assets/netbird.ico -manifest client/manifest.xml -product-name ${{ env.PRODUCT_NAME }} -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/resources_windows_arm64.syso
       - name: Run GoReleaser
         id: goreleaser
-        uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
+        uses: goreleaser/goreleaser-action@4c6ab561adb47e50c45ef534e2155934e91c40c1 # v7.2.0
         with:
           version: ${{ env.GORELEASER_VER }}
           args: release --clean ${{ env.flags }}
@@ -347,7 +347,7 @@ jobs:
       release_ui_artifact_url: ${{ steps.upload_release_ui.outputs.artifact-url }}
     steps:
       - name: Checkout
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           fetch-depth: 0 # It is required for GoReleaser to work properly
           persist-credentials: false
@@ -374,12 +374,12 @@ jobs:
           fi
 
       - name: Set up Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ~/go/pkg/mod
@@ -420,7 +420,7 @@ jobs:
         run: goversioninfo -arm -64 -icon client/ui/assets/netbird.ico -manifest client/ui/manifest.xml -product-name ${{ env.PRODUCT_NAME }}-"UI" -copyright "${{ env.COPYRIGHT }}" -ver-major ${{ steps.semver_parser.outputs.major }} -ver-minor ${{ steps.semver_parser.outputs.minor }} -ver-patch ${{ steps.semver_parser.outputs.patch }} -ver-build 0 -file-version ${{ steps.semver_parser.outputs.fullversion }}.0 -product-version ${{ steps.semver_parser.outputs.fullversion }}.0 -o client/ui/resources_windows_arm64.syso
 
       - name: Run GoReleaser
-        uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
+        uses: goreleaser/goreleaser-action@4c6ab561adb47e50c45ef534e2155934e91c40c1 # v7.2.0
         with:
           version: ${{ env.GORELEASER_VER }}
           args: release --config .goreleaser_ui.yaml --clean ${{ env.flags }}
@@ -464,17 +464,17 @@ jobs:
       - if: ${{ !startsWith(github.ref, 'refs/tags/v') }}
         run: echo "flags=--snapshot" >> $GITHUB_ENV
       - name: Checkout
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           fetch-depth: 0 # It is required for GoReleaser to work properly
           persist-credentials: false
       - name: Set up Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
           cache: false
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: |
             ~/go/pkg/mod
@@ -488,7 +488,7 @@ jobs:
         run: git --no-pager diff --exit-code
       - name: Run GoReleaser
         id: goreleaser
-        uses: goreleaser/goreleaser-action@5daf1e915a5f0af01ddbcd89a43b8061ff4f1a89 # v7.2.2
+        uses: goreleaser/goreleaser-action@4c6ab561adb47e50c45ef534e2155934e91c40c1 # v7.2.0
         with:
           version: ${{ env.GORELEASER_VER }}
           args: release --config .goreleaser_ui_darwin.yaml --clean ${{ env.flags }}
@@ -522,7 +522,7 @@ jobs:
       downloadPath: '${{ github.workspace }}\temp'
     steps:
       - name: Checkout
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
@@ -534,13 +534,13 @@ jobs:
         run: echo "C:\Program Files\7-Zip" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
 
       - name: Download release artifacts
-        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
+        uses: actions/download-artifact@70fc10c6e5e1ce46ad2ea6f2b72d43f7d47b13c3 # v8.0.1
         with:
           name: release
           path: release
 
       - name: Download UI release artifacts
-        uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
+        uses: actions/download-artifact@70fc10c6e5e1ce46ad2ea6f2b72d43f7d47b13c3 # v8.0.1
         with:
           name: release-ui
           path: release-ui

.github/workflows/test-infrastructure-files.yml

@@ -68,17 +68,17 @@ jobs:
         run: sudo apt-get install -y curl
 
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
 
       - name: Cache Go modules
-        uses: actions/cache@2c8a9bd7457de244a408f35966fab2fb45fda9c8 # v6.0.0
+        uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
         with:
           path: ~/go/pkg/mod
           key: ${{ runner.os }}-go-${{ hashFiles('**/go.sum') }}
@@ -207,7 +207,7 @@ jobs:
       - name: Build management docker image
         working-directory: management
         run: |
-          docker build -t netbirdio/management:latest --build-arg TARGETPLATFORM=. .
+          docker build -t netbirdio/management:latest .
 
       - name: Build signal binary
         working-directory: signal
@@ -216,7 +216,7 @@ jobs:
       - name: Build signal docker image
         working-directory: signal
         run: |
-          docker build -t netbirdio/signal:latest --build-arg TARGETPLATFORM=. .
+          docker build -t netbirdio/signal:latest .
 
       - name: Build relay binary
         working-directory: relay
@@ -225,7 +225,7 @@ jobs:
       - name: Build relay docker image
         working-directory: relay
         run: |
-          docker build -t netbirdio/relay:latest --build-arg TARGETPLATFORM=. .
+          docker build -t netbirdio/relay:latest .
 
       - name: run docker compose up
         working-directory: infrastructure_files/artifacts
@@ -256,7 +256,7 @@ jobs:
         run: sudo apt-get install -y jq
 
       - name: Checkout code
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
 

.github/workflows/wasm-build-validation.yml

@@ -19,11 +19,11 @@ jobs:
       GOARCH: wasm
     steps:
       - name: Checkout repository
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
       - name: Install dependencies
@@ -44,11 +44,11 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout repository
-        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
+        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Install Go
-        uses: actions/setup-go@924ae3a1cded613372ab5595356fb5720e22ba16 # v6.5.0
+        uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
           go-version-file: "go.mod"
       - name: Build Wasm client

.goreleaser.yaml

@@ -247,7 +247,7 @@ dockers_v2:
        - netbirdio/netbird
        - ghcr.io/netbirdio/netbird
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: client/Dockerfile
      extra_files:
@@ -295,7 +295,7 @@ dockers_v2:
        - netbirdio/relay
        - ghcr.io/netbirdio/relay
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: relay/Dockerfile
      platforms:
@@ -317,7 +317,7 @@ dockers_v2:
        - netbirdio/signal
        - ghcr.io/netbirdio/signal
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: signal/Dockerfile
      platforms:
@@ -339,7 +339,7 @@ dockers_v2:
        - netbirdio/management
        - ghcr.io/netbirdio/management
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: management/Dockerfile
      platforms:
@@ -361,7 +361,7 @@ dockers_v2:
        - netbirdio/upload
        - ghcr.io/netbirdio/upload
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: upload-server/Dockerfile
      platforms:
@@ -383,7 +383,7 @@ dockers_v2:
        - netbirdio/netbird-server
        - ghcr.io/netbirdio/netbird-server
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: combined/Dockerfile
      platforms:
@@ -405,7 +405,7 @@ dockers_v2:
        - netbirdio/reverse-proxy
        - ghcr.io/netbirdio/reverse-proxy
      tags:
-       - "{{ .Version }}"
+       - "v{{ .Version }}"
        - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: proxy/Dockerfile
      platforms:
@@ -462,13 +462,9 @@ checksum:
     - glob: ./infrastructure_files/getting-started-with-zitadel.sh
     - glob: ./release_files/install.sh
     - glob: ./infrastructure_files/getting-started.sh
-    - glob: ./infrastructure_files/getting-started-enterprise.sh
-    - glob: ./infrastructure_files/migrate-to-enterprise.sh
 
 release:
   extra_files:
     - glob: ./infrastructure_files/getting-started-with-zitadel.sh
     - glob: ./release_files/install.sh
     - glob: ./infrastructure_files/getting-started.sh
-    - glob: ./infrastructure_files/getting-started-enterprise.sh
-    - glob: ./infrastructure_files/migrate-to-enterprise.sh

Makefile

@@ -1,4 +1,4 @@
-.PHONY: lint lint-all lint-install setup-hooks test-unit test-privileged
+.PHONY: lint lint-all lint-install setup-hooks
 GOLANGCI_LINT := $(shell pwd)/bin/golangci-lint
 
 # Install golangci-lint locally if needed
@@ -25,15 +25,3 @@ setup-hooks:
 	@git config core.hooksPath .githooks
 	@chmod +x .githooks/pre-push
 	@echo "✅ Git hooks configured! Pre-push will now run 'make lint'"
-
-# Host-safe unit tests: excludes the privileged-tagged tests (root / system-mutating).
-# Runs as a normal user with no sudo and leaves host networking untouched.
-test-unit:
-	@go test -tags devcert -timeout 10m ./...
-
-# Privileged suite: runs the `privileged`-tagged tests inside a --privileged
-# --cap-add=NET_ADMIN container via the ory/dockertest harness. Requires Docker.
-# Narrow the run with env vars, e.g.:
-#   PRIV_RUN=TestNftablesManager PRIV_PKGS=./client/firewall/nftables/... make test-privileged
-test-privileged:
-	@go test -tags 'devcert privileged' -timeout 30m -run TestRunPrivilegedSuiteInDocker -v ./client/testutil/privileged/...

README.md

@@ -33,15 +33,10 @@
   <br/>
   <br/>
   <strong>
-    🚀 <a href="https://netbird.io/careers">We are hiring! Join us at https://netbird.io/careers</a>
+    🚀 <a href="https://careers.netbird.io">We are hiring! Join us at careers.netbird.io</a>
   </strong>
 </p>
 
-> ### 🤖 NetBird Agent Network (Beta)
-> Identity-aware access control for AI agents — keyless access to LLM APIs and private
-> resources over the encrypted NetBird tunnel. See [`agent-network/`](agent-network/) or
-> read the docs at **[netbird.ai](https://netbird.ai)**.
-
 **NetBird combines a configuration-free peer-to-peer private network and a centralized access control system in a single platform, making it easy to create secure private networks for your organization or home.**
 
 **Connect.** NetBird creates a WireGuard-based overlay network that automatically connects your machines over an encrypted tunnel, leaving behind the hassle of opening ports, complex firewall rules, VPN gateways, and so forth.

agent-network/README.md

@@ -1,39 +0,0 @@
-# NetBird Agent Network
-
-Agent Network is NetBird's access control layer for AI agents and the people who run
-them. It gives every agent a real identity, tied to your identity provider (IdP), and
-governs what it can reach — the LLM APIs and AI gateways it can call, and the internal
-resources it can access. Traffic flows only over the encrypted NetBird tunnel, scoped by
-policy, with no API keys to leak.
-
-> **Beta.** Agent Network is open source and can be self-hosted on your own
-> infrastructure.
-
-## How it works
-
-Agent Network is built on two existing NetBird capabilities:
-
-- **Overlay network** — the encrypted WireGuard mesh between peers.
-- **Reverse proxy** — a NetBird peer that terminates LLM requests, establishes the
-  caller's identity, evaluates policies/limits/guardrails, injects the upstream provider
-  key server-side, forwards to the API or gateway, and records usage.
-
-LLM traffic is routed through the proxy's identity-aware pipeline, while internal
-resources (databases, internal APIs, self-hosted models) are reached directly over
-peer-to-peer WireGuard tunnels, governed by the same identities and access policies.
-
-## Where the code lives
-
-There is no separate "agent-network" service — it reuses the reverse-proxy and management
-components:
-
-- [`proxy/`](../proxy) — the NetBird reverse proxy that serves the agent network endpoint
-  and runs the per-request middleware pipeline.
-- [`management/internals/modules/reverseproxy/`](../management/internals/modules/reverseproxy)
-  — the management-side control plane: providers, policies, guardrails, limits, routing,
-  and usage/access logs.
-
-## Documentation
-
-Full documentation, architecture, and quickstart:
-**https://docs.netbird.io/agent-network**

client/android/client.go

@@ -151,9 +151,9 @@ func (c *Client) Run(platformFiles PlatformFiles, urlOpener URLOpener, isAndroid
 
 	// todo do not throw error in case of cancelled context
 	ctx = internal.CtxInitState(ctx)
-	connectClient := internal.NewConnectClient(ctx, cfg, c.recorder)
+	connectClient := internal.NewConnectClient(ctx, c.recorder)
 	c.setState(cfg, cacheDir, connectClient)
-	return connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
+	return connectClient.RunOnAndroid(cfg, c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
 }
 
 // RunWithoutLogin we apply this type of run function when the backed has been started without UI (i.e. after reboot).
@@ -186,9 +186,9 @@ func (c *Client) RunWithoutLogin(platformFiles PlatformFiles, dns *DNSList, dnsR
 
 	// todo do not throw error in case of cancelled context
 	ctx = internal.CtxInitState(ctx)
-	connectClient := internal.NewConnectClient(ctx, cfg, c.recorder)
+	connectClient := internal.NewConnectClient(ctx, c.recorder)
 	c.setState(cfg, cacheDir, connectClient)
-	return connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
+	return connectClient.RunOnAndroid(cfg, c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
 }
 
 // Stop the internal client and free the resources

client/cmd/debug.go

@@ -130,7 +130,7 @@ func debugConfigDump(cmd *cobra.Command, _ []string) error {
 
 	client := proto.NewDaemonServiceClient(conn)
 	resp, err := client.GetConfig(cmd.Context(), &proto.GetConfigRequest{
-		ProfileName: string(activeProf.ID),
+		ProfileName: activeProf.Name,
 		Username:    currUser.Username,
 	})
 	if err != nil {

client/cmd/login.go

@@ -227,7 +227,7 @@ func switchProfile(ctx context.Context, handle string, username string) (profile
 		Username:    &username,
 	})
 	if err != nil {
-		return "", fmt.Errorf("switch profile failed: %w", err)
+		return "", fmt.Errorf("switch profile failed: %v", err)
 	}
 
 	return profilemanager.ID(resp.Id), nil

client/cmd/profile.go

@@ -138,23 +138,26 @@ func addProfileFunc(cmd *cobra.Command, args []string) error {
 		return err
 	}
 
-	currUser, err := user.Current()
-	if err != nil {
-		return fmt.Errorf("get current user: %w", err)
-	}
-
 	conn, err := DialClientGRPCServer(cmd.Context(), daemonAddr)
 	if err != nil {
 		return fmt.Errorf("connect to service CLI interface: %w", err)
 	}
 	defer conn.Close()
 
+	currUser, err := user.Current()
+	if err != nil {
+		return fmt.Errorf("get current user: %w", err)
+	}
+
 	daemonClient := proto.NewDaemonServiceClient(conn)
 	profileName := args[0]
 
-	id, err := addProfileOnDaemon(cmd.Context(), daemonClient, profileName, currUser.Username)
+	resp, err := daemonClient.AddProfile(cmd.Context(), &proto.AddProfileRequest{
+		ProfileName: profileName,
+		Username:    currUser.Username,
+	})
 	if err != nil {
-		return err
+		return fmt.Errorf("add profile request: %w", err)
 	}
 
 	dupCount, _ := countProfilesWithName(cmd.Context(), daemonClient, currUser.Username, profileName)
@@ -163,6 +166,7 @@ func addProfileFunc(cmd *cobra.Command, args []string) error {
 		cmd.Println("Use `netbird profile list --show-id` to disambiguate later.")
 	}
 
+	id := profilemanager.ID(resp.Id)
 	cmd.Printf("Profile added: %s  %s\n", id.ShortID(), profilemanager.StripCtrlChars(profileName))
 	return nil
 
@@ -326,19 +330,3 @@ func wrapAmbiguityError(err error, handle string) error {
 	}
 	return err
 }
-
-// addProfileOnDaemon issues the AddProfile RPC on an existing daemon client
-// and returns the new profile's ID. It is the single entry point for profile
-// creation, shared by `netbird profile add` and the `netbird up --profile
-// <name>` auto-create path.
-func addProfileOnDaemon(ctx context.Context, client proto.DaemonServiceClient, profileName, username string) (profilemanager.ID, error) {
-	resp, err := client.AddProfile(ctx, &proto.AddProfileRequest{
-		ProfileName: profileName,
-		Username:    username,
-	})
-	if err != nil {
-		return "", fmt.Errorf("add profile failed: %w", err)
-	}
-
-	return profilemanager.ID(resp.Id), nil
-}

client/cmd/root.go

@@ -20,6 +20,7 @@ import (
 	"github.com/spf13/cobra"
 	"github.com/spf13/pflag"
 	"google.golang.org/grpc"
+	"google.golang.org/grpc/connectivity"
 	"google.golang.org/grpc/credentials/insecure"
 
 	daddr "github.com/netbirdio/netbird/client/internal/daemonaddr"
@@ -261,17 +262,46 @@ func FlagNameToEnvVar(cmdFlag string, prefix string) string {
 	return prefix + upper
 }
 
-// DialClientGRPCServer returns client connection to the daemon server.
+// DialClientGRPCServer returns client connection to the daemon server. It waits
+// (up to the timeout) for the daemon to become reachable so an `up` issued right
+// after `service start` tolerates the startup race. Instead of grpc's blocking
+// dial — whose raw "transport failed" retry warnings are silenced by the logger
+// config — we drive the wait ourselves and emit one clean line per failed attempt.
 func DialClientGRPCServer(ctx context.Context, addr string) (*grpc.ClientConn, error) {
 	ctx, cancel := context.WithTimeout(ctx, time.Second*10)
 	defer cancel()
 
-	return grpc.DialContext(
+	conn, err := grpc.DialContext(
 		ctx,
 		strings.TrimPrefix(addr, "tcp://"),
 		grpc.WithTransportCredentials(insecure.NewCredentials()),
-		grpc.WithBlock(),
 	)
+	if err != nil {
+		return nil, err
+	}
+
+	conn.Connect()
+	for {
+		state := conn.GetState()
+		if state == connectivity.Ready {
+			return conn, nil
+		}
+		// Log only once the connection has actually failed — not during the
+		// brief Idle/Connecting phase on a healthy daemon (avoids a spurious
+		// line + wait when the daemon is already up).
+		if state == connectivity.TransientFailure {
+			log.Infof("waiting for the netbird daemon to become available at %s...", addr)
+		}
+		// Wake on the next state change, but at least every second so a stuck
+		// TransientFailure re-logs at a steady cadence until the timeout.
+		waitCtx, waitCancel := context.WithTimeout(ctx, time.Second)
+		conn.WaitForStateChange(waitCtx, state)
+		waitCancel()
+		if ctx.Err() != nil {
+			_ = conn.Close()
+			return nil, fmt.Errorf("daemon not reachable at %s: %w", addr, ctx.Err())
+		}
+	}
 }
 
 // WithBackOff execute function in backoff cycle.

client/cmd/service_privileged_test.go

@@ -1,196 +0,0 @@
-//go:build privileged
-
-package cmd
-
-import (
-	"context"
-	"fmt"
-	"os"
-	"runtime"
-	"testing"
-	"time"
-
-	"github.com/kardianos/service"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-const (
-	serviceStartTimeout = 10 * time.Second
-	serviceStopTimeout  = 5 * time.Second
-	statusPollInterval  = 500 * time.Millisecond
-)
-
-// waitForServiceStatus waits for service to reach expected status with timeout
-func waitForServiceStatus(expectedStatus service.Status, timeout time.Duration) (bool, error) {
-	cfg, err := newSVCConfig()
-	if err != nil {
-		return false, err
-	}
-
-	ctxSvc, cancel := context.WithCancel(context.Background())
-	defer cancel()
-
-	s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
-	if err != nil {
-		return false, err
-	}
-
-	ctx, timeoutCancel := context.WithTimeout(context.Background(), timeout)
-	defer timeoutCancel()
-
-	ticker := time.NewTicker(statusPollInterval)
-	defer ticker.Stop()
-
-	for {
-		select {
-		case <-ctx.Done():
-			return false, fmt.Errorf("timeout waiting for service status %v", expectedStatus)
-		case <-ticker.C:
-			status, err := s.Status()
-			if err != nil {
-				// Continue polling on transient errors
-				continue
-			}
-			if status == expectedStatus {
-				return true, nil
-			}
-		}
-	}
-}
-
-// TestServiceLifecycle tests the complete service lifecycle
-func TestServiceLifecycle(t *testing.T) {
-	// TODO: Add support for Windows and macOS
-	if runtime.GOOS != "linux" && runtime.GOOS != "freebsd" {
-		t.Skipf("Skipping service lifecycle test on unsupported OS: %s", runtime.GOOS)
-	}
-
-	if os.Getenv("CONTAINER") == "true" {
-		t.Skip("Skipping service lifecycle test in container environment")
-	}
-
-	originalServiceName := serviceName
-	serviceName = "netbirdtest" + fmt.Sprintf("%d", time.Now().Unix())
-	defer func() {
-		serviceName = originalServiceName
-	}()
-
-	tempDir := t.TempDir()
-	configPath = fmt.Sprintf("%s/netbird-test-config.json", tempDir)
-	logLevel = "info"
-	daemonAddr = fmt.Sprintf("unix://%s/netbird-test.sock", tempDir)
-
-	// Ensure cleanup even if a subtest fails and Stop/Uninstall subtests don't run.
-	t.Cleanup(func() {
-		cfg, err := newSVCConfig()
-		if err != nil {
-			t.Errorf("cleanup: create service config: %v", err)
-			return
-		}
-		ctxSvc, cancel := context.WithCancel(context.Background())
-		defer cancel()
-		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
-		if err != nil {
-			t.Errorf("cleanup: create service: %v", err)
-			return
-		}
-
-		// If the subtests already cleaned up, there's nothing to do.
-		if _, err := s.Status(); err != nil {
-			return
-		}
-
-		if err := s.Stop(); err != nil {
-			t.Errorf("cleanup: stop service: %v", err)
-		}
-		if err := s.Uninstall(); err != nil {
-			t.Errorf("cleanup: uninstall service: %v", err)
-		}
-	})
-
-	ctx := context.Background()
-
-	t.Run("Install", func(t *testing.T) {
-		installCmd.SetContext(ctx)
-		err := installCmd.RunE(installCmd, []string{})
-		require.NoError(t, err)
-
-		cfg, err := newSVCConfig()
-		require.NoError(t, err)
-
-		ctxSvc, cancel := context.WithCancel(context.Background())
-		defer cancel()
-
-		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
-		require.NoError(t, err)
-
-		status, err := s.Status()
-		assert.NoError(t, err)
-		assert.NotEqual(t, service.StatusUnknown, status)
-	})
-
-	t.Run("Start", func(t *testing.T) {
-		startCmd.SetContext(ctx)
-		err := startCmd.RunE(startCmd, []string{})
-		require.NoError(t, err)
-
-		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
-		require.NoError(t, err)
-		assert.True(t, running)
-	})
-
-	t.Run("Restart", func(t *testing.T) {
-		restartCmd.SetContext(ctx)
-		err := restartCmd.RunE(restartCmd, []string{})
-		require.NoError(t, err)
-
-		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
-		require.NoError(t, err)
-		assert.True(t, running)
-	})
-
-	t.Run("Reconfigure", func(t *testing.T) {
-		originalLogLevel := logLevel
-		logLevel = "debug"
-		defer func() {
-			logLevel = originalLogLevel
-		}()
-
-		reconfigureCmd.SetContext(ctx)
-		err := reconfigureCmd.RunE(reconfigureCmd, []string{})
-		require.NoError(t, err)
-
-		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
-		require.NoError(t, err)
-		assert.True(t, running)
-	})
-
-	t.Run("Stop", func(t *testing.T) {
-		stopCmd.SetContext(ctx)
-		err := stopCmd.RunE(stopCmd, []string{})
-		require.NoError(t, err)
-
-		stopped, err := waitForServiceStatus(service.StatusStopped, serviceStopTimeout)
-		require.NoError(t, err)
-		assert.True(t, stopped)
-	})
-
-	t.Run("Uninstall", func(t *testing.T) {
-		uninstallCmd.SetContext(ctx)
-		err := uninstallCmd.RunE(uninstallCmd, []string{})
-		require.NoError(t, err)
-
-		cfg, err := newSVCConfig()
-		require.NoError(t, err)
-
-		ctxSvc, cancel := context.WithCancel(context.Background())
-		defer cancel()
-
-		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
-		require.NoError(t, err)
-
-		_, err = s.Status()
-		assert.Error(t, err)
-	})
-}

client/cmd/service_test.go

@@ -1,12 +1,16 @@
 package cmd
 
 import (
+	"context"
+	"fmt"
 	"os"
 	"os/signal"
 	"runtime"
 	"syscall"
 	"testing"
+	"time"
 
+	"github.com/kardianos/service"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 )
@@ -27,6 +31,186 @@ func TestMain(m *testing.M) {
 	os.Exit(m.Run())
 }
 
+const (
+	serviceStartTimeout = 10 * time.Second
+	serviceStopTimeout  = 5 * time.Second
+	statusPollInterval  = 500 * time.Millisecond
+)
+
+// waitForServiceStatus waits for service to reach expected status with timeout
+func waitForServiceStatus(expectedStatus service.Status, timeout time.Duration) (bool, error) {
+	cfg, err := newSVCConfig()
+	if err != nil {
+		return false, err
+	}
+
+	ctxSvc, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
+	if err != nil {
+		return false, err
+	}
+
+	ctx, timeoutCancel := context.WithTimeout(context.Background(), timeout)
+	defer timeoutCancel()
+
+	ticker := time.NewTicker(statusPollInterval)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ctx.Done():
+			return false, fmt.Errorf("timeout waiting for service status %v", expectedStatus)
+		case <-ticker.C:
+			status, err := s.Status()
+			if err != nil {
+				// Continue polling on transient errors
+				continue
+			}
+			if status == expectedStatus {
+				return true, nil
+			}
+		}
+	}
+}
+
+// TestServiceLifecycle tests the complete service lifecycle
+func TestServiceLifecycle(t *testing.T) {
+	// TODO: Add support for Windows and macOS
+	if runtime.GOOS != "linux" && runtime.GOOS != "freebsd" {
+		t.Skipf("Skipping service lifecycle test on unsupported OS: %s", runtime.GOOS)
+	}
+
+	if os.Getenv("CONTAINER") == "true" {
+		t.Skip("Skipping service lifecycle test in container environment")
+	}
+
+	originalServiceName := serviceName
+	serviceName = "netbirdtest" + fmt.Sprintf("%d", time.Now().Unix())
+	defer func() {
+		serviceName = originalServiceName
+	}()
+
+	tempDir := t.TempDir()
+	configPath = fmt.Sprintf("%s/netbird-test-config.json", tempDir)
+	logLevel = "info"
+	daemonAddr = fmt.Sprintf("unix://%s/netbird-test.sock", tempDir)
+
+	// Ensure cleanup even if a subtest fails and Stop/Uninstall subtests don't run.
+	t.Cleanup(func() {
+		cfg, err := newSVCConfig()
+		if err != nil {
+			t.Errorf("cleanup: create service config: %v", err)
+			return
+		}
+		ctxSvc, cancel := context.WithCancel(context.Background())
+		defer cancel()
+		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
+		if err != nil {
+			t.Errorf("cleanup: create service: %v", err)
+			return
+		}
+
+		// If the subtests already cleaned up, there's nothing to do.
+		if _, err := s.Status(); err != nil {
+			return
+		}
+
+		if err := s.Stop(); err != nil {
+			t.Errorf("cleanup: stop service: %v", err)
+		}
+		if err := s.Uninstall(); err != nil {
+			t.Errorf("cleanup: uninstall service: %v", err)
+		}
+	})
+
+	ctx := context.Background()
+
+	t.Run("Install", func(t *testing.T) {
+		installCmd.SetContext(ctx)
+		err := installCmd.RunE(installCmd, []string{})
+		require.NoError(t, err)
+
+		cfg, err := newSVCConfig()
+		require.NoError(t, err)
+
+		ctxSvc, cancel := context.WithCancel(context.Background())
+		defer cancel()
+
+		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
+		require.NoError(t, err)
+
+		status, err := s.Status()
+		assert.NoError(t, err)
+		assert.NotEqual(t, service.StatusUnknown, status)
+	})
+
+	t.Run("Start", func(t *testing.T) {
+		startCmd.SetContext(ctx)
+		err := startCmd.RunE(startCmd, []string{})
+		require.NoError(t, err)
+
+		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
+		require.NoError(t, err)
+		assert.True(t, running)
+	})
+
+	t.Run("Restart", func(t *testing.T) {
+		restartCmd.SetContext(ctx)
+		err := restartCmd.RunE(restartCmd, []string{})
+		require.NoError(t, err)
+
+		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
+		require.NoError(t, err)
+		assert.True(t, running)
+	})
+
+	t.Run("Reconfigure", func(t *testing.T) {
+		originalLogLevel := logLevel
+		logLevel = "debug"
+		defer func() {
+			logLevel = originalLogLevel
+		}()
+
+		reconfigureCmd.SetContext(ctx)
+		err := reconfigureCmd.RunE(reconfigureCmd, []string{})
+		require.NoError(t, err)
+
+		running, err := waitForServiceStatus(service.StatusRunning, serviceStartTimeout)
+		require.NoError(t, err)
+		assert.True(t, running)
+	})
+
+	t.Run("Stop", func(t *testing.T) {
+		stopCmd.SetContext(ctx)
+		err := stopCmd.RunE(stopCmd, []string{})
+		require.NoError(t, err)
+
+		stopped, err := waitForServiceStatus(service.StatusStopped, serviceStopTimeout)
+		require.NoError(t, err)
+		assert.True(t, stopped)
+	})
+
+	t.Run("Uninstall", func(t *testing.T) {
+		uninstallCmd.SetContext(ctx)
+		err := uninstallCmd.RunE(uninstallCmd, []string{})
+		require.NoError(t, err)
+
+		cfg, err := newSVCConfig()
+		require.NoError(t, err)
+
+		ctxSvc, cancel := context.WithCancel(context.Background())
+		defer cancel()
+
+		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
+		require.NoError(t, err)
+
+		_, err = s.Status()
+		assert.Error(t, err)
+	})
+}
+
 // TestServiceEnvVars tests environment variable parsing
 func TestServiceEnvVars(t *testing.T) {
 	tests := []struct {

client/cmd/status.go

@@ -11,6 +11,7 @@ import (
 	"google.golang.org/grpc/status"
 
 	"github.com/netbirdio/netbird/client/internal"
+	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	"github.com/netbirdio/netbird/client/proto"
 	nbstatus "github.com/netbirdio/netbird/client/status"
 	"github.com/netbirdio/netbird/util"
@@ -110,10 +111,11 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 		return nil
 	}
 
-	// Resolve the active profile's display name via the daemon, which runs
-	// as root and can read the per-user profile files. The local profile
-	// manager only knows the active profile ID, not its display name.
-	profName := getActiveProfileName(ctx)
+	pm := profilemanager.NewProfileManager()
+	var profName string
+	if activeProf, err := pm.GetActiveProfile(); err == nil {
+		profName = activeProf.Name
+	}
 
 	var outputInformationHolder = nbstatus.ConvertToStatusOutputOverview(resp.GetFullStatus(), nbstatus.ConvertOptions{
 		Anonymize:            anonymizeFlag,
@@ -165,25 +167,6 @@ func getStatus(ctx context.Context, fullPeerStatus bool, shouldRunProbes bool) (
 	return resp, nil
 }
 
-// getActiveProfileName asks the daemon for the active profile's display
-// name. The daemon runs as root and can read the per-user profile files to
-// resolve the ID to its human-readable name. Returns an empty string on any
-// error so status output degrades gracefully.
-func getActiveProfileName(ctx context.Context) string {
-	conn, err := DialClientGRPCServer(ctx, daemonAddr)
-	if err != nil {
-		return ""
-	}
-	defer conn.Close()
-
-	resp, err := proto.NewDaemonServiceClient(conn).GetActiveProfile(ctx, &proto.GetActiveProfileRequest{})
-	if err != nil {
-		return ""
-	}
-
-	return resp.GetProfileName()
-}
-
 func parseFilters() error {
 	switch strings.ToLower(statusFilter) {
 	case "", "idle", "connecting", "connected":

client/cmd/up.go

@@ -128,9 +128,15 @@ func upFunc(cmd *cobra.Command, args []string) error {
 	var profileSwitched bool
 	// switch profile if provided
 	if profileName != "" {
-		if err := switchOrCreateProfile(cmd.Context(), pm, profileName, username.Username); err != nil {
+		resolvedID, err := switchProfile(cmd.Context(), profileName, username.Username)
+		if err != nil {
 			return fmt.Errorf("switch profile: %v", err)
 		}
+
+		if err := pm.SwitchProfile(resolvedID); err != nil {
+			return fmt.Errorf("switch profile: %v", err)
+		}
+
 		profileSwitched = true
 	}
 
@@ -145,52 +151,6 @@ func upFunc(cmd *cobra.Command, args []string) error {
 	return runInDaemonMode(ctx, cmd, pm, activeProf, profileSwitched)
 }
 
-// switchOrCreateProfile switches the active profile to the one identified by
-// handle, creating it first when it does not exist yet. This restores the
-// pre-0.73 behaviour where `netbird up --profile <name>` auto-creates a
-// missing profile instead of failing.
-func switchOrCreateProfile(ctx context.Context, pm *profilemanager.ProfileManager, handle, username string) error {
-	resolvedID, err := switchProfile(ctx, handle, username)
-	if err != nil {
-		st, ok := gstatus.FromError(err)
-		if !ok || st.Code() != codes.NotFound {
-			return err
-		}
-		// Don't fail immediately on a create error: a concurrent run may
-		// have created the profile between the NotFound above and this
-		// call, in which case the retried switch still succeeds. Only
-		// surface the create error if the switch also fails.
-		_, createErr := createProfile(ctx, handle, username)
-		if resolvedID, err = switchProfile(ctx, handle, username); err != nil {
-			if createErr != nil {
-				return fmt.Errorf("create profile: %w", createErr)
-			}
-			return err
-		}
-	}
-
-	if err := pm.SwitchProfile(resolvedID); err != nil {
-		return err
-	}
-	return nil
-}
-
-// createProfile dials the daemon and creates a new profile with the given
-// display name, returning its generated ID. Use addProfileOnDaemon directly
-// when a daemon client is already available to reuse the connection.
-func createProfile(ctx context.Context, profileName, username string) (profilemanager.ID, error) {
-	conn, err := DialClientGRPCServer(ctx, daemonAddr)
-	if err != nil {
-		//nolint
-		return "", fmt.Errorf("failed to connect to daemon error: %v\n"+
-			"If the daemon is not running please run: "+
-			"\nnetbird service install \nnetbird service start\n", err)
-	}
-	defer conn.Close()
-
-	return addProfileOnDaemon(ctx, proto.NewDaemonServiceClient(conn), profileName, username)
-}
-
 func runInForegroundMode(ctx context.Context, cmd *cobra.Command, activeProf *profilemanager.Profile) error {
 	// override the default profile filepath if provided
 	if configPath != "" {
@@ -241,10 +201,10 @@ func runInForegroundMode(ctx context.Context, cmd *cobra.Command, activeProf *pr
 	r := peer.NewRecorder(config.ManagementURL.String())
 	r.GetFullStatus()
 
-	connectClient := internal.NewConnectClient(ctx, config, r)
+	connectClient := internal.NewConnectClient(ctx, r)
 	SetupDebugHandler(ctx, config, r, connectClient, "")
 
-	return connectClient.Run(nil, util.FindFirstLogPath(logFiles))
+	return connectClient.Run(config, nil, util.FindFirstLogPath(logFiles))
 }
 
 func runInDaemonMode(ctx context.Context, cmd *cobra.Command, pm *profilemanager.ProfileManager, activeProf *profilemanager.Profile, profileSwitched bool) error {

client/embed/embed.go

@@ -264,34 +264,24 @@ func (c *Client) Start(startCtx context.Context) error {
 	if err, _ := authClient.Login(ctx, c.setupKey, c.jwtToken); err != nil {
 		return fmt.Errorf("login: %w", err)
 	}
-	client := internal.NewConnectClient(ctx, c.config, c.recorder)
+	client := internal.NewConnectClient(ctx, c.recorder)
 	client.SetSyncResponsePersistence(true)
 
-	// either startup error (permanent backoff err) or nil err (successful engine up)
+	// The supervisor owns the run; we wait until it is established, ends with a
+	// startup error (permanent backoff err), or startCtx expires.
 	// TODO: make after-startup backoff err available
-	run := make(chan struct{})
-	clientErr := make(chan error, 1)
-	go func() {
-		if err := client.Run(run, ""); err != nil {
-			clientErr <- err
-		}
-	}()
+	client.RunAsync(c.config, nil)
 
-	select {
-	case <-startCtx.Done():
-		// ConnectClient.Stop now cancels its own run context and waits for the
-		// run loop to tear the engine down, so this cancel() is no longer
-		// required to break the deadlock and could be removed. It is kept as a
-		// defensive belt-and-suspenders: cancelling the parent context first
-		// guarantees the run loop is unblocked even if Stop's contract regresses.
+	if err := client.WaitEstablishedOrDone(startCtx); err != nil {
+		// Either startCtx expired while connecting, or the run ended before it
+		// established. Cancel the client context before stopping: Engine.Start
+		// blocks on the signal stream while holding the engine mutex and only
+		// unblocks on cancellation. Stopping first would deadlock on that mutex.
 		cancel()
 		if stopErr := client.Stop(); stopErr != nil {
-			return fmt.Errorf("stop error after context done. Stop error: %w. Context done: %w", stopErr, startCtx.Err())
+			return fmt.Errorf("stop error after startup failure. Stop error: %w. Startup: %w", stopErr, err)
 		}
-		return startCtx.Err()
-	case err := <-clientErr:
 		return fmt.Errorf("startup: %w", err)
-	case <-run:
 	}
 
 	c.connect = client

client/firewall/iptables/manager_linux_test.go

@@ -1,5 +1,3 @@
-//go:build privileged
-
 package iptables
 
 import (

client/firewall/iptables/router_linux_test.go

@@ -1,4 +1,4 @@
-//go:build !android && privileged
+//go:build !android
 
 package iptables
 

client/firewall/nftables/manager_linux_test.go

@@ -1,5 +1,3 @@
-//go:build privileged
-
 package nftables
 
 import (

client/firewall/nftables/router_linux_test.go

@@ -1,4 +1,4 @@
-//go:build !android && privileged
+//go:build !android
 
 package nftables
 

client/iface/iface_test.go

@@ -1,5 +1,3 @@
-//go:build privileged
-
 package iface
 
 import (

client/iface/wgproxy/bind/proxy.go

@@ -136,11 +136,6 @@ func (p *ProxyBind) CloseConn() error {
 	return p.close()
 }
 
-// InjectPacket is a no-op for the userspace proxy: first-packet reinjection is kernel-only.
-func (p *ProxyBind) InjectPacket(_ []byte) error {
-	return nil
-}
-
 func (p *ProxyBind) close() error {
 	if p.remoteConn == nil {
 		return nil

client/iface/wgproxy/ebpf/wrapper.go

@@ -219,17 +219,6 @@ func (p *ProxyWrapper) RedirectAs(endpoint *net.UDPAddr) {
 	p.pausedCond.L.Unlock()
 }
 
-// InjectPacket writes b to the remote peer over the underlying transport.
-func (p *ProxyWrapper) InjectPacket(b []byte) error {
-	if p.remoteConn == nil {
-		return errors.New("proxy not started")
-	}
-	if _, err := p.remoteConn.Write(b); err != nil {
-		return err
-	}
-	return nil
-}
-
 // CloseConn close the remoteConn and automatically remove the conn instance from the map
 func (p *ProxyWrapper) CloseConn() error {
 	if p.cancel == nil {

client/iface/wgproxy/proxy.go

@@ -18,9 +18,4 @@ type Proxy interface {
 	RedirectAs(endpoint *net.UDPAddr)
 	CloseConn() error
 	SetDisconnectListener(disconnected func())
-
-	// InjectPacket writes a raw packet directly to the remote peer over the underlying transport,
-	// bypassing WireGuard. Used to replay the captured lazyconn handshake initiation. Only the
-	// kernel-mode proxies act on it; the userspace proxy is a no-op since reinjection is kernel-only.
-	InjectPacket(b []byte) error
 }

client/iface/wgproxy/proxy_linux_test.go

@@ -1,4 +1,4 @@
-//go:build linux && !android && privileged
+//go:build linux && !android
 
 package wgproxy
 

client/iface/wgproxy/proxy_seed_test.go

@@ -1,4 +1,4 @@
-//go:build !linux || !privileged
+//go:build !linux
 
 package wgproxy
 

client/iface/wgproxy/redirect_test.go

@@ -1,4 +1,4 @@
-//go:build linux && !android && privileged
+//go:build linux && !android
 
 package wgproxy
 
@@ -26,6 +26,64 @@ func compareUDPAddr(addr1, addr2 net.Addr) bool {
 	return udpAddr1.IP.Equal(udpAddr2.IP) && udpAddr1.Port == udpAddr2.Port
 }
 
+// TestRedirectAs_eBPF_IPv4 tests RedirectAs with eBPF proxy using IPv4 addresses
+func TestRedirectAs_eBPF_IPv4(t *testing.T) {
+	wgPort := 51850
+	ebpfProxy := ebpf.NewWGEBPFProxy(wgPort, 1280)
+	if err := ebpfProxy.Listen(); err != nil {
+		t.Fatalf("failed to initialize ebpf proxy: %v", err)
+	}
+	defer func() {
+		if err := ebpfProxy.Free(); err != nil {
+			t.Errorf("failed to free ebpf proxy: %v", err)
+		}
+	}()
+
+	proxy := ebpf.NewProxyWrapper(ebpfProxy)
+
+	// NetBird UDP address of the remote peer
+	nbAddr := &net.UDPAddr{
+		IP:   net.ParseIP("100.108.111.177"),
+		Port: 38746,
+	}
+
+	p2pEndpoint := &net.UDPAddr{
+		IP:   net.ParseIP("192.168.0.56"),
+		Port: 51820,
+	}
+
+	testRedirectAs(t, proxy, wgPort, nbAddr, p2pEndpoint)
+}
+
+// TestRedirectAs_eBPF_IPv6 tests RedirectAs with eBPF proxy using IPv6 addresses
+func TestRedirectAs_eBPF_IPv6(t *testing.T) {
+	wgPort := 51851
+	ebpfProxy := ebpf.NewWGEBPFProxy(wgPort, 1280)
+	if err := ebpfProxy.Listen(); err != nil {
+		t.Fatalf("failed to initialize ebpf proxy: %v", err)
+	}
+	defer func() {
+		if err := ebpfProxy.Free(); err != nil {
+			t.Errorf("failed to free ebpf proxy: %v", err)
+		}
+	}()
+
+	proxy := ebpf.NewProxyWrapper(ebpfProxy)
+
+	// NetBird UDP address of the remote peer
+	nbAddr := &net.UDPAddr{
+		IP:   net.ParseIP("100.108.111.177"),
+		Port: 38746,
+	}
+
+	p2pEndpoint := &net.UDPAddr{
+		IP:   net.ParseIP("fe80::56"),
+		Port: 51820,
+	}
+
+	testRedirectAs(t, proxy, wgPort, nbAddr, p2pEndpoint)
+}
+
 // TestRedirectAs_UDP_IPv4 tests RedirectAs with UDP proxy using IPv4 addresses
 func TestRedirectAs_UDP_IPv4(t *testing.T) {
 	wgPort := 51852
@@ -198,64 +256,6 @@ func testRedirectAs(t *testing.T, proxy Proxy, wgPort int, nbAddr, p2pEndpoint *
 	}
 }
 
-// TestRedirectAs_eBPF_IPv4 tests RedirectAs with eBPF proxy using IPv4 addresses
-func TestRedirectAs_eBPF_IPv4(t *testing.T) {
-	wgPort := 51850
-	ebpfProxy := ebpf.NewWGEBPFProxy(wgPort, 1280)
-	if err := ebpfProxy.Listen(); err != nil {
-		t.Fatalf("failed to initialize ebpf proxy: %v", err)
-	}
-	defer func() {
-		if err := ebpfProxy.Free(); err != nil {
-			t.Errorf("failed to free ebpf proxy: %v", err)
-		}
-	}()
-
-	proxy := ebpf.NewProxyWrapper(ebpfProxy)
-
-	// NetBird UDP address of the remote peer
-	nbAddr := &net.UDPAddr{
-		IP:   net.ParseIP("100.108.111.177"),
-		Port: 38746,
-	}
-
-	p2pEndpoint := &net.UDPAddr{
-		IP:   net.ParseIP("192.168.0.56"),
-		Port: 51820,
-	}
-
-	testRedirectAs(t, proxy, wgPort, nbAddr, p2pEndpoint)
-}
-
-// TestRedirectAs_eBPF_IPv6 tests RedirectAs with eBPF proxy using IPv6 addresses
-func TestRedirectAs_eBPF_IPv6(t *testing.T) {
-	wgPort := 51851
-	ebpfProxy := ebpf.NewWGEBPFProxy(wgPort, 1280)
-	if err := ebpfProxy.Listen(); err != nil {
-		t.Fatalf("failed to initialize ebpf proxy: %v", err)
-	}
-	defer func() {
-		if err := ebpfProxy.Free(); err != nil {
-			t.Errorf("failed to free ebpf proxy: %v", err)
-		}
-	}()
-
-	proxy := ebpf.NewProxyWrapper(ebpfProxy)
-
-	// NetBird UDP address of the remote peer
-	nbAddr := &net.UDPAddr{
-		IP:   net.ParseIP("100.108.111.177"),
-		Port: 38746,
-	}
-
-	p2pEndpoint := &net.UDPAddr{
-		IP:   net.ParseIP("fe80::56"),
-		Port: 51820,
-	}
-
-	testRedirectAs(t, proxy, wgPort, nbAddr, p2pEndpoint)
-}
-
 // TestRedirectAs_Multiple_Switches tests switching between multiple endpoints
 func TestRedirectAs_Multiple_Switches(t *testing.T) {
 	wgPort := 51856

client/iface/wgproxy/udp/proxy.go

@@ -147,17 +147,6 @@ func (p *WGUDPProxy) RedirectAs(endpoint *net.UDPAddr) {
 	p.sendPkg = p.srcFakerConn.SendPkg
 }
 
-// InjectPacket writes b to the remote peer over the underlying transport.
-func (p *WGUDPProxy) InjectPacket(b []byte) error {
-	if p.remoteConn == nil {
-		return errors.New("proxy not started")
-	}
-	if _, err := p.remoteConn.Write(b); err != nil {
-		return err
-	}
-	return nil
-}
-
 // CloseConn close the localConn
 func (p *WGUDPProxy) CloseConn() error {
 	if p.cancel == nil {

client/internal/acl/manager.go

@@ -11,7 +11,6 @@ import (
 	"time"
 
 	"github.com/hashicorp/go-multierror"
-	"github.com/mitchellh/hashstructure/v2"
 	log "github.com/sirupsen/logrus"
 
 	nberrors "github.com/netbirdio/netbird/client/errors"
@@ -31,13 +30,11 @@ type Manager interface {
 
 // DefaultManager uses firewall manager to handle
 type DefaultManager struct {
-	firewall           firewall.Manager
-	ipsetCounter       int
-	peerRulesPairs     map[id.RuleID][]firewall.Rule
-	routeRules         map[id.RuleID]struct{}
-	previousConfigHash uint64
-	hasAppliedConfig   bool
-	mutex              sync.Mutex
+	firewall       firewall.Manager
+	ipsetCounter   int
+	peerRulesPairs map[id.RuleID][]firewall.Rule
+	routeRules     map[id.RuleID]struct{}
+	mutex          sync.Mutex
 }
 
 func NewDefaultManager(fm firewall.Manager) *DefaultManager {
@@ -60,23 +57,6 @@ func (d *DefaultManager) ApplyFiltering(networkMap *mgmProto.NetworkMap, dnsRout
 		return
 	}
 
-	// Skip the full rebuild + flush when the inputs that drive the firewall
-	// state are byte-for-byte identical to the last successfully applied
-	// update. Management re-sends the same network map far more often than it
-	// actually changes (account-wide updates, peer meta churn), and rebuilding
-	// every peer/route ACL and flushing the firewall on every such sync is the
-	// dominant client-side cost when nothing changed. Mirrors the same guard the
-	// DNS server already uses (previousConfigHash). Only the fields ApplyFiltering
-	// consumes participate in the hash, so an unrelated map change cannot mask a
-	// real ACL change.
-	hash, err := d.firewallConfigHash(networkMap, dnsRouteFeatureFlag)
-	if err != nil {
-		log.Errorf("unable to hash firewall configuration, applying unconditionally: %v", err)
-	} else if d.hasAppliedConfig && d.previousConfigHash == hash {
-		log.Debugf("not applying the firewall configuration update as there is nothing new (hash: %d)", hash)
-		return
-	}
-
 	start := time.Now()
 	defer func() {
 		total := 0
@@ -90,49 +70,13 @@ func (d *DefaultManager) ApplyFiltering(networkMap *mgmProto.NetworkMap, dnsRout
 
 	d.applyPeerACLs(networkMap)
 
-	routeErr := d.applyRouteACLs(networkMap.RoutesFirewallRules, dnsRouteFeatureFlag)
-	if routeErr != nil {
-		log.Errorf("Failed to apply route ACLs: %v", routeErr)
+	if err := d.applyRouteACLs(networkMap.RoutesFirewallRules, dnsRouteFeatureFlag); err != nil {
+		log.Errorf("Failed to apply route ACLs: %v", err)
 	}
 
-	flushErr := d.firewall.Flush()
-	if flushErr != nil {
-		log.Error("failed to flush firewall rules: ", flushErr)
+	if err := d.firewall.Flush(); err != nil {
+		log.Error("failed to flush firewall rules: ", err)
 	}
-
-	// Only remember the hash once the firewall actually reflects this config.
-	// If applying or flushing failed, leave the previous hash untouched so the
-	// next (possibly identical) update is not skipped and gets a chance to
-	// reconcile the firewall state.
-	if err == nil && routeErr == nil && flushErr == nil {
-		d.previousConfigHash = hash
-		d.hasAppliedConfig = true
-	} else {
-		d.hasAppliedConfig = false
-	}
-}
-
-// firewallConfigHash hashes exactly the inputs ApplyFiltering uses to build the
-// firewall state, so an identical hash means an identical resulting ruleset.
-func (d *DefaultManager) firewallConfigHash(networkMap *mgmProto.NetworkMap, dnsRouteFeatureFlag bool) (uint64, error) {
-	return hashstructure.Hash(struct {
-		PeerRules            []*mgmProto.FirewallRule
-		PeerRulesIsEmpty     bool
-		RouteRules           []*mgmProto.RouteFirewallRule
-		RouteRulesIsEmpty    bool
-		DNSRouteFeatureFlag  bool
-	}{
-		PeerRules:           networkMap.GetFirewallRules(),
-		PeerRulesIsEmpty:    networkMap.GetFirewallRulesIsEmpty(),
-		RouteRules:          networkMap.GetRoutesFirewallRules(),
-		RouteRulesIsEmpty:   networkMap.GetRoutesFirewallRulesIsEmpty(),
-		DNSRouteFeatureFlag: dnsRouteFeatureFlag,
-	}, hashstructure.FormatV2, &hashstructure.HashOptions{
-		ZeroNil:         true,
-		IgnoreZeroValue: true,
-		SlicesAsSets:    true,
-		UseStringer:     true,
-	})
 }
 
 func (d *DefaultManager) applyPeerACLs(networkMap *mgmProto.NetworkMap) {

client/internal/acl/manager_test.go

@@ -1,7 +1,6 @@
 package acl
 
 import (
-	"fmt"
 	"net/netip"
 	"testing"
 
@@ -486,149 +485,3 @@ func TestPortInfoEmpty(t *testing.T) {
 		})
 	}
 }
-
-// TestApplyFilteringSkipsUnchangedConfig verifies that an identical network map
-// re-applied is recognized as a no-op (hash unchanged), while a real change to
-// any firewall-relevant input forces a re-apply (hash changes). This is the
-// guard that prevents a full ruleset rebuild + flush on every redundant sync.
-func TestApplyFilteringSkipsUnchangedConfig(t *testing.T) {
-	t.Setenv("NB_WG_KERNEL_DISABLED", "true")
-	t.Setenv(firewall.EnvForceUserspaceFirewall, "true")
-
-	ctrl := gomock.NewController(t)
-	defer ctrl.Finish()
-
-	ifaceMock := mocks.NewMockIFaceMapper(ctrl)
-	ifaceMock.EXPECT().IsUserspaceBind().Return(true).AnyTimes()
-	ifaceMock.EXPECT().SetFilter(gomock.Any())
-	network := netip.MustParsePrefix("172.0.0.1/32")
-	ifaceMock.EXPECT().Name().Return("lo").AnyTimes()
-	ifaceMock.EXPECT().Address().Return(wgaddr.Address{
-		IP:      network.Addr(),
-		Network: network,
-	}).AnyTimes()
-	ifaceMock.EXPECT().GetWGDevice().Return(nil).AnyTimes()
-
-	fw, err := firewall.NewFirewall(ifaceMock, nil, flowLogger, false, iface.DefaultMTU)
-	require.NoError(t, err)
-	defer func() {
-		require.NoError(t, fw.Close(nil))
-	}()
-
-	acl := NewDefaultManager(fw)
-
-	networkMap := &mgmProto.NetworkMap{
-		FirewallRules: []*mgmProto.FirewallRule{
-			{
-				PeerIP:    "10.93.0.1",
-				Direction: mgmProto.RuleDirection_IN,
-				Action:    mgmProto.RuleAction_ACCEPT,
-				Protocol:  mgmProto.RuleProtocol_TCP,
-				Port:      "22",
-			},
-		},
-		FirewallRulesIsEmpty: false,
-	}
-
-	acl.ApplyFiltering(networkMap, false)
-	require.True(t, acl.hasAppliedConfig, "config should be marked applied after first apply")
-	firstHash := acl.previousConfigHash
-	require.NotZero(t, firstHash)
-
-	// Re-applying the identical map must not change the recorded hash: the
-	// expensive rebuild path was skipped.
-	acl.ApplyFiltering(networkMap, false)
-	assert.Equal(t, firstHash, acl.previousConfigHash,
-		"identical re-apply must be a no-op (hash unchanged)")
-
-	// A real change must produce a different hash and re-apply.
-	networkMap.FirewallRules[0].Action = mgmProto.RuleAction_DROP
-	acl.ApplyFiltering(networkMap, false)
-	assert.NotEqual(t, firstHash, acl.previousConfigHash,
-		"changing a rule's action must force a re-apply (hash changed)")
-
-	// The dnsRouteFeatureFlag also participates in the hash.
-	changedHash := acl.previousConfigHash
-	acl.ApplyFiltering(networkMap, true)
-	assert.NotEqual(t, changedHash, acl.previousConfigHash,
-		"flipping dnsRouteFeatureFlag must force a re-apply (hash changed)")
-}
-
-func buildNetworkMap(peerRules, routeRules int) *mgmProto.NetworkMap {
-	nm := &mgmProto.NetworkMap{
-		FirewallRulesIsEmpty:      peerRules == 0,
-		RoutesFirewallRulesIsEmpty: routeRules == 0,
-	}
-	for i := range peerRules {
-		nm.FirewallRules = append(nm.FirewallRules, &mgmProto.FirewallRule{
-			PeerIP:    fmt.Sprintf("10.%d.%d.%d", i>>16&0xff, i>>8&0xff, i&0xff),
-			Direction: mgmProto.RuleDirection_IN,
-			Action:    mgmProto.RuleAction_ACCEPT,
-			Protocol:  mgmProto.RuleProtocol_TCP,
-			Port:      fmt.Sprintf("%d", 1024+i%64511),
-		})
-	}
-	for i := range routeRules {
-		nm.RoutesFirewallRules = append(nm.RoutesFirewallRules, &mgmProto.RouteFirewallRule{
-			Destination:  fmt.Sprintf("192.168.%d.0/24", i%256),
-			SourceRanges: []string{fmt.Sprintf("10.0.%d.0/24", i%256)},
-			Action:       mgmProto.RuleAction_ACCEPT,
-			Protocol:     mgmProto.RuleProtocol_ALL,
-		})
-	}
-	return nm
-}
-
-func BenchmarkFirewallConfigHash_Small(b *testing.B) {
-	d := &DefaultManager{}
-	nm := buildNetworkMap(10, 5)
-	b.ResetTimer()
-	for b.Loop() {
-		_, _ = d.firewallConfigHash(nm, false)
-	}
-}
-
-func BenchmarkFirewallConfigHash_Medium(b *testing.B) {
-	d := &DefaultManager{}
-	nm := buildNetworkMap(100, 50)
-	b.ResetTimer()
-	for b.Loop() {
-		_, _ = d.firewallConfigHash(nm, false)
-	}
-}
-
-func BenchmarkFirewallConfigHash_Large(b *testing.B) {
-	d := &DefaultManager{}
-	nm := buildNetworkMap(1000, 200)
-	b.ResetTimer()
-	for b.Loop() {
-		_, _ = d.firewallConfigHash(nm, false)
-	}
-}
-
-// TestFirewallConfigHashDeterministic verifies the hash is stable for equal
-// inputs and order-independent for the rule slices (management does not
-// guarantee rule order).
-func TestFirewallConfigHashDeterministic(t *testing.T) {
-	d := &DefaultManager{}
-
-	nm1 := &mgmProto.NetworkMap{
-		FirewallRules: []*mgmProto.FirewallRule{
-			{PeerIP: "10.0.0.1", Direction: mgmProto.RuleDirection_IN, Action: mgmProto.RuleAction_ACCEPT, Protocol: mgmProto.RuleProtocol_TCP, Port: "22"},
-			{PeerIP: "10.0.0.2", Direction: mgmProto.RuleDirection_IN, Action: mgmProto.RuleAction_DROP, Protocol: mgmProto.RuleProtocol_TCP, Port: "80"},
-		},
-	}
-	// Same rules, reversed order.
-	nm2 := &mgmProto.NetworkMap{
-		FirewallRules: []*mgmProto.FirewallRule{
-			nm1.FirewallRules[1],
-			nm1.FirewallRules[0],
-		},
-	}
-
-	h1, err := d.firewallConfigHash(nm1, false)
-	require.NoError(t, err)
-	h2, err := d.firewallConfigHash(nm2, false)
-	require.NoError(t, err)
-	assert.Equal(t, h1, h2, "hash must be order-independent for rule slices")
-}

client/internal/connect.go

@@ -11,7 +11,6 @@ import (
 	"runtime/debug"
 	"strings"
 	"sync"
-	"sync/atomic"
 	"time"
 
 	"github.com/cenkalti/backoff/v4"
@@ -19,6 +18,7 @@ import (
 
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
 	"google.golang.org/grpc/codes"
+	"google.golang.org/grpc/metadata"
 	gstatus "google.golang.org/grpc/status"
 
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
@@ -49,17 +49,23 @@ import (
 	"github.com/netbirdio/netbird/version"
 )
 
-// androidRunOverride is set on Android to inject mobile dependencies
-// when using embed.Client (which calls Run() with empty MobileDependency).
-var androidRunOverride func(c *ConnectClient, runningChan chan struct{}, logPath string) error
+// androidMobileDep is set on Android to inject the MobileDependency for runs
+// started through the generic entry points (Run/RunAsync, e.g. embed.Client).
+// nil on other platforms, where the dependency is empty.
+var androidMobileDep func(config *profilemanager.Config) MobileDependency
+
+// mobileDependency returns the MobileDependency for a run started via the
+// generic entry points. On Android the androidMobileDep provider supplies
+// platform stubs (or real implementations); elsewhere it is empty.
+func (c *ConnectClient) mobileDependency(config *profilemanager.Config) MobileDependency {
+	if androidMobileDep != nil {
+		return androidMobileDep(config)
+	}
+	return MobileDependency{}
+}
 
 type ConnectClient struct {
 	ctx            context.Context
-	runCancel      context.CancelFunc
-	runExited      chan struct{}
-	runOnce        sync.Once
-	runStarted     atomic.Bool
-	config         *profilemanager.Config
 	statusRecorder *peer.Status
 
 	engine        *Engine
@@ -68,41 +74,62 @@ type ConnectClient struct {
 	updateManager *updater.Manager
 
 	persistSyncResponse bool
+
+	// sup serializes all start/stop requests so two lifecycle operations can
+	// never overlap. See connect_lifecycle.go.
+	sup *supervisor
 }
 
 func NewConnectClient(
 	ctx context.Context,
-	config *profilemanager.Config,
 	statusRecorder *peer.Status,
 ) *ConnectClient {
-	// Derive the run context here so Stop owns the cancel that unblocks the run
-	// loop. runCancel is set once at construction, so Stop can call it without
-	// racing the run loop's startup. Callers therefore need not cancel before Stop.
-	runCtx, runCancel := context.WithCancel(ctx)
-	return &ConnectClient{
-		ctx:            runCtx,
-		runCancel:      runCancel,
-		runExited:      make(chan struct{}),
-		config:         config,
+	c := &ConnectClient{
+		ctx:            ctx,
 		statusRecorder: statusRecorder,
 		engineMutex:    sync.Mutex{},
 	}
+	c.sup = newSupervisor(ctx, c.run)
+	return c
 }
 
 func (c *ConnectClient) SetUpdateManager(um *updater.Manager) {
 	c.updateManager = um
 }
 
-// Run with main logic.
-func (c *ConnectClient) Run(runningChan chan struct{}, logPath string) error {
-	if androidRunOverride != nil {
-		return androidRunOverride(c, runningChan, logPath)
-	}
-	return c.run(MobileDependency{}, runningChan, logPath)
+// Run with main logic. md carries optional gRPC metadata (e.g. the UI
+// user-agent) to forward to the management/signal services; nil when none.
+func (c *ConnectClient) Run(config *profilemanager.Config, md metadata.MD, logPath string) error {
+	return c.sup.start(config, md, c.mobileDependency(config), logPath)
+}
+
+// RunAsync starts a client run without blocking. Used by the daemon and embed,
+// which drive the lifecycle through the supervisor rather than blocking on Run;
+// they then wait for the outcome via WaitEstablishedOrDone. The run's lifecycle
+// channels are created and owned by the supervisor — callers never hold them.
+func (c *ConnectClient) RunAsync(config *profilemanager.Config, md metadata.MD) {
+	c.sup.startAsync(config, md, c.mobileDependency(config), "", nil)
+}
+
+// Restart atomically stops any in-flight run and starts a fresh one with the
+// given config. The stop+start happens as a single supervisor operation, so no
+// other lifecycle request can interleave between them — used for explicit
+// restarts (e.g. an MDM policy change) that must not expose a "stopped" window.
+func (c *ConnectClient) Restart(config *profilemanager.Config, md metadata.MD) {
+	c.sup.restartAsync(config, md, c.mobileDependency(config), "")
+}
+
+// WaitEstablishedOrDone blocks until the in-flight run becomes established (nil),
+// ends before that (the run error, or a sentinel on a clean stop), or ctx is
+// cancelled. Returns errNoRunInFlight if no run is in flight. Wraps the wait on
+// the supervisor-owned channels so callers never touch them directly.
+func (c *ConnectClient) WaitEstablishedOrDone(ctx context.Context) error {
+	return c.sup.waitEstablishedOrDone(ctx)
 }
 
 // RunOnAndroid with main logic on mobile system
 func (c *ConnectClient) RunOnAndroid(
+	config *profilemanager.Config,
 	tunAdapter device.TunAdapter,
 	iFaceDiscover stdnet.ExternalIFaceDiscover,
 	networkChangeListener listener.NetworkChangeListener,
@@ -121,10 +148,11 @@ func (c *ConnectClient) RunOnAndroid(
 		StateFilePath:         stateFilePath,
 		TempDir:               cacheDir,
 	}
-	return c.run(mobileDependency, nil, "")
+	return c.sup.start(config, nil, mobileDependency, "")
 }
 
 func (c *ConnectClient) RunOniOS(
+	config *profilemanager.Config,
 	fileDescriptor int32,
 	networkChangeListener listener.NetworkChangeListener,
 	dnsManager dns.IosDnsManager,
@@ -142,15 +170,12 @@ func (c *ConnectClient) RunOniOS(
 		StateFilePath:         stateFilePath,
 		TempDir:               cacheDir,
 	}
-	return c.run(mobileDependency, nil, logFilePath)
+	return c.sup.start(config, nil, mobileDependency, logFilePath)
 }
 
-func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan struct{}, logPath string) error {
-	// Mark the loop as started and signal exit on return so Stop can wait for
-	// the loop to finish (and skip the wait if the loop never ran).
-	c.runStarted.Store(true)
-	defer c.runOnce.Do(func() { close(c.runExited) })
-
+// run executes a single client run. runCtx is owned by the supervisor: cancelling
+// it tears the run down (it is the parent of the per-attempt engine context).
+func (c *ConnectClient) run(runCtx context.Context, config *profilemanager.Config, mobileDependency MobileDependency, connEstablishedChan chan struct{}, logPath string) error {
 	defer func() {
 		if r := recover(); r != nil {
 			rec := c.statusRecorder
@@ -214,18 +239,18 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 	}()
 
 	wrapErr := state.Wrap
-	myPrivateKey, err := wgtypes.ParseKey(c.config.PrivateKey)
+	myPrivateKey, err := wgtypes.ParseKey(config.PrivateKey)
 	if err != nil {
-		log.Errorf("failed parsing Wireguard key %s: [%s]", c.config.PrivateKey, err.Error())
+		log.Errorf("failed parsing Wireguard key %s: [%s]", config.PrivateKey, err.Error())
 		return wrapErr(err)
 	}
 
 	var mgmTlsEnabled bool
-	if c.config.ManagementURL.Scheme == "https" {
+	if config.ManagementURL.Scheme == "https" {
 		mgmTlsEnabled = true
 	}
 
-	publicSSHKey, err := ssh.GeneratePublicKey([]byte(c.config.SSHKey))
+	publicSSHKey, err := ssh.GeneratePublicKey([]byte(config.SSHKey))
 	if err != nil {
 		return err
 	}
@@ -259,13 +284,13 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 	defer c.statusRecorder.ClientStop()
 	operation := func() error {
 		// if context cancelled we not start new backoff cycle
-		if c.ctx.Err() != nil {
+		if runCtx.Err() != nil {
 			return nil
 		}
 
 		state.Set(StatusConnecting)
 
-		engineCtx, cancel := context.WithCancel(c.ctx)
+		engineCtx, cancel := context.WithCancel(runCtx)
 		defer func() {
 			_, err := state.Status()
 			c.statusRecorder.MarkManagementDisconnected(err)
@@ -273,8 +298,8 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 			cancel()
 		}()
 
-		log.Debugf("connecting to the Management service %s", c.config.ManagementURL.Host)
-		mgmClient, err := mgm.NewClient(engineCtx, c.config.ManagementURL.Host, myPrivateKey, mgmTlsEnabled)
+		log.Debugf("connecting to the Management service %s", config.ManagementURL.Host)
+		mgmClient, err := mgm.NewClient(engineCtx, config.ManagementURL.Host, myPrivateKey, mgmTlsEnabled)
 		if err != nil {
 			return wrapErr(gstatus.Errorf(codes.FailedPrecondition, "failed connecting to Management Service : %s", err))
 		}
@@ -291,7 +316,7 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		}
 		c.clientMetrics.UpdateAgentInfo(agentInfo, myPrivateKey.PublicKey().String())
 
-		log.Debugf("connected to the Management service %s", c.config.ManagementURL.Host)
+		log.Debugf("connected to the Management service %s", config.ManagementURL.Host)
 		defer func() {
 			if err = mgmClient.Close(); err != nil {
 				log.Warnf("failed to close the Management service client %v", err)
@@ -300,13 +325,14 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 
 		// connect (just a connection, no stream yet) and login to Management Service to get an initial global Netbird config
 		loginStarted := time.Now()
-		loginResp, err := loginToManagement(engineCtx, mgmClient, publicSSHKey, c.config)
+		loginResp, err := loginToManagement(engineCtx, mgmClient, publicSSHKey, config)
 		if err != nil {
 			c.clientMetrics.RecordLoginDuration(engineCtx, time.Since(loginStarted), false)
 			log.Debug(err)
 			if s, ok := gstatus.FromError(err); ok && (s.Code() == codes.PermissionDenied) {
 				state.Set(StatusNeedsLogin)
-				c.runCancel()
+				// No teardown needed: login fails before the engine is started
+				// (engine.Start is below), so there is nothing running to stop.
 				return backoff.Permanent(wrapErr(err)) // unrecoverable error
 			}
 			return wrapErr(err)
@@ -314,10 +340,6 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		c.clientMetrics.RecordLoginDuration(engineCtx, time.Since(loginStarted), true)
 		c.statusRecorder.MarkManagementConnected()
 
-		if metricsConfig := loginResp.GetNetbirdConfig().GetMetrics(); metricsConfig != nil {
-			c.clientMetrics.UpdatePushFromMgm(c.ctx, metricsConfig.GetEnabled())
-		}
-
 		localPeerState := peer.LocalPeerState{
 			IP:              loginResp.GetPeerConfig().GetAddress(),
 			PubKey:          myPrivateKey.PublicKey().String(),
@@ -364,7 +386,7 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		}
 		peerConfig := loginResp.GetPeerConfig()
 
-		engineConfig, err := createEngineConfig(myPrivateKey, c.config, peerConfig, logPath)
+		engineConfig, err := createEngineConfig(myPrivateKey, config, peerConfig, logPath)
 		if err != nil {
 			log.Error(err)
 			return wrapErr(err)
@@ -403,13 +425,12 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 			StateManager:   stateManager,
 			UpdateManager:  c.updateManager,
 			ClientMetrics:  c.clientMetrics,
-			MetricsCtx:     c.ctx,
 		}, mobileDependency)
 		engine.SetSyncResponsePersistence(c.persistSyncResponse)
 		c.engine = engine
 		c.engineMutex.Unlock()
 
-		if err := engine.Start(loginResp.GetNetbirdConfig(), c.config.ManagementURL); err != nil {
+		if err := engine.Start(loginResp.GetNetbirdConfig(), config.ManagementURL); err != nil {
 			log.Errorf("error while starting Netbird Connection Engine: %s", err)
 			return wrapErr(err)
 		}
@@ -417,12 +438,13 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		log.Infof("Netbird engine started, the IP is: %s", peerConfig.GetAddress())
 		state.Set(StatusConnected)
 
-		if runningChan != nil {
-			select {
-			case <-runningChan:
-			default:
-				close(runningChan)
-			}
+		// The supervisor owns connEstablishedChan and it is always present. Guard
+		// against a double close: operation re-runs on ErrResetConnection retries
+		// within the same run, and the channel is closed only on the first connect.
+		select {
+		case <-connEstablishedChan:
+		default:
+			close(connEstablishedChan)
 		}
 
 		<-engineCtx.Done()
@@ -431,8 +453,10 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		c.engine = nil
 		c.engineMutex.Unlock()
 
-		log.Infof("ensuring wg interface is removed, Netbird engine context cancelled")
-
+		// Always tear the engine down once its context is cancelled. engine.Stop
+		// is nil-guarded per component, so calling it unconditionally is safe and
+		// avoids both the data race on engine.wgInterface and skipping teardown
+		// when the interface was never brought up (e.g. a mid-start failure).
 		if err := engine.Stop(); err != nil {
 			log.Errorf("Failed to stop engine: %v", err)
 		}
@@ -450,12 +474,13 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 	}
 
 	c.statusRecorder.ClientStart()
-	err = backoff.Retry(operation, backoff.WithContext(backOff, c.ctx))
+	err = backoff.Retry(operation, backOff)
 	if err != nil {
 		log.Debugf("exiting client retry loop due to unrecoverable error: %s", err)
 		if s, ok := gstatus.FromError(err); ok && (s.Code() == codes.PermissionDenied) {
+			// Login failed permanently: the engine was never started, so there
+			// is nothing to tear down — just record that a login is needed.
 			state.Set(StatusNeedsLogin)
-			c.runCancel()
 		}
 		return err
 	}
@@ -476,6 +501,22 @@ func parseRelayInfo(loginResp *mgmProto.LoginResponse) ([]string, *hmac.Token) {
 	return relayCfg.GetUrls(), token
 }
 
+// ConnectionRunning reports whether a connection run is currently in flight
+// (connecting, connected, or reconnecting). Answered by the supervisor via a
+// serialized query, so it settles behind an in-flight stop. Distinct from
+// ServiceRunning, which reports whether the service itself is alive.
+func (c *ConnectClient) ConnectionRunning() bool {
+	return c.sup.isRunning()
+}
+
+// ServiceRunning reports whether the client's lifecycle supervisor is alive and
+// able to accept start/stop commands — i.e. its context has not been cancelled
+// (the daemon is not shutting down). Independent of whether a connection run is
+// up (that is ConnectionRunning).
+func (c *ConnectClient) ServiceRunning() bool {
+	return c.sup.ctx.Err() == nil
+}
+
 func (c *ConnectClient) Engine() *Engine {
 	if c == nil {
 		return nil
@@ -532,12 +573,10 @@ func (c *ConnectClient) Status() StatusType {
 	return status
 }
 
+// Stop serializes a stop request through the lifecycle supervisor and blocks
+// until the in-flight run is fully torn down.
 func (c *ConnectClient) Stop() error {
-	c.runCancel()
-	if c.runStarted.Load() {
-		<-c.runExited
-	}
-	return nil
+	return c.sup.stop()
 }
 
 // SetSyncResponsePersistence enables or disables sync response persistence.

client/internal/connect_android_default.go

@@ -7,6 +7,7 @@ import (
 
 	"github.com/netbirdio/netbird/client/internal/dns"
 	"github.com/netbirdio/netbird/client/internal/listener"
+	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	"github.com/netbirdio/netbird/client/internal/stdnet"
 )
 
@@ -59,19 +60,17 @@ var _ listener.NetworkChangeListener = noopNetworkChangeListener{}
 var _ dns.ReadyListener = noopDnsReadyListener{}
 
 func init() {
-	// Wire up the default override so embed.Client.Start() works on Android
-	// with netstack mode. Provides complete no-op stubs for all mobile
+	// Wire up the default MobileDependency provider so embed.Client.Start() works
+	// on Android with netstack mode. Provides complete no-op stubs for all mobile
 	// dependencies so the engine's existing Android code paths work unchanged.
-	// Applications that need P2P ICE or real DNS should replace this by
-	// setting androidRunOverride before calling Start().
-	androidRunOverride = func(c *ConnectClient, runningChan chan struct{}, logPath string) error {
-		return c.runOnAndroidEmbed(
+	// Applications that need P2P ICE or real DNS should replace this by setting
+	// androidMobileDep before calling Start().
+	androidMobileDep = func(config *profilemanager.Config) MobileDependency {
+		return mobileDependencyForEmbed(
 			noopIFaceDiscover{},
 			noopNetworkChangeListener{},
 			[]netip.AddrPort{},
 			noopDnsReadyListener{},
-			runningChan,
-			logPath,
 		)
 	}
 }

client/internal/connect_android_embed.go

@@ -10,23 +10,18 @@ import (
 	"github.com/netbirdio/netbird/client/internal/stdnet"
 )
 
-// runOnAndroidEmbed is like RunOnAndroid but accepts a runningChan
-// so embed.Client.Start() can detect when the engine is ready.
-// It provides complete MobileDependency so the engine's existing
-// Android code paths work unchanged.
-func (c *ConnectClient) runOnAndroidEmbed(
+// mobileDependencyForEmbed builds the MobileDependency used by embed.Client on
+// Android so the engine's existing Android code paths work unchanged.
+func mobileDependencyForEmbed(
 	iFaceDiscover stdnet.ExternalIFaceDiscover,
 	networkChangeListener listener.NetworkChangeListener,
 	dnsAddresses []netip.AddrPort,
 	dnsReadyListener dns.ReadyListener,
-	runningChan chan struct{},
-	logPath string,
-) error {
-	mobileDependency := MobileDependency{
+) MobileDependency {
+	return MobileDependency{
 		IFaceDiscover:         iFaceDiscover,
 		NetworkChangeListener: networkChangeListener,
 		HostDNSAddresses:      dnsAddresses,
 		DnsReadyListener:      dnsReadyListener,
 	}
-	return c.run(mobileDependency, runningChan, logPath)
 }

client/internal/connect_lifecycle.go

@@ -0,0 +1,362 @@
+package internal
+
+import (
+	"context"
+	"errors"
+
+	"google.golang.org/grpc/metadata"
+
+	"github.com/netbirdio/netbird/client/internal/profilemanager"
+)
+
+// errAlreadyRunning is returned when a start is requested while a run is already
+// in flight.
+var errAlreadyRunning = errors.New("client is already running")
+
+// errNoRunInFlight is returned by waitEstablishedOrDone when no run is active.
+var errNoRunInFlight = errors.New("no connection run in flight")
+
+// errStoppedBeforeEstablished is returned when a run ended (cleanly) before the
+// connection was established.
+var errStoppedBeforeEstablished = errors.New("run stopped before the connection was established")
+
+// lifecycleOp is a serialized lifecycle operation processed by the supervisor.
+type lifecycleOp int
+
+const (
+	opStart lifecycleOp = iota
+	opStop
+	opRestart
+	opStatus
+	opWaitEstablished
+)
+
+// lifecycleCmd is a single lifecycle request handed to the supervisor goroutine.
+// They all flow through the same cmdCh so they are strictly ordered (FIFO) with
+// respect to each other.
+type lifecycleCmd struct {
+	op        lifecycleOp
+	config    *profilemanager.Config
+	md        metadata.MD
+	mobileDep MobileDependency
+	logPath   string
+
+	// done is the caller's notification channel (nil for fire-and-forget). Its
+	// meaning depends on op:
+	//   - opStart: receives the run's end result when the run terminates, or
+	//     errAlreadyRunning immediately if a run is already in flight.
+	//   - opStop: receives nil once the in-flight run has fully unwound.
+	//   - opWaitEstablished: receives the wait outcome (see waitEstablishedOrDone).
+	done chan error
+
+	reply   chan bool       // opStatus only: receives whether a run is in flight
+	waitCtx context.Context // opWaitEstablished only: the waiter's cancellation context
+}
+
+// runState holds the lifecycle channels of a single in-flight run, owned by the
+// loop goroutine. It never escapes the supervisor as an API; the only readers
+// are the per-wait goroutines the loop spawns for opWaitEstablished.
+//
+// connEstablishedChan is closed by the run once the connection is established.
+// The supervisor creates and owns it — callers no longer supply it; they observe
+// it through waitEstablishedOrDone. ended is closed (broadcast) when the run
+// terminates, so any number of waiters can observe it; err is the run's end
+// result, valid only after ended is closed.
+type runState struct {
+	connEstablishedChan chan struct{} // closed by the run on established
+	ended               chan struct{} // closed by finishRun when the run terminates
+	err                 error         // run end result, valid after ended is closed
+}
+
+// runEndResult is sent by the run goroutine to the supervisor when a run ends,
+// whether on its own (error / external context cancellation) or because of a Stop.
+type runEndResult struct {
+	err error
+}
+
+// runFunc executes a single client run bound to the supervisor-owned context,
+// with the config supplied by the start request.
+type runFunc func(ctx context.Context, config *profilemanager.Config, mobileDep MobileDependency, connEstablishedChan chan struct{}, logPath string) error
+
+// supervisor serializes start/stop of a single client run. Every request goes
+// through cmdCh and is handled one at a time by the loop goroutine, so two
+// lifecycle operations can never overlap and their order is preserved (FIFO).
+// The loop goroutine is the sole owner of curStart/runCancel, so that state
+// needs no locking. The loop exits when the parent context is cancelled.
+type supervisor struct {
+	ctx      context.Context
+	run      runFunc
+	cmdCh    chan lifecycleCmd
+	runEnded chan runEndResult
+
+	// owned exclusively by the loop goroutine. curStart is the in-flight start
+	// command (nil = idle); its done channel is notified when the run ends.
+	// curRun holds that run's lifecycle channels; runCancel cancels it.
+	curStart  *lifecycleCmd
+	curRun    *runState
+	runCancel context.CancelFunc
+}
+
+func newSupervisor(ctx context.Context, run runFunc) *supervisor {
+	s := &supervisor{
+		ctx:      ctx,
+		run:      run,
+		cmdCh:    make(chan lifecycleCmd, 16),
+		runEnded: make(chan runEndResult, 1),
+	}
+	go s.loop()
+	return s
+}
+
+func (s *supervisor) loop() {
+	for {
+		select {
+		case <-s.ctx.Done():
+			s.shutdown()
+			return
+		case cmd := <-s.cmdCh:
+			switch cmd.op {
+			case opStart:
+				s.handleStart(cmd)
+			case opStop:
+				s.handleStop(cmd)
+			case opRestart:
+				s.handleRestart(cmd)
+			case opStatus:
+				cmd.reply <- (s.isRunningInternal())
+			case opWaitEstablished:
+				s.handleWaitEstablished(cmd)
+			}
+		case res := <-s.runEnded:
+			// Run ended on its own, without an explicit Stop.
+			s.finishRun(res.err)
+		}
+	}
+}
+
+func (s *supervisor) handleStart(cmd lifecycleCmd) {
+	if s.isRunningInternal() {
+		notify(cmd.done, errAlreadyRunning)
+		return
+	}
+
+	runCtx, cancel := context.WithCancel(s.ctx)
+	if cmd.md != nil {
+		// Carry caller-supplied gRPC metadata (e.g. UI user-agent) into the run
+		// context so the engine's management/signal calls forward it. The cancel
+		// still drives runCtx (metadata wrapping preserves cancellation).
+		runCtx = metadata.NewOutgoingContext(runCtx, cmd.md)
+	}
+	s.runCancel = cancel
+	s.curStart = &cmd
+	s.curRun = &runState{connEstablishedChan: make(chan struct{}), ended: make(chan struct{})}
+
+	go func(ctx context.Context, cfg *profilemanager.Config, m MobileDependency, established chan struct{}, lp string) {
+		err := s.run(ctx, cfg, m, established, lp)
+		s.runEnded <- runEndResult{err: err}
+	}(runCtx, cmd.config, cmd.mobileDep, s.curRun.connEstablishedChan, cmd.logPath)
+}
+
+func (s *supervisor) handleStop(cmd lifecycleCmd) {
+	if !s.isRunningInternal() {
+		notify(cmd.done, nil)
+		return
+	}
+	s.stopCurrentRun()
+	notify(cmd.done, nil)
+}
+
+// handleRestart tears down any in-flight run and starts a fresh one in a single
+// loop turn. No other command can interleave between the stop and the start
+// (the loop is single-threaded), so the swap is atomic without relying on any
+// daemon-side lock — that is what an explicit restart (e.g. MDM config change)
+// needs to avoid a window where the client is observably stopped.
+func (s *supervisor) handleRestart(cmd lifecycleCmd) {
+	if s.isRunningInternal() {
+		s.stopCurrentRun()
+	}
+	s.handleStart(cmd)
+}
+
+// stopCurrentRun cancels the in-flight run and blocks the supervisor until it
+// has fully unwound, so the next action starts from a clean slate. The run
+// goroutine reports completion via runEnded. Caller must hold an in-flight run
+// (curStart != nil).
+func (s *supervisor) stopCurrentRun() {
+	s.runCancel()
+	res := <-s.runEnded
+	s.finishRun(res.err)
+}
+
+// finishRun resets lifecycle state after a run terminates and hands the run
+// error back to whoever asked to be notified of the start.
+func (s *supervisor) finishRun(err error) {
+	s.runCancel = nil
+	if s.isRunningInternal() {
+		// Publish the result to the broadcast channel before nil-ing curRun, so
+		// any opWaitEstablished goroutines blocked on ended observe err.
+		s.curRun.err = err
+		close(s.curRun.ended)
+		s.curRun = nil
+
+		notify(s.curStart.done, err)
+		s.curStart = nil
+	}
+}
+
+// handleWaitEstablished answers an opWaitEstablished request. The select itself
+// runs in a spawned goroutine on the run's channels so it never blocks the loop;
+// the loop only snapshots the in-flight run's channels (which it owns) here.
+func (s *supervisor) handleWaitEstablished(cmd lifecycleCmd) {
+	caller := cmd.done
+	if !s.isRunningInternal() {
+		notify(caller, errNoRunInFlight)
+		return
+	}
+	rs := s.curRun
+	established := rs.connEstablishedChan
+	ctx := cmd.waitCtx
+	go func() {
+		select {
+		case <-established:
+			notify(caller, nil)
+		case <-rs.ended:
+			if rs.err != nil {
+				notify(caller, rs.err)
+				return
+			}
+			notify(caller, errStoppedBeforeEstablished)
+		case <-ctx.Done():
+			notify(caller, ctx.Err())
+		}
+	}()
+}
+
+// shutdown tears down the in-flight run when the parent context is cancelled,
+// then fails any still-queued commands so their callers never hang.
+func (s *supervisor) shutdown() {
+	if s.runCancel != nil {
+		s.runCancel()
+		res := <-s.runEnded
+		s.finishRun(res.err)
+	}
+	for {
+		select {
+		case cmd := <-s.cmdCh:
+			notify(cmd.done, s.ctx.Err())
+		default:
+			return
+		}
+	}
+}
+
+// startAsync enqueues a start without blocking. If done is non-nil it receives
+// the run's end result (or errAlreadyRunning on rejection, or the context error
+// on shutdown).
+func (s *supervisor) startAsync(config *profilemanager.Config, md metadata.MD, mobileDep MobileDependency, logPath string, done chan error) {
+	cmd := lifecycleCmd{op: opStart, config: config, md: md, mobileDep: mobileDep, logPath: logPath, done: done}
+	select {
+	case s.cmdCh <- cmd:
+	case <-s.ctx.Done():
+		notify(done, s.ctx.Err())
+	}
+}
+
+// restartAsync enqueues an atomic stop+start without blocking. The supervisor
+// tears down any in-flight run and starts a fresh one with the supplied config
+// in a single loop turn (see handleRestart). Fire-and-forget: the new run owns
+// its lifecycle channels, observed via waitEstablishedOrDone.
+func (s *supervisor) restartAsync(config *profilemanager.Config, md metadata.MD, mobileDep MobileDependency, logPath string) {
+	cmd := lifecycleCmd{op: opRestart, config: config, md: md, mobileDep: mobileDep, logPath: logPath}
+	select {
+	case s.cmdCh <- cmd:
+	case <-s.ctx.Done():
+	}
+}
+
+// start enqueues a start and blocks until the run terminates, preserving the
+// blocking contract of the legacy Run entry points.
+func (s *supervisor) start(config *profilemanager.Config, md metadata.MD, mobileDep MobileDependency, logPath string) error {
+	done := make(chan error, 1)
+	s.startAsync(config, md, mobileDep, logPath, done)
+	select {
+	case err := <-done:
+		return err
+	case <-s.ctx.Done():
+		return s.ctx.Err()
+	}
+}
+
+// isRunning asks the loop whether a run is in flight. The query is serialized
+// with start/stop, so during a stop it waits for the teardown to settle and
+// then reports the final state — never a transient "half-stopped".
+func (s *supervisor) isRunning() bool {
+	reply := make(chan bool, 1)
+	select {
+	case s.cmdCh <- lifecycleCmd{op: opStatus, reply: reply}:
+	case <-s.ctx.Done():
+		return false
+	}
+	select {
+	case r := <-reply:
+		return r
+	case <-s.ctx.Done():
+		return false
+	}
+}
+
+func (s *supervisor) isRunningInternal() bool {
+	return s.curStart != nil
+}
+
+// waitEstablishedOrDone blocks until the in-flight run becomes established
+// (returns nil) or ends before that (returns the run error, or
+// errStoppedBeforeEstablished on a clean stop), or ctx is cancelled. Returns
+// errNoRunInFlight if no run is in flight. The wait is performed by a goroutine
+// spawned inside the loop (see handleWaitEstablished); the run's channels never
+// leave the supervisor.
+func (s *supervisor) waitEstablishedOrDone(ctx context.Context) error {
+	reply := make(chan error, 1)
+	select {
+	case s.cmdCh <- lifecycleCmd{op: opWaitEstablished, waitCtx: ctx, done: reply}:
+	case <-ctx.Done():
+		return ctx.Err()
+	case <-s.ctx.Done():
+		return s.ctx.Err()
+	}
+	select {
+	case err := <-reply:
+		return err
+	case <-s.ctx.Done():
+		return s.ctx.Err()
+	}
+}
+
+// stop enqueues a stop and blocks until the in-flight run is fully torn down.
+func (s *supervisor) stop() error {
+	done := make(chan error, 1)
+	select {
+	case s.cmdCh <- lifecycleCmd{op: opStop, done: done}:
+	case <-s.ctx.Done():
+		return s.ctx.Err()
+	}
+	select {
+	case err := <-done:
+		return err
+	case <-s.ctx.Done():
+		return s.ctx.Err()
+	}
+}
+
+// notify sends on a caller-supplied channel without blocking. The channel is
+// expected to be buffered (cap 1); a nil channel means the caller did not ask
+// to be notified.
+func notify(ch chan error, err error) {
+	if ch == nil {
+		return
+	}
+	select {
+	case ch <- err:
+	default:
+	}
+}

client/internal/dns/mgmt/mgmt.go

@@ -51,20 +51,13 @@ type cachedRecord struct {
 }
 
 // Resolver caches critical NetBird infrastructure domains.
-// records, refreshing, failedResolves, mgmtDomain and serverDomains are all
-// guarded by mutex.
+// records, refreshing, mgmtDomain and serverDomains are all guarded by mutex.
 type Resolver struct {
 	records       map[dns.Question]*cachedRecord
 	mgmtDomain    *domain.Domain
 	serverDomains *dnsconfig.ServerDomains
 	mutex         sync.RWMutex
 
-	// failedResolves records the last failed initial resolve per domain so a
-	// domain that never resolves isn't retried on every server-domains update
-	// until refreshBackoff elapses. Entries are cleared on success and pruned
-	// to the current server-domains set.
-	failedResolves map[domain.Domain]time.Time
-
 	chain            ChainResolver
 	chainMaxPriority int
 	refreshGroup     singleflight.Group
@@ -83,10 +76,9 @@ type Resolver struct {
 // NewResolver creates a new management domains cache resolver.
 func NewResolver() *Resolver {
 	return &Resolver{
-		records:        make(map[dns.Question]*cachedRecord),
-		refreshing:     make(map[dns.Question]*atomic.Bool),
-		failedResolves: make(map[domain.Domain]time.Time),
-		cacheTTL:       resolveCacheTTL(),
+		records:    make(map[dns.Question]*cachedRecord),
+		refreshing: make(map[dns.Question]*atomic.Bool),
+		cacheTTL:   resolveCacheTTL(),
 	}
 }
 
@@ -181,9 +173,7 @@ func (m *Resolver) continueToNext(w dns.ResponseWriter, r *dns.Msg) {
 
 // AddDomain resolves a domain and stores its A/AAAA records in the cache.
 // A family that resolves NODATA (nil err, zero records) evicts any stale
-// entry for that qtype. When one family hard-errors while the other succeeds,
-// the resolved family is still cached but AddDomain returns an error so the
-// caller retries the incomplete resolve rather than treating it as complete.
+// entry for that qtype.
 func (m *Resolver) AddDomain(ctx context.Context, d domain.Domain) error {
 	dnsName := strings.ToLower(dns.Fqdn(d.PunycodeString()))
 
@@ -213,10 +203,6 @@ func (m *Resolver) AddDomain(ctx context.Context, d domain.Domain) error {
 	log.Debugf("added/updated domain=%s with %d A records and %d AAAA records",
 		d.SafeString(), len(aRecords), len(aaaaRecords))
 
-	if errA != nil || errAAAA != nil {
-		return fmt.Errorf("resolve %s: incomplete, a family failed: %w", d.SafeString(), errors.Join(errA, errAAAA))
-	}
-
 	return nil
 }
 
@@ -476,7 +462,6 @@ func (m *Resolver) RemoveDomain(d domain.Domain) error {
 	delete(m.records, qAAAA)
 	delete(m.refreshing, qA)
 	delete(m.refreshing, qAAAA)
-	delete(m.failedResolves, d)
 
 	log.Debugf("removed domain=%s from cache", d.SafeString())
 	return nil
@@ -520,7 +505,6 @@ func (m *Resolver) UpdateFromServerDomains(ctx context.Context, serverDomains dn
 		allDomains := m.extractDomainsFromServerDomains(updatedServerDomains)
 		currentDomains := m.GetCachedDomains()
 		removedDomains = m.removeStaleDomains(currentDomains, allDomains)
-		m.pruneFailedResolves(allDomains)
 	}
 
 	m.addNewDomains(ctx, newDomains)
@@ -593,85 +577,13 @@ func (m *Resolver) isManagementDomain(domain domain.Domain) bool {
 	return m.mgmtDomain != nil && domain == *m.mgmtDomain
 }
 
-// addNewDomains resolves and caches domains that are not yet in the cache,
-// running the lookups concurrently. Domains already cached are skipped and left
-// to the stale-while-revalidate refresh path, so a sync never re-resolves them
-// synchronously: once NetBird owns the OS resolver the resolve runs through the
-// handler chain and would otherwise dial the managed upstreams under the engine
-// sync lock on every update.
+// addNewDomains resolves and caches all domains from the update
 func (m *Resolver) addNewDomains(ctx context.Context, newDomains domain.List) {
-	var wg sync.WaitGroup
-	seen := make(map[domain.Domain]struct{}, len(newDomains))
 	for _, newDomain := range newDomains {
-		if _, dup := seen[newDomain]; dup {
-			continue
-		}
-		seen[newDomain] = struct{}{}
-
-		if !m.needsResolve(newDomain) {
-			continue
-		}
-
-		wg.Add(1)
-		go func(d domain.Domain) {
-			defer wg.Done()
-			if err := m.AddDomain(ctx, d); err != nil {
-				m.markResolveFailed(d)
-				log.Warnf("failed to add/update domain=%s: %v", d.SafeString(), err)
-				return
-			}
-			m.clearResolveFailed(d)
-			log.Debugf("added/updated management cache domain=%s", d.SafeString())
-		}(newDomain)
-	}
-	wg.Wait()
-}
-
-// needsResolve reports whether d should be resolved now. A recent failed or
-// incomplete resolve gates retries on the backoff even when one family is
-// already cached, so a transiently-failed family is retried instead of being
-// treated as fully resolved. Otherwise a domain with any cached record is left
-// to the stale-while-revalidate refresh path.
-func (m *Resolver) needsResolve(d domain.Domain) bool {
-	dnsName := strings.ToLower(dns.Fqdn(d.PunycodeString()))
-
-	m.mutex.RLock()
-	defer m.mutex.RUnlock()
-
-	if failedAt, ok := m.failedResolves[d]; ok {
-		return time.Since(failedAt) >= refreshBackoff
-	}
-
-	for _, qtype := range []uint16{dns.TypeA, dns.TypeAAAA} {
-		q := dns.Question{Name: dnsName, Qtype: qtype, Qclass: dns.ClassINET}
-		if _, ok := m.records[q]; ok {
-			return false
-		}
-	}
-	return true
-}
-
-func (m *Resolver) markResolveFailed(d domain.Domain) {
-	m.mutex.Lock()
-	m.failedResolves[d] = time.Now()
-	m.mutex.Unlock()
-}
-
-func (m *Resolver) clearResolveFailed(d domain.Domain) {
-	m.mutex.Lock()
-	delete(m.failedResolves, d)
-	m.mutex.Unlock()
-}
-
-// pruneFailedResolves drops failure markers for domains no longer present in
-// the server-domains set, keeping the map bounded to the current set (a
-// failed-only domain has no cached record, so RemoveDomain never sees it).
-func (m *Resolver) pruneFailedResolves(domains domain.List) {
-	m.mutex.Lock()
-	defer m.mutex.Unlock()
-	for d := range m.failedResolves {
-		if !slices.Contains(domains, d) {
-			delete(m.failedResolves, d)
+		if err := m.AddDomain(ctx, newDomain); err != nil {
+			log.Warnf("failed to add/update domain=%s: %v", newDomain.SafeString(), err)
+		} else {
+			log.Debugf("added/updated management cache domain=%s", newDomain.SafeString())
 		}
 	}
 }

client/internal/dns/mgmt/mgmt_refresh_test.go

@@ -21,7 +21,6 @@ type fakeChain struct {
 	mu       sync.Mutex
 	calls    map[string]int
 	answers  map[string][]dns.RR
-	qErr     map[string]error
 	err      error
 	hasRoot  bool
 	onLookup func()
@@ -31,7 +30,6 @@ func newFakeChain() *fakeChain {
 	return &fakeChain{
 		calls:   map[string]int{},
 		answers: map[string][]dns.RR{},
-		qErr:    map[string]error{},
 		hasRoot: true,
 	}
 }
@@ -49,9 +47,6 @@ func (f *fakeChain) ResolveInternal(ctx context.Context, msg *dns.Msg, maxPriori
 	f.calls[key]++
 	answers := f.answers[key]
 	err := f.err
-	if err == nil {
-		err = f.qErr[key]
-	}
 	onLookup := f.onLookup
 	f.mu.Unlock()
 
@@ -80,12 +75,6 @@ func (f *fakeChain) setAnswer(name string, qtype uint16, ip string) {
 	}
 }
 
-func (f *fakeChain) setErr(name string, qtype uint16, err error) {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	f.qErr[name+"|"+dns.TypeToString[qtype]] = err
-}
-
 func (f *fakeChain) callCount(name string, qtype uint16) int {
 	f.mu.Lock()
 	defer f.mu.Unlock()

client/internal/dns/mgmt/mgmt_resolve_test.go

@@ -1,183 +0,0 @@
-package mgmt
-
-import (
-	"context"
-	"errors"
-	"sync/atomic"
-	"testing"
-	"time"
-
-	"github.com/miekg/dns"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-
-	dnsconfig "github.com/netbirdio/netbird/client/internal/dns/config"
-	"github.com/netbirdio/netbird/shared/management/domain"
-)
-
-// A domain already in the cache must not be re-resolved on a subsequent server
-// domains update; it is left to the stale-while-revalidate refresh path.
-func TestResolver_UpdateFromServerDomains_SkipsCached(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("signal.example.com.", dns.TypeA, "10.0.0.2")
-	r.SetChainResolver(chain, 50)
-
-	sd := dnsconfig.ServerDomains{Signal: domain.Domain("signal.example.com")}
-
-	_, err := r.UpdateFromServerDomains(context.Background(), sd)
-	require.NoError(t, err)
-	require.Equal(t, 1, chain.callCount("signal.example.com.", dns.TypeA),
-		"first update must resolve the domain")
-
-	_, err = r.UpdateFromServerDomains(context.Background(), sd)
-	require.NoError(t, err)
-	assert.Equal(t, 1, chain.callCount("signal.example.com.", dns.TypeA),
-		"cached domain must not be re-resolved on a subsequent update")
-}
-
-// New domains in a single update must resolve concurrently rather than serially.
-func TestResolver_AddNewDomains_ResolvesConcurrently(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-
-	var inflight, maxInflight atomic.Int32
-	chain.onLookup = func() {
-		n := inflight.Add(1)
-		for {
-			old := maxInflight.Load()
-			if n <= old || maxInflight.CompareAndSwap(old, n) {
-				break
-			}
-		}
-		time.Sleep(50 * time.Millisecond)
-		inflight.Add(-1)
-	}
-
-	relays := []domain.Domain{"a.example.com", "b.example.com", "c.example.com", "d.example.com"}
-	for _, d := range relays {
-		chain.setAnswer(dns.Fqdn(string(d)), dns.TypeA, "10.0.0.2")
-	}
-	r.SetChainResolver(chain, 50)
-
-	start := time.Now()
-	_, err := r.UpdateFromServerDomains(context.Background(), dnsconfig.ServerDomains{Relay: relays})
-	require.NoError(t, err)
-	elapsed := time.Since(start)
-
-	assert.GreaterOrEqual(t, int(maxInflight.Load()), 2, "domains must resolve concurrently")
-	// Serial resolution of 4 domains would take at least 4*50ms; concurrent is far less.
-	assert.Less(t, elapsed, 300*time.Millisecond, "resolution should not be serial")
-}
-
-// A domain that fails to resolve must not be retried on every update; the
-// failure backoff suppresses re-resolution until it expires.
-func TestResolver_UpdateFromServerDomains_BacksOffFailures(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.err = errors.New("resolve boom")
-	r.SetChainResolver(chain, 50)
-
-	sd := dnsconfig.ServerDomains{Signal: domain.Domain("signal.example.com")}
-
-	_, err := r.UpdateFromServerDomains(context.Background(), sd)
-	require.NoError(t, err)
-	require.Equal(t, 1, chain.callCount("signal.example.com.", dns.TypeA),
-		"first update must attempt the resolve")
-
-	_, err = r.UpdateFromServerDomains(context.Background(), sd)
-	require.NoError(t, err)
-	assert.Equal(t, 1, chain.callCount("signal.example.com.", dns.TypeA),
-		"failed resolve must back off and not retry on the next update")
-}
-
-// A domain listed under more than one server-domain type (e.g. STUN and TURN on
-// the same host) must be resolved once per update, not once per occurrence.
-func TestResolver_AddNewDomains_DedupesDuplicateDomains(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("dup.example.com.", dns.TypeA, "10.0.0.9")
-	r.SetChainResolver(chain, 50)
-
-	sd := dnsconfig.ServerDomains{
-		Stuns: []domain.Domain{"dup.example.com"},
-		Turns: []domain.Domain{"dup.example.com"},
-	}
-
-	_, err := r.UpdateFromServerDomains(context.Background(), sd)
-	require.NoError(t, err)
-	assert.Equal(t, 1, chain.callCount("dup.example.com.", dns.TypeA),
-		"a domain appearing under multiple server-domain types must resolve once")
-}
-
-// A failure marker must be dropped once its domain leaves the server-domains set
-// so the map stays bounded to the current set.
-func TestResolver_UpdateFromServerDomains_PrunesFailedResolves(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.err = errors.New("resolve boom")
-	r.SetChainResolver(chain, 50)
-
-	_, err := r.UpdateFromServerDomains(context.Background(), dnsconfig.ServerDomains{Signal: domain.Domain("gone.example.com")})
-	require.NoError(t, err)
-	r.mutex.RLock()
-	_, marked := r.failedResolves[domain.Domain("gone.example.com")]
-	r.mutex.RUnlock()
-	require.True(t, marked, "failed resolve must be recorded")
-
-	_, err = r.UpdateFromServerDomains(context.Background(), dnsconfig.ServerDomains{Signal: domain.Domain("other.example.com")})
-	require.NoError(t, err)
-	r.mutex.RLock()
-	_, stillMarked := r.failedResolves[domain.Domain("gone.example.com")]
-	r.mutex.RUnlock()
-	assert.False(t, stillMarked, "failure marker for a domain no longer in the set must be pruned")
-}
-
-// When one family hard-errors while the other resolves, the domain is cached
-// for the working family but recorded as incomplete so the failed family is
-// retried under backoff instead of being treated as fully resolved forever.
-func TestResolver_AddNewDomains_RetriesPartialFamilyFailure(t *testing.T) {
-	d := domain.Domain("relay.example.com")
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("relay.example.com.", dns.TypeA, "10.0.0.2")
-	chain.setErr("relay.example.com.", dns.TypeAAAA, errors.New("servfail"))
-	r.SetChainResolver(chain, 50)
-
-	_, err := r.UpdateFromServerDomains(context.Background(), dnsconfig.ServerDomains{Relay: []domain.Domain{d}})
-	require.NoError(t, err)
-
-	r.mutex.RLock()
-	_, aCached := r.records[dns.Question{Name: "relay.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}]
-	_, marked := r.failedResolves[d]
-	r.mutex.RUnlock()
-	require.True(t, aCached, "the working family must still be cached")
-	require.True(t, marked, "a partial failure must be recorded so the failed family is retried")
-
-	assert.False(t, r.needsResolve(d), "within the backoff window the domain is not retried")
-
-	r.mutex.Lock()
-	r.failedResolves[d] = time.Now().Add(-2 * refreshBackoff)
-	r.mutex.Unlock()
-	assert.True(t, r.needsResolve(d), "after the backoff elapses the domain is retried to pick up the missing family")
-}
-
-// A family that returns NODATA (legitimately absent, e.g. an IPv4-only host) is
-// not a failure: the domain must not be marked for retry, otherwise it would be
-// re-resolved on every sync.
-func TestResolver_AddNewDomains_NodataIsNotFailure(t *testing.T) {
-	d := domain.Domain("v4only.example.com")
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("v4only.example.com.", dns.TypeA, "10.0.0.2")
-	r.SetChainResolver(chain, 50)
-
-	_, err := r.UpdateFromServerDomains(context.Background(), dnsconfig.ServerDomains{Relay: []domain.Domain{d}})
-	require.NoError(t, err)
-
-	r.mutex.RLock()
-	_, marked := r.failedResolves[d]
-	r.mutex.RUnlock()
-	assert.False(t, marked, "a NODATA family must not be recorded as a failure")
-	assert.False(t, r.needsResolve(d), "an IPv4-only host must not be re-resolved on later syncs")
-}

client/internal/dns/resutil/resolve.go

@@ -8,7 +8,6 @@ import (
 	"errors"
 	"net"
 	"net/netip"
-	"slices"
 	"strings"
 
 	"github.com/miekg/dns"
@@ -168,10 +167,7 @@ func getRcodeForNotFound(ctx context.Context, r resolver, domain string, origina
 	case dns.TypeA:
 		alternativeNetwork = "ip6"
 	default:
-		// Non-address types reach LookupIP only unexpectedly; without an
-		// address pair to probe we cannot prove the name is absent, so answer
-		// NODATA rather than a poisoning NXDOMAIN.
-		return dns.RcodeSuccess
+		return dns.RcodeNameError
 	}
 
 	if _, err := r.LookupNetIP(ctx, alternativeNetwork, domain); err != nil {
@@ -188,230 +184,6 @@ func getRcodeForNotFound(ctx context.Context, r resolver, domain string, origina
 	return dns.RcodeSuccess
 }
 
-// RecordResolver is the host resolver surface used to forward non-address
-// record queries. net.DefaultResolver satisfies it.
-type RecordResolver interface {
-	LookupMX(ctx context.Context, name string) ([]*net.MX, error)
-	LookupTXT(ctx context.Context, name string) ([]string, error)
-	LookupNS(ctx context.Context, name string) ([]*net.NS, error)
-	LookupSRV(ctx context.Context, service, proto, name string) (string, []*net.SRV, error)
-	LookupCNAME(ctx context.Context, host string) (string, error)
-	LookupAddr(ctx context.Context, addr string) ([]string, error)
-}
-
-// LookupRecords resolves a non-address DNS record type through the host
-// resolver and returns the resource records and the DNS rcode. Types the host
-// resolver cannot answer (anything not covered by the net.Resolver Lookup*
-// methods) yield NODATA so that a routed name is never poisoned with NXDOMAIN
-// for an unsupported type.
-func LookupRecords(ctx context.Context, r RecordResolver, name string, qtype uint16, ttl uint32) ([]dns.RR, int) {
-	fqdn := dns.Fqdn(name)
-
-	switch qtype {
-	case dns.TypeMX:
-		return lookupMX(ctx, r, name, fqdn, ttl)
-	case dns.TypeTXT:
-		return lookupTXT(ctx, r, name, fqdn, ttl)
-	case dns.TypeNS:
-		return lookupNS(ctx, r, name, fqdn, ttl)
-	case dns.TypeSRV:
-		return lookupSRV(ctx, r, name, fqdn, ttl)
-	case dns.TypeCNAME:
-		return lookupCNAME(ctx, r, name, fqdn, ttl)
-	case dns.TypePTR:
-		return lookupPTR(ctx, r, name, fqdn, ttl)
-	default:
-		return nil, dns.RcodeSuccess
-	}
-}
-
-func recordHeader(fqdn string, rrtype uint16, ttl uint32) dns.RR_Header {
-	return dns.RR_Header{Name: fqdn, Rrtype: rrtype, Class: dns.ClassINET, Ttl: ttl}
-}
-
-func lookupMX(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	recs, err := r.LookupMX(ctx, name)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	rrs := make([]dns.RR, 0, len(recs))
-	for _, mx := range recs {
-		rrs = append(rrs, &dns.MX{
-			Hdr:        recordHeader(fqdn, dns.TypeMX, ttl),
-			Preference: mx.Pref,
-			Mx:         dns.Fqdn(mx.Host),
-		})
-	}
-	return rrs, dns.RcodeSuccess
-}
-
-func lookupTXT(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	recs, err := r.LookupTXT(ctx, name)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	rrs := make([]dns.RR, 0, len(recs))
-	for _, txt := range recs {
-		rrs = append(rrs, &dns.TXT{
-			Hdr: recordHeader(fqdn, dns.TypeTXT, ttl),
-			Txt: chunkTXT(txt),
-		})
-	}
-	return rrs, dns.RcodeSuccess
-}
-
-func lookupNS(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	recs, err := r.LookupNS(ctx, name)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	rrs := make([]dns.RR, 0, len(recs))
-	for _, ns := range recs {
-		rrs = append(rrs, &dns.NS{
-			Hdr: recordHeader(fqdn, dns.TypeNS, ttl),
-			Ns:  dns.Fqdn(ns.Host),
-		})
-	}
-	return rrs, dns.RcodeSuccess
-}
-
-func lookupSRV(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	_, recs, err := r.LookupSRV(ctx, "", "", name)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	rrs := make([]dns.RR, 0, len(recs))
-	for _, srv := range recs {
-		rrs = append(rrs, &dns.SRV{
-			Hdr:      recordHeader(fqdn, dns.TypeSRV, ttl),
-			Priority: srv.Priority,
-			Weight:   srv.Weight,
-			Port:     srv.Port,
-			Target:   dns.Fqdn(srv.Target),
-		})
-	}
-	return rrs, dns.RcodeSuccess
-}
-
-func lookupCNAME(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	cname, err := r.LookupCNAME(ctx, name)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	// LookupCNAME returns the queried name itself when the name resolves but
-	// has no CNAME record; that is a NODATA result, not a CNAME.
-	if strings.EqualFold(dns.Fqdn(cname), fqdn) {
-		return nil, dns.RcodeSuccess
-	}
-	return []dns.RR{&dns.CNAME{
-		Hdr:    recordHeader(fqdn, dns.TypeCNAME, ttl),
-		Target: dns.Fqdn(cname),
-	}}, dns.RcodeSuccess
-}
-
-func lookupPTR(ctx context.Context, r RecordResolver, name, fqdn string, ttl uint32) ([]dns.RR, int) {
-	addr, ok := ptrQueryAddr(name)
-	if !ok {
-		return nil, dns.RcodeSuccess
-	}
-	names, err := r.LookupAddr(ctx, addr)
-	if err != nil {
-		return nil, rcodeForRecordError(err)
-	}
-	rrs := make([]dns.RR, 0, len(names))
-	for _, n := range names {
-		rrs = append(rrs, &dns.PTR{
-			Hdr: recordHeader(fqdn, dns.TypePTR, ttl),
-			Ptr: dns.Fqdn(n),
-		})
-	}
-	return rrs, dns.RcodeSuccess
-}
-
-// ptrQueryAddr converts a reverse-DNS query name (in-addr.arpa or ip6.arpa)
-// into the address string expected by net.Resolver.LookupAddr. It reports false
-// when the name is not a well-formed reverse name.
-func ptrQueryAddr(qname string) (string, bool) {
-	name := strings.TrimSuffix(strings.ToLower(dns.Fqdn(qname)), ".")
-
-	switch {
-	case strings.HasSuffix(name, ".in-addr.arpa"):
-		return parseInAddrArpa(strings.TrimSuffix(name, ".in-addr.arpa"))
-	case strings.HasSuffix(name, ".ip6.arpa"):
-		return parseIP6Arpa(strings.TrimSuffix(name, ".ip6.arpa"))
-	default:
-		return "", false
-	}
-}
-
-// parseInAddrArpa turns the label portion of an in-addr.arpa name into an IPv4
-// address string, reporting false when it is not a well-formed reverse name.
-func parseInAddrArpa(labelPart string) (string, bool) {
-	labels := strings.Split(labelPart, ".")
-	if len(labels) != 4 {
-		return "", false
-	}
-	slices.Reverse(labels)
-	addr, err := netip.ParseAddr(strings.Join(labels, "."))
-	if err != nil || !addr.Is4() {
-		return "", false
-	}
-	return addr.String(), true
-}
-
-// parseIP6Arpa turns the nibble portion of an ip6.arpa name into an IPv6
-// address string, reporting false when it is not a well-formed reverse name.
-func parseIP6Arpa(nibblePart string) (string, bool) {
-	nibbles := strings.Split(nibblePart, ".")
-	if len(nibbles) != 32 {
-		return "", false
-	}
-	slices.Reverse(nibbles)
-	var sb strings.Builder
-	for i, n := range nibbles {
-		if i > 0 && i%4 == 0 {
-			sb.WriteByte(':')
-		}
-		sb.WriteString(n)
-	}
-	addr, err := netip.ParseAddr(sb.String())
-	if err != nil || !addr.Is6() {
-		return "", false
-	}
-	return addr.String(), true
-}
-
-// rcodeForRecordError maps a non-address lookup error to a DNS rcode. A
-// not-found result becomes NODATA rather than NXDOMAIN: net.DNSError.IsNotFound
-// does not distinguish a missing name from a name that exists only with records
-// of other types, so the name cannot be proven absent and must not be poisoned.
-func rcodeForRecordError(err error) int {
-	var dnsErr *net.DNSError
-	if errors.As(err, &dnsErr) && dnsErr.IsNotFound {
-		return dns.RcodeSuccess
-	}
-	return dns.RcodeServerFailure
-}
-
-// chunkTXT splits a TXT string into character-strings no longer than 255 bytes
-// so the record can be packed. The chunks form one TXT resource record.
-func chunkTXT(s string) []string {
-	const maxLen = 255
-	if len(s) <= maxLen {
-		return []string{s}
-	}
-
-	var chunks []string
-	for len(s) > maxLen {
-		chunks = append(chunks, s[:maxLen])
-		s = s[maxLen:]
-	}
-	if len(s) > 0 {
-		chunks = append(chunks, s)
-	}
-	return chunks
-}
-
 // FormatAnswers formats DNS resource records for logging.
 func FormatAnswers(answers []dns.RR) string {
 	if len(answers) == 0 {
@@ -435,35 +207,3 @@ func FormatAnswers(answers []dns.RR) string {
 	}
 	return "[" + strings.Join(parts, ", ") + "]"
 }
-
-// StripOPT removes any OPT pseudo-RRs from the message's Extra section. Per
-// RFC 6891 a responder must not include an OPT RR toward a client that did not
-// advertise EDNS0.
-func StripOPT(msg *dns.Msg) {
-	if len(msg.Extra) == 0 {
-		return
-	}
-	out := msg.Extra[:0]
-	for _, rr := range msg.Extra {
-		if _, ok := rr.(*dns.OPT); ok {
-			continue
-		}
-		out = append(out, rr)
-	}
-	msg.Extra = out
-}
-
-// ExtractEDE returns the first Extended DNS Error (RFC 8914) option carried in
-// the message, if present.
-func ExtractEDE(msg *dns.Msg) (*dns.EDNS0_EDE, bool) {
-	opt := msg.IsEdns0()
-	if opt == nil {
-		return nil, false
-	}
-	for _, o := range opt.Option {
-		if ede, ok := o.(*dns.EDNS0_EDE); ok {
-			return ede, true
-		}
-	}
-	return nil, false
-}

client/internal/dns/resutil/resolve_test.go

@@ -5,7 +5,6 @@ import (
 	"errors"
 	"net"
 	"net/netip"
-	"strings"
 	"testing"
 
 	"github.com/miekg/dns"
@@ -121,200 +120,3 @@ func TestLookupIP_DNSErrorNotIsNotFound(t *testing.T) {
 
 	assert.Equal(t, dns.RcodeServerFailure, result.Rcode, "upstream failure should map to SERVFAIL")
 }
-
-func TestPtrQueryAddr(t *testing.T) {
-	tests := []struct {
-		name   string
-		qname  string
-		want   string
-		wantOK bool
-	}{
-		{name: "ipv4", qname: "4.3.2.1.in-addr.arpa.", want: "1.2.3.4", wantOK: true},
-		{name: "ipv4 no trailing dot", qname: "1.0.0.127.in-addr.arpa", want: "127.0.0.1", wantOK: true},
-		{
-			name:   "ipv6",
-			qname:  "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.arpa.",
-			want:   "2001:db8::1",
-			wantOK: true,
-		},
-		{name: "ipv4 wrong label count", qname: "2.1.in-addr.arpa.", wantOK: false},
-		{name: "ipv6 wrong nibble count", qname: "1.0.ip6.arpa.", wantOK: false},
-		{name: "not a reverse name", qname: "example.com.", wantOK: false},
-		{name: "ipv4 bad octet", qname: "4.3.2.999.in-addr.arpa.", wantOK: false},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			got, ok := ptrQueryAddr(tt.qname)
-			assert.Equal(t, tt.wantOK, ok, "parse success mismatch")
-			if tt.wantOK {
-				assert.Equal(t, tt.want, got, "parsed address mismatch")
-			}
-		})
-	}
-}
-
-type mockRecordResolver struct {
-	mx    []*net.MX
-	txt   []string
-	ns    []*net.NS
-	srv   []*net.SRV
-	cname string
-	ptr   []string
-	err   error
-}
-
-func (m *mockRecordResolver) LookupMX(context.Context, string) ([]*net.MX, error) {
-	return m.mx, m.err
-}
-func (m *mockRecordResolver) LookupTXT(context.Context, string) ([]string, error) {
-	return m.txt, m.err
-}
-func (m *mockRecordResolver) LookupNS(context.Context, string) ([]*net.NS, error) {
-	return m.ns, m.err
-}
-func (m *mockRecordResolver) LookupSRV(context.Context, string, string, string) (string, []*net.SRV, error) {
-	return "", m.srv, m.err
-}
-func (m *mockRecordResolver) LookupCNAME(context.Context, string) (string, error) {
-	return m.cname, m.err
-}
-func (m *mockRecordResolver) LookupAddr(context.Context, string) ([]string, error) {
-	return m.ptr, m.err
-}
-
-func TestLookupRecords(t *testing.T) {
-	notFound := &net.DNSError{IsNotFound: true, Name: "example.com."}
-
-	t.Run("MX success", func(t *testing.T) {
-		r := &mockRecordResolver{mx: []*net.MX{{Host: "mail.example.com.", Pref: 10}}}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeMX, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, "mail.example.com.", rrs[0].(*dns.MX).Mx)
-	})
-
-	t.Run("TXT short string is one character-string", func(t *testing.T) {
-		r := &mockRecordResolver{txt: []string{"v=spf1 -all"}}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeTXT, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, []string{"v=spf1 -all"}, rrs[0].(*dns.TXT).Txt)
-	})
-
-	t.Run("TXT chunks long strings", func(t *testing.T) {
-		long := strings.Repeat("a", 300)
-		r := &mockRecordResolver{txt: []string{long}}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeTXT, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		txt := rrs[0].(*dns.TXT).Txt
-		require.Len(t, txt, 2, "300-byte string should split into two character-strings")
-		assert.Equal(t, 255, len(txt[0]))
-		assert.Equal(t, 45, len(txt[1]))
-	})
-
-	t.Run("NS success", func(t *testing.T) {
-		r := &mockRecordResolver{ns: []*net.NS{{Host: "ns1.example.com."}}}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeNS, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, "ns1.example.com.", rrs[0].(*dns.NS).Ns)
-	})
-
-	t.Run("SRV success", func(t *testing.T) {
-		r := &mockRecordResolver{srv: []*net.SRV{{Target: "sip.example.com.", Port: 5060}}}
-		rrs, rcode := LookupRecords(context.Background(), r, "_sip._tcp.example.com.", dns.TypeSRV, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, uint16(5060), rrs[0].(*dns.SRV).Port)
-	})
-
-	t.Run("CNAME success", func(t *testing.T) {
-		r := &mockRecordResolver{cname: "target.example.com."}
-		rrs, rcode := LookupRecords(context.Background(), r, "www.example.com.", dns.TypeCNAME, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, "target.example.com.", rrs[0].(*dns.CNAME).Target)
-	})
-
-	t.Run("CNAME equal to name is NODATA", func(t *testing.T) {
-		r := &mockRecordResolver{cname: "example.com."}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeCNAME, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		assert.Empty(t, rrs, "self-referential CNAME is NODATA")
-	})
-
-	t.Run("PTR success", func(t *testing.T) {
-		r := &mockRecordResolver{ptr: []string{"host.example.com."}}
-		rrs, rcode := LookupRecords(context.Background(), r, "4.3.2.1.in-addr.arpa.", dns.TypePTR, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		require.Len(t, rrs, 1)
-		assert.Equal(t, "host.example.com.", rrs[0].(*dns.PTR).Ptr)
-	})
-
-	t.Run("PTR malformed name is NODATA", func(t *testing.T) {
-		r := &mockRecordResolver{}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypePTR, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		assert.Empty(t, rrs)
-	})
-
-	t.Run("not found is NODATA never NXDOMAIN", func(t *testing.T) {
-		r := &mockRecordResolver{err: notFound}
-		_, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeMX, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode, "missing record must not poison the name")
-	})
-
-	t.Run("server failure maps to SERVFAIL", func(t *testing.T) {
-		r := &mockRecordResolver{err: &net.DNSError{Err: "server misbehaving", IsTemporary: true}}
-		_, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeMX, 300)
-		assert.Equal(t, dns.RcodeServerFailure, rcode)
-	})
-
-	t.Run("unsupported type is NODATA", func(t *testing.T) {
-		r := &mockRecordResolver{}
-		rrs, rcode := LookupRecords(context.Background(), r, "example.com.", dns.TypeCAA, 300)
-		assert.Equal(t, dns.RcodeSuccess, rcode)
-		assert.Empty(t, rrs)
-	})
-}
-
-func TestStripOPT(t *testing.T) {
-	rm := &dns.Msg{
-		Extra: []dns.RR{
-			&dns.OPT{Hdr: dns.RR_Header{Name: ".", Rrtype: dns.TypeOPT}},
-			&dns.A{Hdr: dns.RR_Header{Name: "x.", Rrtype: dns.TypeA}, A: net.IPv4(1, 2, 3, 4)},
-		},
-	}
-	StripOPT(rm)
-	assert.Len(t, rm.Extra, 1, "OPT should be removed, A kept")
-	_, isOPT := rm.Extra[0].(*dns.OPT)
-	assert.False(t, isOPT, "remaining record must not be OPT")
-}
-
-func TestExtractEDE(t *testing.T) {
-	t.Run("no edns", func(t *testing.T) {
-		_, ok := ExtractEDE(&dns.Msg{})
-		assert.False(t, ok, "message without OPT has no EDE")
-	})
-
-	t.Run("edns without ede", func(t *testing.T) {
-		rm := &dns.Msg{}
-		rm.SetEdns0(4096, false)
-		_, ok := ExtractEDE(rm)
-		assert.False(t, ok, "OPT without EDE option returns false")
-	})
-
-	t.Run("with ede", func(t *testing.T) {
-		rm := &dns.Msg{}
-		opt := &dns.OPT{Hdr: dns.RR_Header{Name: ".", Rrtype: dns.TypeOPT}}
-		opt.Option = append(opt.Option, &dns.EDNS0_EDE{InfoCode: 49152, ExtraText: "upstream timeout"})
-		rm.Extra = append(rm.Extra, opt)
-
-		ede, ok := ExtractEDE(rm)
-		assert.True(t, ok, "EDE option should be found")
-		assert.Equal(t, uint16(49152), ede.InfoCode)
-		assert.Equal(t, "upstream timeout", ede.ExtraText)
-	})
-}

client/internal/dns/server.go

@@ -6,7 +6,6 @@ import (
 	"fmt"
 	"net/netip"
 	"net/url"
-	"os"
 	"slices"
 	"strings"
 	"sync"
@@ -39,15 +38,11 @@ const (
 	// defaultWarningDelayBase is the starting grace window before a
 	// "Nameserver group unreachable" event fires for a group that's
 	// never been healthy and only has overlay upstreams with no
-	// Connected peer. Per-server and overridable via envWarningDelay;
-	// see warningDelay.
-	defaultWarningDelayBase = 60 * time.Second
+	// Connected peer. Per-server and overridable; see warningDelayFor.
+	defaultWarningDelayBase = 30 * time.Second
 	// warningDelayBonusCap caps the route-count bonus added to the
-	// base grace window. See warningDelay.
+	// base grace window. See warningDelayFor.
 	warningDelayBonusCap = 30 * time.Second
-	// envWarningDelay overrides defaultWarningDelayBase with a Go duration
-	// string (e.g. "90s", "2m"). Invalid or non-positive values are ignored.
-	envWarningDelay = "NB_DNS_HEALTH_WARNING_DELAY"
 )
 
 // errNoUsableNameservers signals that a merged-domain group has no usable
@@ -140,7 +135,7 @@ type DefaultServer struct {
 	disableSys         bool
 	mux                sync.Mutex
 	service            service
-	dnsMuxHandlers     []handlerWrapper
+	dnsMuxMap          registeredHandlerMap
 	localResolver      *local.Resolver
 	wgInterface        WGIface
 	hostManager        hostManager
@@ -204,6 +199,8 @@ type handlerWrapper struct {
 	priority int
 }
 
+type registeredHandlerMap map[types.HandlerID]handlerWrapper
+
 // DefaultServerConfig holds configuration parameters for NewDefaultServer
 type DefaultServerConfig struct {
 	WgInterface    WGIface
@@ -292,6 +289,7 @@ func newDefaultServer(
 		service:           dnsService,
 		handlerChain:      handlerChain,
 		extraDomains:      make(map[domain.Domain]int),
+		dnsMuxMap:         make(registeredHandlerMap),
 		localResolver:     local.NewResolver(),
 		wgInterface:       wgInterface,
 		statusRecorder:    statusRecorder,
@@ -300,7 +298,7 @@ func newDefaultServer(
 		hostManager:       &noopHostConfigurator{},
 		mgmtCacheResolver: mgmtCacheResolver,
 		currentConfigHash: ^uint64(0), // Initialize to max uint64 to ensure first config is always applied
-		warningDelayBase:  warningDelayBaseFromEnv(),
+		warningDelayBase:  defaultWarningDelayBase,
 		healthRefresh:     make(chan struct{}, 1),
 	}
 	// Wire the local resolver against the peer status recorder so it can
@@ -330,7 +328,7 @@ func (s *DefaultServer) SetRouteSources(selected, active func() route.HAMap) {
 	type routeSettable interface {
 		setSelectedRoutes(func() route.HAMap)
 	}
-	for _, entry := range s.dnsMuxHandlers {
+	for _, entry := range s.dnsMuxMap {
 		if h, ok := entry.handler.(routeSettable); ok {
 			h.setSelectedRoutes(selected)
 		}
@@ -980,23 +978,19 @@ func (s *DefaultServer) usableNameServers(nameServers []nbdns.NameServer) []neti
 
 func (s *DefaultServer) updateMux(muxUpdates []handlerWrapper) {
 	// this will introduce a short period of time when the server is not able to handle DNS requests
-	for _, existing := range s.dnsMuxHandlers {
+	for _, existing := range s.dnsMuxMap {
 		s.deregisterHandler([]string{existing.domain}, existing.priority)
-		// The local resolver is a persistent singleton shared by every custom
-		// zone and reused across config updates. Its chain registrations are
-		// per-config and must be deregistered, but Stop() cancels its lookup
-		// context (breaking external CNAME-target resolution) and clears its
-		// records, so it must not be torn down here.
-		if existing.handler != s.localResolver {
-			existing.handler.Stop()
-		}
+		existing.handler.Stop()
 	}
 
+	muxUpdateMap := make(registeredHandlerMap)
+
 	for _, update := range muxUpdates {
 		s.registerHandler([]string{update.domain}, update.handler, update.priority)
+		muxUpdateMap[update.handler.ID()] = update
 	}
 
-	s.dnsMuxHandlers = muxUpdates
+	s.dnsMuxMap = muxUpdateMap
 }
 
 // updateNSGroupStates records the new group set and pokes the refresher.
@@ -1160,26 +1154,6 @@ func (s *DefaultServer) projectUnhealthy(p *nsGroupProj, servers []netip.AddrPor
 	return false
 }
 
-// warningDelayBaseFromEnv returns the base grace window, honoring
-// envWarningDelay when it holds a valid positive Go duration. Invalid or
-// non-positive values fall back to defaultWarningDelayBase.
-func warningDelayBaseFromEnv() time.Duration {
-	val := os.Getenv(envWarningDelay)
-	if val == "" {
-		return defaultWarningDelayBase
-	}
-	d, err := time.ParseDuration(val)
-	if err != nil {
-		log.Warnf("invalid %s value %q, using default %v: %v", envWarningDelay, val, defaultWarningDelayBase, err)
-		return defaultWarningDelayBase
-	}
-	if d <= 0 {
-		log.Warnf("%s must be positive, got %v, using default %v", envWarningDelay, d, defaultWarningDelayBase)
-		return defaultWarningDelayBase
-	}
-	return d
-}
-
 // warningDelay returns the grace window for the given selected-route
 // count. Scales gently: +1s per 100 routes, capped by
 // warningDelayBonusCap. Parallel handshakes mean handshake time grows
@@ -1230,7 +1204,7 @@ func (s *DefaultServer) groupHasImmediateUpstream(servers []netip.AddrPort, snap
 // in more than one handler.
 func (s *DefaultServer) collectUpstreamHealth() map[netip.AddrPort]UpstreamHealth {
 	merged := make(map[netip.AddrPort]UpstreamHealth)
-	for _, entry := range s.dnsMuxHandlers {
+	for _, entry := range s.dnsMuxMap {
 		reporter, ok := entry.handler.(upstreamHealthReporter)
 		if !ok {
 			continue

client/internal/dns/server_privileged_test.go

@@ -1,485 +0,0 @@
-//go:build privileged
-
-package dns
-
-import (
-	"context"
-	"fmt"
-	"net/netip"
-	"os"
-	"testing"
-
-	"github.com/golang/mock/gomock"
-	"github.com/miekg/dns"
-	"github.com/stretchr/testify/assert"
-	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
-
-	"github.com/netbirdio/netbird/client/iface"
-	pfmock "github.com/netbirdio/netbird/client/iface/mocks"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
-	"github.com/netbirdio/netbird/client/internal/dns/local"
-	"github.com/netbirdio/netbird/client/internal/dns/test"
-	"github.com/netbirdio/netbird/client/internal/peer"
-	"github.com/netbirdio/netbird/client/internal/stdnet"
-	nbdns "github.com/netbirdio/netbird/dns"
-)
-
-func TestUpdateDNSServer(t *testing.T) {
-
-	nameServers := []nbdns.NameServer{
-		{
-			IP:     netip.MustParseAddr("8.8.8.8"),
-			NSType: nbdns.UDPNameServerType,
-			Port:   53,
-		},
-		{
-			IP:     netip.MustParseAddr("8.8.4.4"),
-			NSType: nbdns.UDPNameServerType,
-			Port:   53,
-		},
-	}
-
-	testCases := []struct {
-		name                string
-		initUpstreamMap     []handlerWrapper
-		initLocalZones      []nbdns.CustomZone
-		initSerial          uint64
-		inputSerial         uint64
-		inputUpdate         nbdns.Config
-		shouldFail          bool
-		expectedUpstreamMap []handlerWrapper
-		expectedLocalQs     []dns.Question
-	}{
-		{
-			name:            "Initial Config Should Succeed",
-			initUpstreamMap: nil,
-			initSerial:      0,
-			inputSerial:     1,
-			inputUpdate: nbdns.Config{
-				ServiceEnable: true,
-				CustomZones: []nbdns.CustomZone{
-					{
-						Domain:  "netbird.cloud",
-						Records: zoneRecords,
-					},
-				},
-				NameServerGroups: []*nbdns.NameServerGroup{
-					{
-						Domains:     []string{"netbird.io"},
-						NameServers: nameServers,
-					},
-					{
-						NameServers: nameServers,
-						Primary:     true,
-					},
-				},
-			},
-			expectedUpstreamMap: []handlerWrapper{
-				{
-					domain:   "netbird.io",
-					priority: PriorityUpstream,
-				},
-				{
-					domain:   "netbird.cloud",
-					priority: PriorityLocal,
-				},
-				{
-					domain:   nbdns.RootZone,
-					priority: PriorityDefault,
-				},
-			},
-			expectedLocalQs: []dns.Question{{Name: "peera.netbird.cloud.", Qtype: dns.TypeA, Qclass: dns.ClassINET}},
-		},
-		{
-			name:           "New Config Should Succeed",
-			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
-			initUpstreamMap: []handlerWrapper{
-				{
-					domain:   "netbird.cloud",
-					handler:  &mockHandler{},
-					priority: PriorityUpstream,
-				},
-			},
-			initSerial:  0,
-			inputSerial: 1,
-			inputUpdate: nbdns.Config{
-				ServiceEnable: true,
-				CustomZones: []nbdns.CustomZone{
-					{
-						Domain:  "netbird.cloud",
-						Records: zoneRecords,
-					},
-				},
-				NameServerGroups: []*nbdns.NameServerGroup{
-					{
-						Domains:     []string{"netbird.io"},
-						NameServers: nameServers,
-					},
-				},
-			},
-			expectedUpstreamMap: []handlerWrapper{
-				{
-					domain:   "netbird.io",
-					priority: PriorityUpstream,
-				},
-				{
-					domain:   "netbird.cloud",
-					priority: PriorityLocal,
-				},
-			},
-			expectedLocalQs: []dns.Question{{Name: zoneRecords[0].Name, Qtype: 1, Qclass: 1}},
-		},
-		{
-			name:            "Smaller Config Serial Should Be Skipped",
-			initLocalZones:  []nbdns.CustomZone{},
-			initUpstreamMap: nil,
-			initSerial:      2,
-			inputSerial:     1,
-			shouldFail:      true,
-		},
-		{
-			name:            "Empty NS Group Domain Or Not Primary Element Should Fail",
-			initLocalZones:  []nbdns.CustomZone{},
-			initUpstreamMap: nil,
-			initSerial:      0,
-			inputSerial:     1,
-			inputUpdate: nbdns.Config{
-				ServiceEnable: true,
-				CustomZones: []nbdns.CustomZone{
-					{
-						Domain:  "netbird.cloud",
-						Records: zoneRecords,
-					},
-				},
-				NameServerGroups: []*nbdns.NameServerGroup{
-					{
-						NameServers: nameServers,
-					},
-				},
-			},
-			shouldFail: true,
-		},
-		{
-			name:            "Invalid NS Group Nameservers list Should Fail",
-			initLocalZones:  []nbdns.CustomZone{},
-			initUpstreamMap: nil,
-			initSerial:      0,
-			inputSerial:     1,
-			inputUpdate: nbdns.Config{
-				ServiceEnable: true,
-				CustomZones: []nbdns.CustomZone{
-					{
-						Domain:  "netbird.cloud",
-						Records: zoneRecords,
-					},
-				},
-				NameServerGroups: []*nbdns.NameServerGroup{
-					{
-						NameServers: nameServers,
-					},
-				},
-			},
-			shouldFail: true,
-		},
-		{
-			name:            "Invalid Custom Zone Records list Should Skip",
-			initLocalZones:  []nbdns.CustomZone{},
-			initUpstreamMap: nil,
-			initSerial:      0,
-			inputSerial:     1,
-			inputUpdate: nbdns.Config{
-				ServiceEnable: true,
-				CustomZones: []nbdns.CustomZone{
-					{
-						Domain: "netbird.cloud",
-					},
-				},
-				NameServerGroups: []*nbdns.NameServerGroup{
-					{
-						NameServers: nameServers,
-						Primary:     true,
-					},
-				},
-			},
-			expectedUpstreamMap: []handlerWrapper{{
-				domain:   ".",
-				priority: PriorityDefault,
-			}},
-		},
-		{
-			name:           "Empty Config Should Succeed and Clean Maps",
-			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
-			initUpstreamMap: []handlerWrapper{
-				{
-					domain:   zoneRecords[0].Name,
-					handler:  &mockHandler{},
-					priority: PriorityUpstream,
-				},
-			},
-			initSerial:          0,
-			inputSerial:         1,
-			inputUpdate:         nbdns.Config{ServiceEnable: true},
-			expectedUpstreamMap: nil,
-			expectedLocalQs:     []dns.Question{},
-		},
-		{
-			name:           "Disabled Service Should clean map",
-			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
-			initUpstreamMap: []handlerWrapper{
-				{
-					domain:   zoneRecords[0].Name,
-					handler:  &mockHandler{},
-					priority: PriorityUpstream,
-				},
-			},
-			initSerial:          0,
-			inputSerial:         1,
-			inputUpdate:         nbdns.Config{ServiceEnable: false},
-			expectedUpstreamMap: nil,
-			expectedLocalQs:     []dns.Question{},
-		},
-	}
-
-	for n, testCase := range testCases {
-		t.Run(testCase.name, func(t *testing.T) {
-			privKey, _ := wgtypes.GenerateKey()
-			newNet, err := stdnet.NewNet(context.Background(), nil)
-			if err != nil {
-				t.Fatal(err)
-			}
-
-			opts := iface.WGIFaceOpts{
-				IFaceName:    fmt.Sprintf("utun230%d", n),
-				Address:      wgaddr.MustParseWGAddress(fmt.Sprintf("100.66.100.%d/32", n+1)),
-				WGPort:       33100,
-				WGPrivKey:    privKey.String(),
-				MTU:          iface.DefaultMTU,
-				TransportNet: newNet,
-			}
-
-			wgIface, err := iface.NewWGIFace(opts)
-			if err != nil {
-				t.Fatal(err)
-			}
-			err = wgIface.Create()
-			if err != nil {
-				t.Fatal(err)
-			}
-			defer func() {
-				err = wgIface.Close()
-				if err != nil {
-					t.Log(err)
-				}
-			}()
-			dnsServer, err := NewDefaultServer(context.Background(), DefaultServerConfig{
-				WgInterface:    wgIface,
-				CustomAddress:  "",
-				StatusRecorder: peer.NewRecorder("mgm"),
-				StateManager:   nil,
-				DisableSys:     false,
-			})
-			if err != nil {
-				t.Fatal(err)
-			}
-			err = dnsServer.Initialize()
-			if err != nil {
-				t.Fatal(err)
-			}
-			defer func() {
-				err = dnsServer.hostManager.restoreHostDNS()
-				if err != nil {
-					t.Log(err)
-				}
-			}()
-
-			dnsServer.dnsMuxHandlers = testCase.initUpstreamMap
-			dnsServer.localResolver.Update(testCase.initLocalZones)
-			dnsServer.updateSerial = testCase.initSerial
-
-			err = dnsServer.UpdateDNSServer(testCase.inputSerial, testCase.inputUpdate)
-			if err != nil {
-				if testCase.shouldFail {
-					return
-				}
-				t.Fatalf("update dns server should not fail, got error: %v", err)
-			}
-
-			if len(dnsServer.dnsMuxHandlers) != len(testCase.expectedUpstreamMap) {
-				t.Fatalf("update upstream failed, map size is different than expected, want %d, got %d", len(testCase.expectedUpstreamMap), len(dnsServer.dnsMuxHandlers))
-			}
-
-			for _, expected := range testCase.expectedUpstreamMap {
-				found := false
-				for _, got := range dnsServer.dnsMuxHandlers {
-					if got.domain == expected.domain && got.priority == expected.priority {
-						found = true
-						break
-					}
-				}
-				if !found {
-					t.Fatalf("update upstream failed, handler for domain=%s priority=%d not found in dnsMuxHandlers: %#v", expected.domain, expected.priority, dnsServer.dnsMuxHandlers)
-				}
-			}
-
-			var responseMSG *dns.Msg
-			responseWriter := &test.MockResponseWriter{
-				WriteMsgFunc: func(m *dns.Msg) error {
-					responseMSG = m
-					return nil
-				},
-			}
-			for _, q := range testCase.expectedLocalQs {
-				dnsServer.localResolver.ServeDNS(responseWriter, &dns.Msg{
-					Question: []dns.Question{q},
-				})
-			}
-
-			if len(testCase.expectedLocalQs) > 0 {
-				assert.NotNil(t, responseMSG, "response message should not be nil")
-				assert.Equal(t, dns.RcodeSuccess, responseMSG.Rcode, "response code should be success")
-				assert.NotEmpty(t, responseMSG.Answer, "response message should have answers")
-			}
-		})
-	}
-}
-
-func TestDNSFakeResolverHandleUpdates(t *testing.T) {
-	ov := os.Getenv("NB_WG_KERNEL_DISABLED")
-	defer t.Setenv("NB_WG_KERNEL_DISABLED", ov)
-
-	t.Setenv("NB_WG_KERNEL_DISABLED", "true")
-	newNet, err := stdnet.NewNet(context.Background(), []string{"utun2301"})
-	if err != nil {
-		t.Errorf("create stdnet: %v", err)
-		return
-	}
-
-	privKey, _ := wgtypes.GeneratePrivateKey()
-	opts := iface.WGIFaceOpts{
-		IFaceName:    "utun2301",
-		Address:      wgaddr.MustParseWGAddress("100.66.100.1/32"),
-		WGPort:       33100,
-		WGPrivKey:    privKey.String(),
-		MTU:          iface.DefaultMTU,
-		TransportNet: newNet,
-	}
-	wgIface, err := iface.NewWGIFace(opts)
-	if err != nil {
-		t.Errorf("build interface wireguard: %v", err)
-		return
-	}
-
-	err = wgIface.Create()
-	if err != nil {
-		t.Errorf("create and init wireguard interface: %v", err)
-		return
-	}
-	defer func() {
-		if err = wgIface.Close(); err != nil {
-			t.Logf("close wireguard interface: %v", err)
-		}
-	}()
-
-	ctrl := gomock.NewController(t)
-	defer ctrl.Finish()
-
-	packetfilter := pfmock.NewMockPacketFilter(ctrl)
-	packetfilter.EXPECT().FilterOutbound(gomock.Any(), gomock.Any()).AnyTimes()
-	packetfilter.EXPECT().SetUDPPacketHook(gomock.Any(), gomock.Any(), gomock.Any()).AnyTimes()
-	packetfilter.EXPECT().SetTCPPacketHook(gomock.Any(), gomock.Any(), gomock.Any()).AnyTimes()
-
-	if err := wgIface.SetFilter(packetfilter); err != nil {
-		t.Errorf("set packet filter: %v", err)
-		return
-	}
-
-	dnsServer, err := NewDefaultServer(context.Background(), DefaultServerConfig{
-		WgInterface:    wgIface,
-		CustomAddress:  "",
-		StatusRecorder: peer.NewRecorder("mgm"),
-		StateManager:   nil,
-		DisableSys:     false,
-	})
-	if err != nil {
-		t.Errorf("create DNS server: %v", err)
-		return
-	}
-
-	err = dnsServer.Initialize()
-	if err != nil {
-		t.Errorf("run DNS server: %v", err)
-		return
-	}
-	defer func() {
-		if err = dnsServer.hostManager.restoreHostDNS(); err != nil {
-			t.Logf("restore DNS settings on the host: %v", err)
-			return
-		}
-	}()
-
-	dnsServer.dnsMuxHandlers = []handlerWrapper{
-		{
-			domain:   zoneRecords[0].Name,
-			handler:  &local.Resolver{},
-			priority: PriorityUpstream,
-		},
-	}
-	dnsServer.localResolver.Update([]nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}})
-	dnsServer.updateSerial = 0
-
-	nameServers := []nbdns.NameServer{
-		{
-			IP:     netip.MustParseAddr("8.8.8.8"),
-			NSType: nbdns.UDPNameServerType,
-			Port:   53,
-		},
-		{
-			IP:     netip.MustParseAddr("8.8.4.4"),
-			NSType: nbdns.UDPNameServerType,
-			Port:   53,
-		},
-	}
-
-	update := nbdns.Config{
-		ServiceEnable: true,
-		CustomZones: []nbdns.CustomZone{
-			{
-				Domain:  "netbird.cloud",
-				Records: zoneRecords,
-			},
-		},
-		NameServerGroups: []*nbdns.NameServerGroup{
-			{
-				Domains:     []string{"netbird.io"},
-				NameServers: nameServers,
-			},
-			{
-				NameServers: nameServers,
-				Primary:     true,
-			},
-		},
-	}
-
-	// Start the server with regular configuration
-	if err := dnsServer.UpdateDNSServer(1, update); err != nil {
-		t.Fatalf("update dns server should not fail, got error: %v", err)
-		return
-	}
-
-	update2 := update
-	update2.ServiceEnable = false
-	// Disable the server, stop the listener
-	if err := dnsServer.UpdateDNSServer(2, update2); err != nil {
-		t.Fatalf("update dns server should not fail, got error: %v", err)
-		return
-	}
-
-	update3 := update2
-	update3.NameServerGroups = update3.NameServerGroups[:1]
-	// But service still get updates and we checking that we handle
-	// internal state in the right way
-	if err := dnsServer.UpdateDNSServer(3, update3); err != nil {
-		t.Fatalf("update dns server should not fail, got error: %v", err)
-		return
-	}
-}

client/internal/dns/server_test.go

@@ -10,6 +10,7 @@ import (
 	"testing"
 	"time"
 
+	"github.com/golang/mock/gomock"
 	"github.com/miekg/dns"
 	log "github.com/sirupsen/logrus"
 	"github.com/stretchr/testify/assert"
@@ -22,6 +23,7 @@ import (
 	"github.com/netbirdio/netbird/client/iface"
 	"github.com/netbirdio/netbird/client/iface/configurer"
 	"github.com/netbirdio/netbird/client/iface/device"
+	pfmock "github.com/netbirdio/netbird/client/iface/mocks"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/internal/dns/local"
 	"github.com/netbirdio/netbird/client/internal/dns/test"
@@ -102,6 +104,481 @@ func init() {
 	formatter.SetTextFormatter(log.StandardLogger())
 }
 
+func generateDummyHandler(d string, servers []nbdns.NameServer) *upstreamResolverBase {
+	var srvs []netip.AddrPort
+	for _, srv := range servers {
+		srvs = append(srvs, srv.AddrPort())
+	}
+	u := &upstreamResolverBase{
+		domain: domain.Domain(d),
+		cancel: func() {},
+	}
+	u.addRace(srvs)
+	return u
+}
+
+func TestUpdateDNSServer(t *testing.T) {
+
+	nameServers := []nbdns.NameServer{
+		{
+			IP:     netip.MustParseAddr("8.8.8.8"),
+			NSType: nbdns.UDPNameServerType,
+			Port:   53,
+		},
+		{
+			IP:     netip.MustParseAddr("8.8.4.4"),
+			NSType: nbdns.UDPNameServerType,
+			Port:   53,
+		},
+	}
+
+	dummyHandler := local.NewResolver()
+
+	testCases := []struct {
+		name                string
+		initUpstreamMap     registeredHandlerMap
+		initLocalZones      []nbdns.CustomZone
+		initSerial          uint64
+		inputSerial         uint64
+		inputUpdate         nbdns.Config
+		shouldFail          bool
+		expectedUpstreamMap registeredHandlerMap
+		expectedLocalQs     []dns.Question
+	}{
+		{
+			name:            "Initial Config Should Succeed",
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
+			inputUpdate: nbdns.Config{
+				ServiceEnable: true,
+				CustomZones: []nbdns.CustomZone{
+					{
+						Domain:  "netbird.cloud",
+						Records: zoneRecords,
+					},
+				},
+				NameServerGroups: []*nbdns.NameServerGroup{
+					{
+						Domains:     []string{"netbird.io"},
+						NameServers: nameServers,
+					},
+					{
+						NameServers: nameServers,
+						Primary:     true,
+					},
+				},
+			},
+			expectedUpstreamMap: registeredHandlerMap{
+				generateDummyHandler("netbird.io", nameServers).ID(): handlerWrapper{
+					domain:   "netbird.io",
+					handler:  dummyHandler,
+					priority: PriorityUpstream,
+				},
+				dummyHandler.ID(): handlerWrapper{
+					domain:   "netbird.cloud",
+					handler:  dummyHandler,
+					priority: PriorityLocal,
+				},
+				generateDummyHandler(".", nameServers).ID(): handlerWrapper{
+					domain:   nbdns.RootZone,
+					handler:  dummyHandler,
+					priority: PriorityDefault,
+				},
+			},
+			expectedLocalQs: []dns.Question{{Name: "peera.netbird.cloud.", Qtype: dns.TypeA, Qclass: dns.ClassINET}},
+		},
+		{
+			name:           "New Config Should Succeed",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
+			initUpstreamMap: registeredHandlerMap{
+				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
+					domain:   "netbird.cloud",
+					handler:  dummyHandler,
+					priority: PriorityUpstream,
+				},
+			},
+			initSerial:  0,
+			inputSerial: 1,
+			inputUpdate: nbdns.Config{
+				ServiceEnable: true,
+				CustomZones: []nbdns.CustomZone{
+					{
+						Domain:  "netbird.cloud",
+						Records: zoneRecords,
+					},
+				},
+				NameServerGroups: []*nbdns.NameServerGroup{
+					{
+						Domains:     []string{"netbird.io"},
+						NameServers: nameServers,
+					},
+				},
+			},
+			expectedUpstreamMap: registeredHandlerMap{
+				generateDummyHandler("netbird.io", nameServers).ID(): handlerWrapper{
+					domain:   "netbird.io",
+					handler:  dummyHandler,
+					priority: PriorityUpstream,
+				},
+				"local-resolver": handlerWrapper{
+					domain:   "netbird.cloud",
+					handler:  dummyHandler,
+					priority: PriorityLocal,
+				},
+			},
+			expectedLocalQs: []dns.Question{{Name: zoneRecords[0].Name, Qtype: 1, Qclass: 1}},
+		},
+		{
+			name:            "Smaller Config Serial Should Be Skipped",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      2,
+			inputSerial:     1,
+			shouldFail:      true,
+		},
+		{
+			name:            "Empty NS Group Domain Or Not Primary Element Should Fail",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
+			inputUpdate: nbdns.Config{
+				ServiceEnable: true,
+				CustomZones: []nbdns.CustomZone{
+					{
+						Domain:  "netbird.cloud",
+						Records: zoneRecords,
+					},
+				},
+				NameServerGroups: []*nbdns.NameServerGroup{
+					{
+						NameServers: nameServers,
+					},
+				},
+			},
+			shouldFail: true,
+		},
+		{
+			name:            "Invalid NS Group Nameservers list Should Fail",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
+			inputUpdate: nbdns.Config{
+				ServiceEnable: true,
+				CustomZones: []nbdns.CustomZone{
+					{
+						Domain:  "netbird.cloud",
+						Records: zoneRecords,
+					},
+				},
+				NameServerGroups: []*nbdns.NameServerGroup{
+					{
+						NameServers: nameServers,
+					},
+				},
+			},
+			shouldFail: true,
+		},
+		{
+			name:            "Invalid Custom Zone Records list Should Skip",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
+			inputUpdate: nbdns.Config{
+				ServiceEnable: true,
+				CustomZones: []nbdns.CustomZone{
+					{
+						Domain: "netbird.cloud",
+					},
+				},
+				NameServerGroups: []*nbdns.NameServerGroup{
+					{
+						NameServers: nameServers,
+						Primary:     true,
+					},
+				},
+			},
+			expectedUpstreamMap: registeredHandlerMap{generateDummyHandler(".", nameServers).ID(): handlerWrapper{
+				domain:   ".",
+				handler:  dummyHandler,
+				priority: PriorityDefault,
+			}},
+		},
+		{
+			name:           "Empty Config Should Succeed and Clean Maps",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
+			initUpstreamMap: registeredHandlerMap{
+				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
+					domain:   zoneRecords[0].Name,
+					handler:  dummyHandler,
+					priority: PriorityUpstream,
+				},
+			},
+			initSerial:          0,
+			inputSerial:         1,
+			inputUpdate:         nbdns.Config{ServiceEnable: true},
+			expectedUpstreamMap: make(registeredHandlerMap),
+			expectedLocalQs:     []dns.Question{},
+		},
+		{
+			name:           "Disabled Service Should clean map",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
+			initUpstreamMap: registeredHandlerMap{
+				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
+					domain:   zoneRecords[0].Name,
+					handler:  dummyHandler,
+					priority: PriorityUpstream,
+				},
+			},
+			initSerial:          0,
+			inputSerial:         1,
+			inputUpdate:         nbdns.Config{ServiceEnable: false},
+			expectedUpstreamMap: make(registeredHandlerMap),
+			expectedLocalQs:     []dns.Question{},
+		},
+	}
+
+	for n, testCase := range testCases {
+		t.Run(testCase.name, func(t *testing.T) {
+			privKey, _ := wgtypes.GenerateKey()
+			newNet, err := stdnet.NewNet(context.Background(), nil)
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			opts := iface.WGIFaceOpts{
+				IFaceName:    fmt.Sprintf("utun230%d", n),
+				Address:      wgaddr.MustParseWGAddress(fmt.Sprintf("100.66.100.%d/32", n+1)),
+				WGPort:       33100,
+				WGPrivKey:    privKey.String(),
+				MTU:          iface.DefaultMTU,
+				TransportNet: newNet,
+			}
+
+			wgIface, err := iface.NewWGIFace(opts)
+			if err != nil {
+				t.Fatal(err)
+			}
+			err = wgIface.Create()
+			if err != nil {
+				t.Fatal(err)
+			}
+			defer func() {
+				err = wgIface.Close()
+				if err != nil {
+					t.Log(err)
+				}
+			}()
+			dnsServer, err := NewDefaultServer(context.Background(), DefaultServerConfig{
+				WgInterface:    wgIface,
+				CustomAddress:  "",
+				StatusRecorder: peer.NewRecorder("mgm"),
+				StateManager:   nil,
+				DisableSys:     false,
+			})
+			if err != nil {
+				t.Fatal(err)
+			}
+			err = dnsServer.Initialize()
+			if err != nil {
+				t.Fatal(err)
+			}
+			defer func() {
+				err = dnsServer.hostManager.restoreHostDNS()
+				if err != nil {
+					t.Log(err)
+				}
+			}()
+
+			dnsServer.dnsMuxMap = testCase.initUpstreamMap
+			dnsServer.localResolver.Update(testCase.initLocalZones)
+			dnsServer.updateSerial = testCase.initSerial
+
+			err = dnsServer.UpdateDNSServer(testCase.inputSerial, testCase.inputUpdate)
+			if err != nil {
+				if testCase.shouldFail {
+					return
+				}
+				t.Fatalf("update dns server should not fail, got error: %v", err)
+			}
+
+			if len(dnsServer.dnsMuxMap) != len(testCase.expectedUpstreamMap) {
+				t.Fatalf("update upstream failed, map size is different than expected, want %d, got %d", len(testCase.expectedUpstreamMap), len(dnsServer.dnsMuxMap))
+			}
+
+			for key := range testCase.expectedUpstreamMap {
+				_, found := dnsServer.dnsMuxMap[key]
+				if !found {
+					t.Fatalf("update upstream failed, key %s was not found in the dnsMuxMap: %#v", key, dnsServer.dnsMuxMap)
+				}
+			}
+
+			var responseMSG *dns.Msg
+			responseWriter := &test.MockResponseWriter{
+				WriteMsgFunc: func(m *dns.Msg) error {
+					responseMSG = m
+					return nil
+				},
+			}
+			for _, q := range testCase.expectedLocalQs {
+				dnsServer.localResolver.ServeDNS(responseWriter, &dns.Msg{
+					Question: []dns.Question{q},
+				})
+			}
+
+			if len(testCase.expectedLocalQs) > 0 {
+				assert.NotNil(t, responseMSG, "response message should not be nil")
+				assert.Equal(t, dns.RcodeSuccess, responseMSG.Rcode, "response code should be success")
+				assert.NotEmpty(t, responseMSG.Answer, "response message should have answers")
+			}
+		})
+	}
+}
+
+func TestDNSFakeResolverHandleUpdates(t *testing.T) {
+	ov := os.Getenv("NB_WG_KERNEL_DISABLED")
+	defer t.Setenv("NB_WG_KERNEL_DISABLED", ov)
+
+	t.Setenv("NB_WG_KERNEL_DISABLED", "true")
+	newNet, err := stdnet.NewNet(context.Background(), []string{"utun2301"})
+	if err != nil {
+		t.Errorf("create stdnet: %v", err)
+		return
+	}
+
+	privKey, _ := wgtypes.GeneratePrivateKey()
+	opts := iface.WGIFaceOpts{
+		IFaceName:    "utun2301",
+		Address:      wgaddr.MustParseWGAddress("100.66.100.1/32"),
+		WGPort:       33100,
+		WGPrivKey:    privKey.String(),
+		MTU:          iface.DefaultMTU,
+		TransportNet: newNet,
+	}
+	wgIface, err := iface.NewWGIFace(opts)
+	if err != nil {
+		t.Errorf("build interface wireguard: %v", err)
+		return
+	}
+
+	err = wgIface.Create()
+	if err != nil {
+		t.Errorf("create and init wireguard interface: %v", err)
+		return
+	}
+	defer func() {
+		if err = wgIface.Close(); err != nil {
+			t.Logf("close wireguard interface: %v", err)
+		}
+	}()
+
+	ctrl := gomock.NewController(t)
+	defer ctrl.Finish()
+
+	packetfilter := pfmock.NewMockPacketFilter(ctrl)
+	packetfilter.EXPECT().FilterOutbound(gomock.Any(), gomock.Any()).AnyTimes()
+	packetfilter.EXPECT().SetUDPPacketHook(gomock.Any(), gomock.Any(), gomock.Any()).AnyTimes()
+	packetfilter.EXPECT().SetTCPPacketHook(gomock.Any(), gomock.Any(), gomock.Any()).AnyTimes()
+
+	if err := wgIface.SetFilter(packetfilter); err != nil {
+		t.Errorf("set packet filter: %v", err)
+		return
+	}
+
+	dnsServer, err := NewDefaultServer(context.Background(), DefaultServerConfig{
+		WgInterface:    wgIface,
+		CustomAddress:  "",
+		StatusRecorder: peer.NewRecorder("mgm"),
+		StateManager:   nil,
+		DisableSys:     false,
+	})
+	if err != nil {
+		t.Errorf("create DNS server: %v", err)
+		return
+	}
+
+	err = dnsServer.Initialize()
+	if err != nil {
+		t.Errorf("run DNS server: %v", err)
+		return
+	}
+	defer func() {
+		if err = dnsServer.hostManager.restoreHostDNS(); err != nil {
+			t.Logf("restore DNS settings on the host: %v", err)
+			return
+		}
+	}()
+
+	dnsServer.dnsMuxMap = registeredHandlerMap{
+		"id1": handlerWrapper{
+			domain:   zoneRecords[0].Name,
+			handler:  &local.Resolver{},
+			priority: PriorityUpstream,
+		},
+	}
+	dnsServer.localResolver.Update([]nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}})
+	dnsServer.updateSerial = 0
+
+	nameServers := []nbdns.NameServer{
+		{
+			IP:     netip.MustParseAddr("8.8.8.8"),
+			NSType: nbdns.UDPNameServerType,
+			Port:   53,
+		},
+		{
+			IP:     netip.MustParseAddr("8.8.4.4"),
+			NSType: nbdns.UDPNameServerType,
+			Port:   53,
+		},
+	}
+
+	update := nbdns.Config{
+		ServiceEnable: true,
+		CustomZones: []nbdns.CustomZone{
+			{
+				Domain:  "netbird.cloud",
+				Records: zoneRecords,
+			},
+		},
+		NameServerGroups: []*nbdns.NameServerGroup{
+			{
+				Domains:     []string{"netbird.io"},
+				NameServers: nameServers,
+			},
+			{
+				NameServers: nameServers,
+				Primary:     true,
+			},
+		},
+	}
+
+	// Start the server with regular configuration
+	if err := dnsServer.UpdateDNSServer(1, update); err != nil {
+		t.Fatalf("update dns server should not fail, got error: %v", err)
+		return
+	}
+
+	update2 := update
+	update2.ServiceEnable = false
+	// Disable the server, stop the listener
+	if err := dnsServer.UpdateDNSServer(2, update2); err != nil {
+		t.Fatalf("update dns server should not fail, got error: %v", err)
+		return
+	}
+
+	update3 := update2
+	update3.NameServerGroups = update3.NameServerGroups[:1]
+	// But service still get updates and we checking that we handle
+	// internal state in the right way
+	if err := dnsServer.UpdateDNSServer(3, update3); err != nil {
+		t.Fatalf("update dns server should not fail, got error: %v", err)
+		return
+	}
+}
+
 func TestDNSServerStartStop(t *testing.T) {
 	testCases := []struct {
 		name     string
@@ -552,15 +1029,15 @@ func (m *mockService) RegisterMux(string, dns.Handler) {}
 func (m *mockService) DeregisterMux(string)            {}
 
 func TestDefaultServer_UpdateMux(t *testing.T) {
-	baseMatchHandlers := []handlerWrapper{
-		{
+	baseMatchHandlers := registeredHandlerMap{
+		"upstream-group1": {
 			domain: "example.com",
 			handler: &mockHandler{
 				Id: "upstream-group1",
 			},
 			priority: PriorityUpstream,
 		},
-		{
+		"upstream-group2": {
 			domain: "example.com",
 			handler: &mockHandler{
 				Id: "upstream-group2",
@@ -569,15 +1046,15 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 		},
 	}
 
-	baseRootHandlers := []handlerWrapper{
-		{
+	baseRootHandlers := registeredHandlerMap{
+		"upstream-root1": {
 			domain: ".",
 			handler: &mockHandler{
 				Id: "upstream-root1",
 			},
 			priority: PriorityDefault,
 		},
-		{
+		"upstream-root2": {
 			domain: ".",
 			handler: &mockHandler{
 				Id: "upstream-root2",
@@ -586,22 +1063,22 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 		},
 	}
 
-	baseMixedHandlers := []handlerWrapper{
-		{
+	baseMixedHandlers := registeredHandlerMap{
+		"upstream-group1": {
 			domain: "example.com",
 			handler: &mockHandler{
 				Id: "upstream-group1",
 			},
 			priority: PriorityUpstream,
 		},
-		{
+		"upstream-group2": {
 			domain: "example.com",
 			handler: &mockHandler{
 				Id: "upstream-group2",
 			},
 			priority: PriorityUpstream - 1,
 		},
-		{
+		"upstream-other": {
 			domain: "other.com",
 			handler: &mockHandler{
 				Id: "upstream-other",
@@ -612,7 +1089,7 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 
 	tests := []struct {
 		name             string
-		initialHandlers  []handlerWrapper
+		initialHandlers  registeredHandlerMap
 		updates          []handlerWrapper
 		expectedHandlers map[string]string // map[HandlerID]domain
 		description      string
@@ -896,38 +1373,32 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			server := &DefaultServer{
-				dnsMuxHandlers: tt.initialHandlers,
-				handlerChain:   NewHandlerChain(),
-				service:        &mockService{},
+				dnsMuxMap:    tt.initialHandlers,
+				handlerChain: NewHandlerChain(),
+				service:      &mockService{},
 			}
 
 			// Perform the update
 			server.updateMux(tt.updates)
 
 			// Verify the results
-			assert.Equal(t, len(tt.expectedHandlers), len(server.dnsMuxHandlers),
+			assert.Equal(t, len(tt.expectedHandlers), len(server.dnsMuxMap),
 				"Number of handlers after update doesn't match expected")
 
 			// Check each expected handler
 			for id, expectedDomain := range tt.expectedHandlers {
-				var found *handlerWrapper
-				for i := range server.dnsMuxHandlers {
-					if server.dnsMuxHandlers[i].handler.ID() == types.HandlerID(id) {
-						found = &server.dnsMuxHandlers[i]
-						break
-					}
-				}
-				assert.NotNil(t, found, "Expected handler %s not found", id)
-				if found != nil {
-					assert.Equal(t, expectedDomain, found.domain,
+				handler, exists := server.dnsMuxMap[types.HandlerID(id)]
+				assert.True(t, exists, "Expected handler %s not found", id)
+				if exists {
+					assert.Equal(t, expectedDomain, handler.domain,
 						"Domain mismatch for handler %s", id)
 				}
 			}
 
 			// Verify no unexpected handlers exist
-			for _, entry := range server.dnsMuxHandlers {
-				_, expected := tt.expectedHandlers[string(entry.handler.ID())]
-				assert.True(t, expected, "Unexpected handler found: %s", entry.handler.ID())
+			for HandlerID := range server.dnsMuxMap {
+				_, expected := tt.expectedHandlers[string(HandlerID)]
+				assert.True(t, expected, "Unexpected handler found: %s", HandlerID)
 			}
 
 			// Verify the handlerChain state and order
@@ -942,7 +1413,7 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 
 				// Verify handler exists in mux
 				foundInMux := false
-				for _, muxEntry := range server.dnsMuxHandlers {
+				for _, muxEntry := range server.dnsMuxMap {
 					if chainEntry.Handler == muxEntry.handler &&
 						chainEntry.Priority == muxEntry.priority &&
 						chainEntry.Pattern == dns.Fqdn(muxEntry.domain) {
@@ -951,108 +1422,12 @@ func TestDefaultServer_UpdateMux(t *testing.T) {
 					}
 				}
 				assert.True(t, foundInMux,
-					"Handler in chain not found in dnsMuxHandlers")
+					"Handler in chain not found in dnsMuxMap")
 			}
 		})
 	}
 }
 
-// chainHasPattern reports whether the handler chain holds an entry registered
-// for the given fqdn pattern at the given priority.
-func chainHasPattern(s *DefaultServer, pattern string, priority int) bool {
-	for _, h := range s.handlerChain.handlers {
-		if h.OrigPattern == pattern && h.Priority == priority {
-			return true
-		}
-	}
-	return false
-}
-
-// TestDefaultServer_UpdateMux_SharedHandlerZoneRemoval verifies that updateMux
-// tracks each (handler, domain) registration independently when one handler
-// serves multiple zones. Every custom zone is served by the same handler
-// instance (the local resolver, whose ID is the constant "local-resolver"), so
-// removing one zone must deregister exactly that zone's chain entry and leave
-// the others in place. Tracking registrations by handler ID alone collapses all
-// zones onto one entry, leaving removed zones in the chain to answer
-// authoritatively with no records.
-func TestDefaultServer_UpdateMux_SharedHandlerZoneRemoval(t *testing.T) {
-	// One handler serves every custom zone, mirroring s.localResolver.
-	shared := &mockHandler{Id: "local-resolver"}
-
-	server := &DefaultServer{
-		handlerChain: NewHandlerChain(),
-		service:      &mockService{},
-	}
-
-	// Two custom zones under the same handler. The surviving zone is registered
-	// last, mirroring the management emission order.
-	server.updateMux([]handlerWrapper{
-		{domain: "userzone.test", handler: shared, priority: PriorityLocal},
-		{domain: "peerzone.test", handler: shared, priority: PriorityLocal},
-	})
-
-	require.True(t, chainHasPattern(server, "userzone.test.", PriorityLocal),
-		"userzone.test should be registered after the first update")
-	require.True(t, chainHasPattern(server, "peerzone.test.", PriorityLocal),
-		"peerzone.test should be registered after the first update")
-
-	// Remove one zone, keep the other.
-	server.updateMux([]handlerWrapper{
-		{domain: "peerzone.test", handler: shared, priority: PriorityLocal},
-	})
-
-	assert.True(t, chainHasPattern(server, "peerzone.test.", PriorityLocal),
-		"peerzone.test should remain after removing userzone.test")
-	assert.False(t, chainHasPattern(server, "userzone.test.", PriorityLocal),
-		"userzone.test handler must be deregistered, not leaked in the chain")
-}
-
-// TestDefaultServer_UpdateMux_PreservesLocalResolver verifies that updateMux
-// does not tear down the shared local resolver during reconfiguration. The
-// resolver is a process-lifetime singleton reused across config updates;
-// Stop() cancels its lookup context (breaking external CNAME-target
-// resolution) and clears its records. updateMux must deregister its chain
-// entries without stopping it. Records surviving a teardown update is the
-// observable proxy: Stop() would have cleared them.
-func TestDefaultServer_UpdateMux_PreservesLocalResolver(t *testing.T) {
-	resolver := local.NewResolver()
-	require.NoError(t, resolver.RegisterRecord(nbdns.SimpleRecord{
-		Name:  "peer.netbird.cloud.",
-		Type:  int(dns.TypeA),
-		Class: nbdns.DefaultClass,
-		TTL:   300,
-		RData: "10.0.0.1",
-	}))
-
-	server := &DefaultServer{
-		handlerChain:  NewHandlerChain(),
-		service:       &mockService{},
-		localResolver: resolver,
-	}
-
-	server.updateMux([]handlerWrapper{
-		{domain: "netbird.cloud", handler: resolver, priority: PriorityLocal},
-	})
-
-	// Remove the zone. The resolver must survive so its records and lookup
-	// context stay intact for the next registration.
-	server.updateMux(nil)
-
-	var response *dns.Msg
-	resolver.ServeDNS(&test.MockResponseWriter{
-		WriteMsgFunc: func(m *dns.Msg) error {
-			response = m
-			return nil
-		},
-	}, &dns.Msg{Question: []dns.Question{{Name: "peer.netbird.cloud.", Qtype: dns.TypeA, Qclass: dns.ClassINET}}})
-
-	require.NotNil(t, response, "local resolver should answer after teardown")
-	assert.Equal(t, dns.RcodeSuccess, response.Rcode,
-		"local resolver records must survive teardown; updateMux must not Stop() the shared resolver")
-	assert.NotEmpty(t, response.Answer, "answer should contain the surviving record")
-}
-
 func TestExtraDomains(t *testing.T) {
 	tests := []struct {
 		name                string
@@ -1674,6 +2049,7 @@ func TestBuildUpstreamHandler_MergesGroupsPerDomain(t *testing.T) {
 		localResolver: local.NewResolver(),
 		handlerChain:  NewHandlerChain(),
 		hostManager:   &noopHostConfigurator{},
+		dnsMuxMap:     make(registeredHandlerMap),
 	}
 
 	groups := []*nbdns.NameServerGroup{
@@ -1831,7 +2207,7 @@ func TestEvaluateNSGroupHealth(t *testing.T) {
 	}
 }
 
-// healthStubHandler is a minimal dnsMuxHandlers entry that exposes a fixed
+// healthStubHandler is a minimal dnsMuxMap entry that exposes a fixed
 // UpstreamHealth snapshot, letting tests drive recomputeNSGroupStates
 // without spinning up real handlers.
 type healthStubHandler struct {
@@ -1907,11 +2283,12 @@ func newProjTestFixture(t *testing.T) *projTestFixture {
 		ctx:              context.Background(),
 		wgInterface:      &mocWGIface{},
 		statusRecorder:   recorder,
+		dnsMuxMap:        make(registeredHandlerMap),
 		selectedRoutes:   func() route.HAMap { return fx.selected },
 		activeRoutes:     func() route.HAMap { return fx.active },
 		warningDelayBase: defaultWarningDelayBase,
 	}
-	fx.server.dnsMuxHandlers = []handlerWrapper{{domain: "example.com", handler: fx.stub, priority: PriorityUpstream}}
+	fx.server.dnsMuxMap["example.com"] = handlerWrapper{domain: "example.com", handler: fx.stub, priority: PriorityUpstream}
 
 	fx.server.mux.Lock()
 	fx.server.updateNSGroupStates([]*nbdns.NameServerGroup{fx.group})
@@ -2018,6 +2395,7 @@ func TestProjection_OverlayAddrNoRouteDelaysWarning(t *testing.T) {
 		ctx:              context.Background(),
 		wgInterface:      &mocWGIface{},
 		statusRecorder:   recorder,
+		dnsMuxMap:        make(registeredHandlerMap),
 		selectedRoutes:   func() route.HAMap { return nil },
 		activeRoutes:     func() route.HAMap { return nil },
 		warningDelayBase: 50 * time.Millisecond,
@@ -2029,7 +2407,7 @@ func TestProjection_OverlayAddrNoRouteDelaysWarning(t *testing.T) {
 	stub := &healthStubHandler{health: map[netip.AddrPort]UpstreamHealth{
 		overlayPeer: {LastFail: time.Now(), LastErr: "timeout"},
 	}}
-	server.dnsMuxHandlers = []handlerWrapper{{domain: "example.com", handler: stub, priority: PriorityUpstream}}
+	server.dnsMuxMap["example.com"] = handlerWrapper{domain: "example.com", handler: stub, priority: PriorityUpstream}
 
 	server.mux.Lock()
 	server.updateNSGroupStates([]*nbdns.NameServerGroup{group})
@@ -2066,6 +2444,7 @@ func TestProjection_StopClearsHealthState(t *testing.T) {
 		service:           NewServiceViaMemory(wgIface),
 		hostManager:       &noopHostConfigurator{},
 		extraDomains:      map[domain.Domain]int{},
+		dnsMuxMap:         make(registeredHandlerMap),
 		statusRecorder:    peer.NewRecorder("mgm"),
 		selectedRoutes:    func() route.HAMap { return nil },
 		activeRoutes:      func() route.HAMap { return nil },
@@ -2080,7 +2459,7 @@ func TestProjection_StopClearsHealthState(t *testing.T) {
 		NameServers: []nbdns.NameServer{{IP: srv.Addr(), NSType: nbdns.UDPNameServerType, Port: int(srv.Port())}},
 	}
 	stub := &healthStubHandler{health: map[netip.AddrPort]UpstreamHealth{srv: {LastOk: time.Now()}}}
-	server.dnsMuxHandlers = []handlerWrapper{{domain: "example.com", handler: stub, priority: PriorityUpstream}}
+	server.dnsMuxMap["example.com"] = handlerWrapper{domain: "example.com", handler: stub, priority: PriorityUpstream}
 
 	server.mux.Lock()
 	server.updateNSGroupStates([]*nbdns.NameServerGroup{group})
@@ -2105,32 +2484,6 @@ func TestProjection_StopClearsHealthState(t *testing.T) {
 // rule 3: startup failures while the peer is handshaking, then the peer
 // comes up and a query succeeds before the grace window elapses. No
 // warning should ever have fired, and no recovery either.
-func TestWarningDelayBaseFromEnv(t *testing.T) {
-	tests := []struct {
-		name string
-		set  bool
-		val  string
-		want time.Duration
-	}{
-		{name: "unset uses default", set: false, want: defaultWarningDelayBase},
-		{name: "valid override", set: true, val: "90s", want: 90 * time.Second},
-		{name: "valid minutes", set: true, val: "2m", want: 2 * time.Minute},
-		{name: "invalid falls back", set: true, val: "notaduration", want: defaultWarningDelayBase},
-		{name: "zero falls back", set: true, val: "0s", want: defaultWarningDelayBase},
-		{name: "negative falls back", set: true, val: "-30s", want: defaultWarningDelayBase},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			t.Setenv(envWarningDelay, tc.val)
-			if !tc.set {
-				os.Unsetenv(envWarningDelay)
-			}
-			assert.Equal(t, tc.want, warningDelayBaseFromEnv(), "grace window base")
-		})
-	}
-}
-
 func TestProjection_OverlayRecoversDuringGrace(t *testing.T) {
 	fx := newProjTestFixture(t)
 	fx.server.warningDelayBase = 200 * time.Millisecond
@@ -2242,6 +2595,7 @@ func TestProjection_MixedGroupEmitsImmediately(t *testing.T) {
 	server := &DefaultServer{
 		ctx:              context.Background(),
 		statusRecorder:   recorder,
+		dnsMuxMap:        make(registeredHandlerMap),
 		selectedRoutes:   func() route.HAMap { return overlayMap },
 		activeRoutes:     func() route.HAMap { return nil },
 		warningDelayBase: time.Hour,
@@ -2259,7 +2613,7 @@ func TestProjection_MixedGroupEmitsImmediately(t *testing.T) {
 			overlay: {LastFail: time.Now(), LastErr: "timeout"},
 		},
 	}
-	server.dnsMuxHandlers = []handlerWrapper{{domain: "example.com", handler: stub, priority: PriorityUpstream}}
+	server.dnsMuxMap["example.com"] = handlerWrapper{domain: "example.com", handler: stub, priority: PriorityUpstream}
 
 	server.mux.Lock()
 	server.updateNSGroupStates([]*nbdns.NameServerGroup{group})
@@ -2286,6 +2640,7 @@ func TestDNSLoopPrevention(t *testing.T) {
 		localResolver: local.NewResolver(),
 		handlerChain:  NewHandlerChain(),
 		hostManager:   &noopHostConfigurator{},
+		dnsMuxMap:     make(registeredHandlerMap),
 	}
 
 	tests := []struct {

client/internal/dns/upstream.go

@@ -443,32 +443,29 @@ func (u *upstreamResolverBase) queryUpstream(parentCtx context.Context, r *dns.M
 		return raceResult{}, &upstreamFailure{upstream: upstream, reason: "no response"}
 	}
 
-	// A valid response means the upstream is reachable, whatever the Rcode.
-	u.markUpstreamOk(upstream)
-
 	proto := ""
 	if upstreamProto != nil {
 		proto = upstreamProto.protocol
 	}
 
 	if rm.Rcode == dns.RcodeServerFailure || rm.Rcode == dns.RcodeRefused {
-		// SERVFAIL and REFUSED are per-question outcomes (DNSSEC-bogus names,
-		// refused zones, transient recursion errors), not reachability
-		// problems: fail over for a better answer but keep the upstream healthy.
 		if code, ok := nonRetryableEDE(rm); ok {
 			if !hadEdns {
-				resutil.StripOPT(rm)
+				stripOPT(rm)
 			}
+			u.markUpstreamOk(upstream)
 			return raceResult{msg: rm, upstream: upstream, protocol: proto, ede: edeName(code)}, nil
 		}
 		reason := dns.RcodeToString[rm.Rcode]
+		u.markUpstreamFail(upstream, reason)
 		return raceResult{}, &upstreamFailure{upstream: upstream, reason: reason}
 	}
 
 	if !hadEdns {
-		resutil.StripOPT(rm)
+		stripOPT(rm)
 	}
 
+	u.markUpstreamOk(upstream)
 	return raceResult{msg: rm, upstream: upstream, protocol: proto}, nil
 }
 
@@ -523,6 +520,22 @@ func upstreamUDPSize() uint16 {
 	return dns.MinMsgSize
 }
 
+// stripOPT removes any OPT pseudo-RRs from the response's Extra section so
+// the response complies with RFC 6891 when the client did not advertise EDNS0.
+func stripOPT(rm *dns.Msg) {
+	if len(rm.Extra) == 0 {
+		return
+	}
+	out := rm.Extra[:0]
+	for _, rr := range rm.Extra {
+		if _, ok := rr.(*dns.OPT); ok {
+			continue
+		}
+		out = append(out, rr)
+	}
+	rm.Extra = out
+}
+
 func (u *upstreamResolverBase) handleUpstreamError(err error, upstream netip.AddrPort, startTime time.Time) *upstreamFailure {
 	if !errors.Is(err, context.DeadlineExceeded) && !isTimeout(err) {
 		return &upstreamFailure{upstream: upstream, reason: err.Error()}

client/internal/dns/upstream_test.go

@@ -517,78 +517,6 @@ func TestUpstreamResolver_HealthTracking(t *testing.T) {
 	assert.NotContains(t, health, bad, "sibling upstream should not be queried when primary answers")
 }
 
-// TestUpstreamResolver_HealthTracking_ResponseMeansReachable verifies that an
-// upstream which answers with SERVFAIL or REFUSED is recorded as healthy:
-// those are per-question outcomes from a reachable server and must not mark
-// the upstream unhealthy. Only transport failures (timeouts) do.
-func TestUpstreamResolver_HealthTracking_ResponseMeansReachable(t *testing.T) {
-	a := netip.MustParseAddrPort("192.0.2.10:53")
-	b := netip.MustParseAddrPort("192.0.2.11:53")
-	timeoutErr := &net.OpError{Op: "read", Err: fmt.Errorf("i/o timeout")}
-
-	tests := []struct {
-		name        string
-		respA       mockUpstreamResponse
-		respB       mockUpstreamResponse
-		wantHealthy bool
-	}{
-		{
-			name:        "both SERVFAIL are reachable",
-			respA:       mockUpstreamResponse{msg: buildMockResponse(dns.RcodeServerFailure, "")},
-			respB:       mockUpstreamResponse{msg: buildMockResponse(dns.RcodeServerFailure, "")},
-			wantHealthy: true,
-		},
-		{
-			name:        "both REFUSED are reachable",
-			respA:       mockUpstreamResponse{msg: buildMockResponse(dns.RcodeRefused, "")},
-			respB:       mockUpstreamResponse{msg: buildMockResponse(dns.RcodeRefused, "")},
-			wantHealthy: true,
-		},
-		{
-			name:        "timeout marks unhealthy",
-			respA:       mockUpstreamResponse{err: timeoutErr},
-			respB:       mockUpstreamResponse{err: timeoutErr},
-			wantHealthy: false,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			mockClient := &mockUpstreamResolverPerServer{
-				responses: map[string]mockUpstreamResponse{
-					a.String(): tc.respA,
-					b.String(): tc.respB,
-				},
-				rtt: time.Millisecond,
-			}
-
-			ctx, cancel := context.WithCancel(context.Background())
-			defer cancel()
-
-			resolver := &upstreamResolverBase{
-				ctx:             ctx,
-				upstreamClient:  mockClient,
-				upstreamTimeout: UpstreamTimeout,
-			}
-			resolver.addRace([]netip.AddrPort{a, b})
-
-			responseWriter := &test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { return nil }}
-			resolver.ServeDNS(responseWriter, new(dns.Msg).SetQuestion("example.com.", dns.TypeA))
-
-			health := resolver.UpstreamHealth()
-			require.Contains(t, health, a, "primary upstream should have a health record")
-			if tc.wantHealthy {
-				assert.False(t, health[a].LastOk.IsZero(), "responding upstream should have LastOk set")
-				assert.True(t, health[a].LastFail.IsZero(), "responding upstream should not be marked failed")
-				assert.Empty(t, health[a].LastErr, "responding upstream should have no error")
-			} else {
-				assert.False(t, health[a].LastFail.IsZero(), "timed-out upstream should be marked failed")
-				assert.NotEmpty(t, health[a].LastErr, "timed-out upstream should record an error")
-			}
-		})
-	}
-}
-
 func TestFormatFailures(t *testing.T) {
 	testCases := []struct {
 		name     string
@@ -985,6 +913,19 @@ func TestEDEName(t *testing.T) {
 	assert.Equal(t, "EDE 9999", edeName(9999), "unknown code falls back to numeric")
 }
 
+func TestStripOPT(t *testing.T) {
+	rm := &dns.Msg{
+		Extra: []dns.RR{
+			&dns.OPT{Hdr: dns.RR_Header{Name: ".", Rrtype: dns.TypeOPT}},
+			&dns.A{Hdr: dns.RR_Header{Name: "x.", Rrtype: dns.TypeA}, A: net.IPv4(1, 2, 3, 4)},
+		},
+	}
+	stripOPT(rm)
+	assert.Len(t, rm.Extra, 1, "OPT should be removed, A kept")
+	_, isOPT := rm.Extra[0].(*dns.OPT)
+	assert.False(t, isOPT, "remaining record must not be OPT")
+}
+
 func TestUpstreamResolver_NonRetryableEDEShortCircuits(t *testing.T) {
 	upstream1 := netip.MustParseAddrPort("192.0.2.1:53")
 	upstream2 := netip.MustParseAddrPort("192.0.2.2:53")

client/internal/dnsfwd/forwarder.go

@@ -26,23 +26,8 @@ import (
 const errResolveFailed = "failed to resolve query for domain=%s: %v"
 const upstreamTimeout = 15 * time.Second
 
-// EDE info codes the forwarder emits on upstream failures so the querying
-// client can see the reason without inspecting this peer's logs. They live in
-// the RFC 8914 Private Use range (49152-65535); the Go resolver never exposes a
-// real upstream EDE here, so these cannot collide with a genuine code.
-const (
-	edeNetbirdUpstreamTimeout uint16 = 49152
-	edeNetbirdUpstreamFailure uint16 = 49153
-)
-
 type resolver interface {
 	LookupNetIP(ctx context.Context, network, host string) ([]netip.Addr, error)
-	LookupMX(ctx context.Context, name string) ([]*net.MX, error)
-	LookupTXT(ctx context.Context, name string) ([]string, error)
-	LookupNS(ctx context.Context, name string) ([]*net.NS, error)
-	LookupSRV(ctx context.Context, service, proto, name string) (string, []*net.SRV, error)
-	LookupCNAME(ctx context.Context, host string) (string, error)
-	LookupAddr(ctx context.Context, addr string) ([]string, error)
 }
 
 type firewaller interface {
@@ -216,6 +201,12 @@ func (f *DNSForwarder) handleDNSQuery(logger *log.Entry, w dns.ResponseWriter, q
 		qname, dns.TypeToString[question.Qtype], dns.ClassToString[question.Qclass])
 
 	resp := query.SetReply(query)
+	network := resutil.NetworkForQtype(question.Qtype)
+	if network == "" {
+		resp.Rcode = dns.RcodeNotImplemented
+		f.writeResponse(logger, w, resp, qname, startTime)
+		return
+	}
 
 	mostSpecificResId, matchingEntries := f.getMatchingEntries(strings.TrimSuffix(qname, "."))
 	if mostSpecificResId == "" {
@@ -227,46 +218,9 @@ func (f *DNSForwarder) handleDNSQuery(logger *log.Entry, w dns.ResponseWriter, q
 	ctx, cancel := context.WithTimeout(context.Background(), upstreamTimeout)
 	defer cancel()
 
-	reqHasEdns := query.IsEdns0() != nil
-
-	switch question.Qtype {
-	case dns.TypeA, dns.TypeAAAA:
-		f.handleAddressQuery(ctx, logger, w, resp, mostSpecificResId, matchingEntries, reqHasEdns, startTime)
-	case dns.TypeMX, dns.TypeTXT, dns.TypeNS, dns.TypeSRV, dns.TypeCNAME, dns.TypePTR:
-		f.handleRecordQuery(ctx, logger, w, resp, startTime)
-	default:
-		// The domain is routed here, so any other type is answered NODATA
-		// (NOERROR, empty answer) rather than falling back to a resolver that
-		// would poison the name with NXDOMAIN. The Extended DNS Error lets a
-		// client tell this capability-driven NODATA apart from an
-		// authoritative one. The OPT pseudo-record must not appear unless the
-		// query advertised EDNS0.
-		if reqHasEdns {
-			attachEDE(resp, dns.ExtendedErrorCodeNotSupported, "netbird forwarder: unsupported query type")
-		}
-		f.writeResponse(logger, w, resp, qname, startTime)
-	}
-}
-
-// handleAddressQuery resolves A/AAAA queries, programs the firewall sets and
-// resolved-IP state, and caches the answer for resilience on upstream failure.
-func (f *DNSForwarder) handleAddressQuery(
-	ctx context.Context,
-	logger *log.Entry,
-	w dns.ResponseWriter,
-	resp *dns.Msg,
-	mostSpecificResId route.ResID,
-	matchingEntries []*ForwarderEntry,
-	reqHasEdns bool,
-	startTime time.Time,
-) {
-	question := resp.Question[0]
-	qname := strings.ToLower(question.Name)
-
-	network := resutil.NetworkForQtype(question.Qtype)
 	result := resutil.LookupIP(ctx, f.resolver, network, qname, question.Qtype)
 	if result.Err != nil {
-		f.handleDNSError(ctx, logger, w, question, resp, qname, result, reqHasEdns, startTime)
+		f.handleDNSError(ctx, logger, w, question, resp, qname, result, startTime)
 		return
 	}
 
@@ -277,25 +231,6 @@ func (f *DNSForwarder) handleAddressQuery(
 	f.writeResponse(logger, w, resp, qname, startTime)
 }
 
-// handleRecordQuery resolves non-address record types (MX, TXT, NS, SRV,
-// CNAME, PTR) through the host resolver. Missing records are answered NODATA so
-// the routed name is never poisoned with NXDOMAIN.
-func (f *DNSForwarder) handleRecordQuery(
-	ctx context.Context,
-	logger *log.Entry,
-	w dns.ResponseWriter,
-	resp *dns.Msg,
-	startTime time.Time,
-) {
-	question := resp.Question[0]
-	qname := strings.ToLower(question.Name)
-
-	records, rcode := resutil.LookupRecords(ctx, f.resolver, qname, question.Qtype, f.ttl)
-	resp.Rcode = rcode
-	resp.Answer = append(resp.Answer, records...)
-	f.writeResponse(logger, w, resp, qname, startTime)
-}
-
 func (f *DNSForwarder) writeResponse(logger *log.Entry, w dns.ResponseWriter, resp *dns.Msg, qname string, startTime time.Time) {
 	if err := w.WriteMsg(resp); err != nil {
 		logger.Errorf("failed to write DNS response: %v", err)
@@ -398,7 +333,6 @@ func (f *DNSForwarder) handleDNSError(
 	resp *dns.Msg,
 	domain string,
 	result resutil.LookupResult,
-	reqHasEdns bool,
 	startTime time.Time,
 ) {
 	qType := question.Qtype
@@ -440,10 +374,6 @@ func (f *DNSForwarder) handleDNSError(
 		logger.Warnf(errResolveFailed, domain, result.Err)
 	}
 
-	if reqHasEdns {
-		attachEDE(resp, edeCodeFor(dnsErr), edeText(dnsErr))
-	}
-
 	f.writeResponse(logger, w, resp, domain, startTime)
 }
 
@@ -484,33 +414,3 @@ func (f *DNSForwarder) getMatchingEntries(domain string) (route.ResID, []*Forwar
 
 	return selectedResId, matches
 }
-
-// edeCodeFor maps an upstream lookup error to the NetBird EDE info code.
-func edeCodeFor(dnsErr *net.DNSError) uint16 {
-	if dnsErr != nil && dnsErr.IsTimeout {
-		return edeNetbirdUpstreamTimeout
-	}
-	return edeNetbirdUpstreamFailure
-}
-
-// edeText builds the EDE extra-text describing the class of upstream failure.
-// It deliberately omits the upstream server address, which may be an internal
-// resolver and is exposed to any client permitted to use the route; the full
-// detail stays in the forwarder's local log.
-func edeText(dnsErr *net.DNSError) string {
-	if dnsErr != nil && dnsErr.IsTimeout {
-		return "netbird forwarder: upstream timeout"
-	}
-	return "netbird forwarder: upstream failure"
-}
-
-// attachEDE adds an Extended DNS Error (RFC 8914) option to the response,
-// creating the OPT pseudo-record if the response does not already carry one.
-func attachEDE(resp *dns.Msg, code uint16, text string) {
-	opt := resp.IsEdns0()
-	if opt == nil {
-		resp.SetEdns0(dns.DefaultMsgSize, false)
-		opt = resp.IsEdns0()
-	}
-	opt.Option = append(opt.Option, &dns.EDNS0_EDE{InfoCode: code, ExtraText: text})
-}

client/internal/dnsfwd/forwarder_test.go

@@ -16,7 +16,6 @@ import (
 	"github.com/stretchr/testify/require"
 
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
-	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/client/internal/dns/test"
 	"github.com/netbirdio/netbird/client/internal/peer"
 	"github.com/netbirdio/netbird/route"
@@ -133,41 +132,6 @@ func (m *MockResolver) LookupNetIP(ctx context.Context, network, host string) ([
 	return args.Get(0).([]netip.Addr), args.Error(1)
 }
 
-func (m *MockResolver) LookupMX(ctx context.Context, name string) ([]*net.MX, error) {
-	args := m.Called(ctx, name)
-	recs, _ := args.Get(0).([]*net.MX)
-	return recs, args.Error(1)
-}
-
-func (m *MockResolver) LookupTXT(ctx context.Context, name string) ([]string, error) {
-	args := m.Called(ctx, name)
-	recs, _ := args.Get(0).([]string)
-	return recs, args.Error(1)
-}
-
-func (m *MockResolver) LookupNS(ctx context.Context, name string) ([]*net.NS, error) {
-	args := m.Called(ctx, name)
-	recs, _ := args.Get(0).([]*net.NS)
-	return recs, args.Error(1)
-}
-
-func (m *MockResolver) LookupSRV(ctx context.Context, service, proto, name string) (string, []*net.SRV, error) {
-	args := m.Called(ctx, service, proto, name)
-	recs, _ := args.Get(1).([]*net.SRV)
-	return args.String(0), recs, args.Error(2)
-}
-
-func (m *MockResolver) LookupCNAME(ctx context.Context, host string) (string, error) {
-	args := m.Called(ctx, host)
-	return args.String(0), args.Error(1)
-}
-
-func (m *MockResolver) LookupAddr(ctx context.Context, addr string) ([]string, error) {
-	args := m.Called(ctx, addr)
-	recs, _ := args.Get(0).([]string)
-	return recs, args.Error(1)
-}
-
 func TestDNSForwarder_SubdomainAccessLogic(t *testing.T) {
 	tests := []struct {
 		name             string
@@ -580,15 +544,12 @@ func TestDNSForwarder_MultipleIPsInSingleUpdate(t *testing.T) {
 }
 
 func TestDNSForwarder_ResponseCodes(t *testing.T) {
-	// A type with no net.Resolver Lookup method (CAA) must answer NODATA
-	// (NOERROR, empty) rather than NXDOMAIN/NOTIMP to avoid poisoning the name.
 	tests := []struct {
 		name         string
 		queryType    uint16
 		queryDomain  string
 		configured   string
 		expectedCode int
-		expectEDE    bool
 		description  string
 	}{
 		{
@@ -600,13 +561,28 @@ func TestDNSForwarder_ResponseCodes(t *testing.T) {
 			description:  "RFC compliant REFUSED for unauthorized queries",
 		},
 		{
-			name:         "unsupported query type returns NODATA",
-			queryType:    dns.TypeCAA,
+			name:         "unsupported query type returns NOTIMP",
+			queryType:    dns.TypeMX,
 			queryDomain:  "example.com",
 			configured:   "example.com",
-			expectedCode: dns.RcodeSuccess,
-			expectEDE:    true,
-			description:  "Unsupported types answer NODATA, not NXDOMAIN/NOTIMP",
+			expectedCode: dns.RcodeNotImplemented,
+			description:  "RFC compliant NOTIMP for unsupported types",
+		},
+		{
+			name:         "CNAME query returns NOTIMP",
+			queryType:    dns.TypeCNAME,
+			queryDomain:  "example.com",
+			configured:   "example.com",
+			expectedCode: dns.RcodeNotImplemented,
+			description:  "CNAME queries not supported",
+		},
+		{
+			name:         "TXT query returns NOTIMP",
+			queryType:    dns.TypeTXT,
+			queryDomain:  "example.com",
+			configured:   "example.com",
+			expectedCode: dns.RcodeNotImplemented,
+			description:  "TXT queries not supported",
 		},
 	}
 
@@ -622,7 +598,6 @@ func TestDNSForwarder_ResponseCodes(t *testing.T) {
 
 			query := &dns.Msg{}
 			query.SetQuestion(dns.Fqdn(tt.queryDomain), tt.queryType)
-			query.SetEdns0(dns.DefaultMsgSize, false)
 
 			// Capture the written response
 			var writtenResp *dns.Msg
@@ -638,288 +613,6 @@ func TestDNSForwarder_ResponseCodes(t *testing.T) {
 			// Check the response written to the writer
 			require.NotNil(t, writtenResp, "Expected response to be written")
 			assert.Equal(t, tt.expectedCode, writtenResp.Rcode, tt.description)
-			assert.Empty(t, writtenResp.Answer, "Non-address response should carry no answers")
-
-			if tt.expectEDE {
-				require.NotNil(t, writtenResp.IsEdns0(), "EDNS0 client should get an OPT in the reply")
-				assert.True(t, hasEDE(writtenResp, dns.ExtendedErrorCodeNotSupported),
-					"unsupported type NODATA should carry EDE Not Supported")
-			}
-		})
-	}
-}
-
-func hasEDE(m *dns.Msg, code uint16) bool {
-	opt := m.IsEdns0()
-	if opt == nil {
-		return false
-	}
-	for _, o := range opt.Option {
-		if ede, ok := o.(*dns.EDNS0_EDE); ok && ede.InfoCode == code {
-			return true
-		}
-	}
-	return false
-}
-
-func TestDNSForwarder_RecordQueries(t *testing.T) {
-	notFound := &net.DNSError{IsNotFound: true, Name: "example.com"}
-
-	t.Run("MX records are forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		mockResolver.On("LookupMX", mock.Anything, "example.com.").
-			Return([]*net.MX{{Host: "mail.example.com.", Pref: 10}}, nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeMX)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		mx, ok := resp.Answer[0].(*dns.MX)
-		require.True(t, ok, "answer should be an MX record")
-		assert.Equal(t, uint16(10), mx.Preference)
-		assert.Equal(t, "mail.example.com.", mx.Mx)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("missing MX is NODATA not NXDOMAIN", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		// A not-found cannot prove the name is absent (it may exist with only
-		// other record types), so it must answer NODATA, never NXDOMAIN.
-		mockResolver.On("LookupMX", mock.Anything, "example.com.").
-			Return(nil, notFound).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeMX)
-		assert.Equal(t, dns.RcodeSuccess, resp.Rcode, "missing record must be NODATA")
-		assert.Empty(t, resp.Answer)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("NS records are forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		mockResolver.On("LookupNS", mock.Anything, "example.com.").
-			Return([]*net.NS{{Host: "ns1.example.com."}}, nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeNS)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		ns, ok := resp.Answer[0].(*dns.NS)
-		require.True(t, ok, "answer should be an NS record")
-		assert.Equal(t, "ns1.example.com.", ns.Ns)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("missing NS is NODATA", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		mockResolver.On("LookupNS", mock.Anything, "example.com.").
-			Return(nil, notFound).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeNS)
-		assert.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		assert.Empty(t, resp.Answer)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("SRV records are forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "_sip._tcp.example.com")
-
-		mockResolver.On("LookupSRV", mock.Anything, "", "", "_sip._tcp.example.com.").
-			Return("", []*net.SRV{{Target: "sip.example.com.", Port: 5060, Priority: 10, Weight: 5}}, nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "_sip._tcp.example.com", dns.TypeSRV)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		srv, ok := resp.Answer[0].(*dns.SRV)
-		require.True(t, ok, "answer should be an SRV record")
-		assert.Equal(t, "sip.example.com.", srv.Target)
-		assert.Equal(t, uint16(5060), srv.Port)
-		assert.Equal(t, uint16(10), srv.Priority)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("missing SRV is NODATA", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "_sip._tcp.example.com")
-
-		mockResolver.On("LookupSRV", mock.Anything, "", "", "_sip._tcp.example.com.").
-			Return("", nil, notFound).Once()
-
-		resp := runRecordQuery(t, forwarder, "_sip._tcp.example.com", dns.TypeSRV)
-		assert.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		assert.Empty(t, resp.Answer)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("TXT records are forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		mockResolver.On("LookupTXT", mock.Anything, "example.com.").
-			Return([]string{"v=spf1 -all"}, nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeTXT)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		txt, ok := resp.Answer[0].(*dns.TXT)
-		require.True(t, ok, "answer should be a TXT record")
-		assert.Equal(t, []string{"v=spf1 -all"}, txt.Txt)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("CNAME record is forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "www.example.com")
-
-		mockResolver.On("LookupCNAME", mock.Anything, "www.example.com.").
-			Return("target.example.com.", nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "www.example.com", dns.TypeCNAME)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		cname, ok := resp.Answer[0].(*dns.CNAME)
-		require.True(t, ok, "answer should be a CNAME record")
-		assert.Equal(t, "target.example.com.", cname.Target)
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("CNAME equal to the name is NODATA", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "example.com")
-
-		// No CNAME exists: LookupCNAME echoes the queried name back.
-		mockResolver.On("LookupCNAME", mock.Anything, "example.com.").
-			Return("example.com.", nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "example.com", dns.TypeCNAME)
-		assert.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		assert.Empty(t, resp.Answer, "self-referential CNAME means no CNAME record")
-		mockResolver.AssertExpectations(t)
-	})
-
-	t.Run("PTR record is forwarded", func(t *testing.T) {
-		mockResolver := &MockResolver{}
-		forwarder := newRecordTestForwarder(t, mockResolver, "*.in-addr.arpa")
-
-		// The reverse name is parsed back to the address LookupAddr expects.
-		mockResolver.On("LookupAddr", mock.Anything, "1.2.3.4").
-			Return([]string{"host.example.com."}, nil).Once()
-
-		resp := runRecordQuery(t, forwarder, "4.3.2.1.in-addr.arpa", dns.TypePTR)
-		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
-		require.Len(t, resp.Answer, 1)
-		ptr, ok := resp.Answer[0].(*dns.PTR)
-		require.True(t, ok, "answer should be a PTR record")
-		assert.Equal(t, "host.example.com.", ptr.Ptr)
-		mockResolver.AssertExpectations(t)
-	})
-}
-
-func newRecordTestForwarder(t *testing.T, r resolver, configured string) *DNSForwarder {
-	t.Helper()
-	forwarder := NewDNSForwarder(netip.MustParseAddrPort("127.0.0.1:0"), 300, nil, &peer.Status{}, nil)
-	forwarder.resolver = r
-
-	d, err := domain.FromString(configured)
-	require.NoError(t, err)
-	forwarder.UpdateDomains([]*ForwarderEntry{{Domain: d, ResID: "test-res"}})
-	return forwarder
-}
-
-func runRecordQuery(t *testing.T, forwarder *DNSForwarder, qname string, qtype uint16) *dns.Msg {
-	t.Helper()
-	query := &dns.Msg{}
-	query.SetQuestion(dns.Fqdn(qname), qtype)
-
-	mockWriter := &test.MockResponseWriter{}
-	forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query, time.Now())
-
-	resp := mockWriter.GetLastResponse()
-	require.NotNil(t, resp, "expected response to be written")
-	return resp
-}
-
-func TestDNSForwarder_UpstreamFailureEDE(t *testing.T) {
-	tests := []struct {
-		name        string
-		lookupErr   error
-		reqEdns     bool
-		wantEDE     bool
-		wantCode    uint16
-		wantTextHas string
-	}{
-		{
-			name:        "timeout with edns0",
-			lookupErr:   &net.DNSError{Err: "i/o timeout", Server: "10.0.0.53:53", IsTimeout: true},
-			reqEdns:     true,
-			wantEDE:     true,
-			wantCode:    edeNetbirdUpstreamTimeout,
-			wantTextHas: "netbird forwarder: upstream timeout",
-		},
-		{
-			name:        "server failure with edns0",
-			lookupErr:   &net.DNSError{Err: "server misbehaving", Server: "10.0.0.53:53"},
-			reqEdns:     true,
-			wantEDE:     true,
-			wantCode:    edeNetbirdUpstreamFailure,
-			wantTextHas: "netbird forwarder: upstream failure",
-		},
-		{
-			name:      "no edns0 in request omits ede",
-			lookupErr: &net.DNSError{Err: "server misbehaving", Server: "10.0.0.53:53"},
-			reqEdns:   false,
-			wantEDE:   false,
-		},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			mockResolver := &MockResolver{}
-			forwarder := NewDNSForwarder(netip.MustParseAddrPort("127.0.0.1:0"), 300, nil, &peer.Status{}, nil)
-			forwarder.resolver = mockResolver
-
-			d, err := domain.FromString("example.com")
-			require.NoError(t, err)
-			forwarder.UpdateDomains([]*ForwarderEntry{{Domain: d, ResID: "test-res"}})
-
-			mockResolver.On("LookupNetIP", mock.Anything, "ip4", "example.com.").
-				Return([]netip.Addr(nil), tt.lookupErr).Once()
-
-			query := &dns.Msg{}
-			query.SetQuestion("example.com.", dns.TypeA)
-			if tt.reqEdns {
-				query.SetEdns0(dns.DefaultMsgSize, false)
-			}
-
-			var writtenResp *dns.Msg
-			mockWriter := &test.MockResponseWriter{
-				WriteMsgFunc: func(m *dns.Msg) error {
-					writtenResp = m
-					return nil
-				},
-			}
-
-			forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query, time.Now())
-			mockResolver.AssertExpectations(t)
-
-			require.NotNil(t, writtenResp, "expected a response")
-			assert.Equal(t, dns.RcodeServerFailure, writtenResp.Rcode, "upstream failure must be SERVFAIL")
-
-			ede, ok := resutil.ExtractEDE(writtenResp)
-			if !tt.wantEDE {
-				assert.False(t, ok, "response must not carry EDE")
-				return
-			}
-			require.True(t, ok, "response must carry EDE")
-			assert.Equal(t, tt.wantCode, ede.InfoCode, "EDE info code")
-			assert.Contains(t, ede.ExtraText, tt.wantTextHas, "EDE extra-text")
-			assert.NotContains(t, ede.ExtraText, "10.0.0.53", "must not leak upstream server address")
 		})
 	}
 }

client/internal/engine.go

@@ -22,6 +22,8 @@ import (
 	log "github.com/sirupsen/logrus"
 	"golang.zx2c4.com/wireguard/tun/netstack"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
+	"google.golang.org/grpc/codes"
+	gstatus "google.golang.org/grpc/status"
 
 	nberrors "github.com/netbirdio/netbird/client/errors"
 	"github.com/netbirdio/netbird/client/firewall"
@@ -82,18 +84,10 @@ const (
 	PeerConnectionTimeoutMax = 45000 // ms
 	PeerConnectionTimeoutMin = 30000 // ms
 	disableAutoUpdate        = "disabled"
-
-	// systemInfoTimeout bounds how long the sync loop waits for system info / posture
-	// check gathering. The gathering runs uncancellable system calls (process scan,
-	// exec, os.Stat); without this bound a single stuck call freezes handleSync, and
-	// thus syncMsgMux, for as long as the call hangs (observed multi-minute freezes).
-	systemInfoTimeout = 15 * time.Second
 )
 
 var ErrResetConnection = fmt.Errorf("reset connection")
 
-var ErrEngineAlreadyStarted = errors.New("engine already started")
-
 type EngineConfig struct {
 	WgPort      int
 	WgIfaceName string
@@ -172,7 +166,6 @@ type EngineServices struct {
 	StateManager   *statemanager.Manager
 	UpdateManager  *updater.Manager
 	ClientMetrics  *metrics.ClientMetrics
-	MetricsCtx     context.Context
 }
 
 // Engine is a mechanism responsible for reacting on Signal and Management stream events and managing connections to the remote peers.
@@ -208,8 +201,6 @@ type Engine struct {
 	ctx    context.Context
 	cancel context.CancelFunc
 
-	started bool
-
 	wgInterface WGIface
 
 	udpMux *udpmux.UniversalUDPMuxDefault
@@ -265,7 +256,6 @@ type Engine struct {
 
 	// clientMetrics collects and pushes metrics
 	clientMetrics *metrics.ClientMetrics
-	metricsCtx    context.Context
 
 	jobExecutor   *jobexec.Executor
 	jobExecutorWG sync.WaitGroup
@@ -291,15 +281,9 @@ func NewEngine(
 	services EngineServices,
 	mobileDep MobileDependency,
 ) *Engine {
-	// The engine is single-use: a fresh instance is built per connection
-	// cycle (see Client.run), so the run context is created once here rather
-	// than in Start.
-	ctx, cancel := context.WithCancel(clientCtx)
 	engine := &Engine{
 		clientCtx:          clientCtx,
 		clientCancel:       clientCancel,
-		ctx:                ctx,
-		cancel:             cancel,
 		signal:             services.SignalClient,
 		signaler:           peer.NewSignaler(services.SignalClient, config.WgPrivateKey),
 		mgmClient:          services.MgmClient,
@@ -318,7 +302,6 @@ func NewEngine(
 		probeStunTurn:      relay.NewStunTurnProbe(relay.DefaultCacheTTL),
 		jobExecutor:        jobexec.NewExecutor(),
 		clientMetrics:      services.ClientMetrics,
-		metricsCtx:         services.MetricsCtx,
 		updateManager:      services.UpdateManager,
 		syncStoreDir:       config.StateDir,
 	}
@@ -333,34 +316,8 @@ func (e *Engine) Stop() error {
 		log.Debugf("tried stopping engine that is nil")
 		return nil
 	}
-	e.cancel()
 	e.syncMsgMux.Lock()
 
-	e.stopLocked()
-
-	e.syncMsgMux.Unlock()
-
-	timeout := e.calculateShutdownTimeout()
-	log.Debugf("waiting for goroutines to finish with timeout: %v", timeout)
-	shutdownCtx, cancel := context.WithTimeout(context.Background(), timeout)
-	defer cancel()
-
-	if err := waitWithContext(shutdownCtx, &e.shutdownWg); err != nil {
-		log.Warnf("shutdown timeout exceeded after %v, some goroutines may still be running", timeout)
-	}
-
-	log.Infof("stopped Netbird Engine")
-
-	return nil
-}
-
-// stopLocked tears down everything Start may have brought up, in the order
-// teardown requires (DNS before the interface goes down, flow manager after).
-// The caller must hold syncMsgMux. It is shared by Stop and by Start's failure
-// path, so a partially-initialized engine is cleaned up the same way; every
-// step is nil-guarded. It does not wait on shutdownWg — the caller does that
-// after releasing the lock, since the goroutines also take syncMsgMux.
-func (e *Engine) stopLocked() {
 	if e.connMgr != nil {
 		e.connMgr.Close()
 	}
@@ -411,6 +368,10 @@ func (e *Engine) stopLocked() {
 	// so dbus and friends don't complain because of a missing interface
 	e.stopDNSServer()
 
+	if e.cancel != nil {
+		e.cancel()
+	}
+
 	e.jobExecutorWG.Wait() // block until job goroutines finish
 
 	e.close()
@@ -429,6 +390,21 @@ func (e *Engine) stopLocked() {
 	if err := e.stateManager.PersistState(context.Background()); err != nil {
 		log.Errorf("failed to persist state: %v", err)
 	}
+
+	e.syncMsgMux.Unlock()
+
+	timeout := e.calculateShutdownTimeout()
+	log.Debugf("waiting for goroutines to finish with timeout: %v", timeout)
+	shutdownCtx, cancel := context.WithTimeout(context.Background(), timeout)
+	defer cancel()
+
+	if err := waitWithContext(shutdownCtx, &e.shutdownWg); err != nil {
+		log.Warnf("shutdown timeout exceeded after %v, some goroutines may still be running", timeout)
+	}
+
+	log.Infof("stopped Netbird Engine")
+
+	return nil
 }
 
 // calculateShutdownTimeout returns shutdown timeout: 10s base + 100ms per peer, capped at 30s.
@@ -466,38 +442,18 @@ func waitWithContext(ctx context.Context, wg *sync.WaitGroup) error {
 // Start creates a new WireGuard tunnel interface and listens to events from Signal and Management services
 // Connections to remote peers are not established here.
 // However, they will be established once an event with a list of peers to connect to will be received from Management Service
-func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL) (err error) {
+func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL) error {
 	e.syncMsgMux.Lock()
 	defer e.syncMsgMux.Unlock()
 
-	// The engine is single-use. Reject a duplicate start and a start on an
-	// already-stopped engine (run context cancelled).
-	if e.started {
-		return ErrEngineAlreadyStarted
-	}
-
-	if ctxErr := e.ctx.Err(); ctxErr != nil {
-		return fmt.Errorf("engine already stopped: %w", ctxErr)
-	}
-
-	e.started = true
-
-	// Tear down any partially-initialized state on a failed start. Cancel the
-	// run context first so goroutines started before the failure (connMgr,
-	// srWatcher, monitors) unwind, then stopLocked mirrors Stop's teardown (we
-	// already hold syncMsgMux), cleaning up route/DNS/flow/state managers too,
-	// not just what close() covers.
-	defer func() {
-		if err != nil {
-			e.cancel()
-			e.stopLocked()
-		}
-	}()
-
-	if err = iface.ValidateMTU(e.config.MTU); err != nil {
+	if err := iface.ValidateMTU(e.config.MTU); err != nil {
 		return fmt.Errorf("invalid MTU configuration: %w", err)
 	}
 
+	if e.cancel != nil {
+		e.cancel()
+	}
+	e.ctx, e.cancel = context.WithCancel(e.clientCtx)
 	e.exposeManager = expose.NewManager(e.ctx, e.mgmClient)
 
 	wgIface, err := e.newWgIface()
@@ -531,11 +487,13 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 
 	initialRoutes, dnsConfig, dnsFeatureFlag, err := e.readInitialSettings()
 	if err != nil {
+		e.close()
 		return fmt.Errorf("read initial settings: %w", err)
 	}
 
 	dnsServer, err := e.newDnsServer(dnsConfig)
 	if err != nil {
+		e.close()
 		return fmt.Errorf("create dns server: %w", err)
 	}
 	e.dnsServer = dnsServer
@@ -570,6 +528,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 
 	if err = e.wgInterfaceCreate(); err != nil {
 		log.Errorf("failed creating tunnel interface %s: [%s]", e.config.WgIfaceName, err.Error())
+		e.close()
 		return fmt.Errorf("create wg interface: %w", err)
 	}
 
@@ -578,6 +537,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	}
 
 	if err := e.createFirewall(); err != nil {
+		e.close()
 		return err
 	}
 
@@ -589,6 +549,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	e.udpMux, err = e.wgInterface.Up()
 	if err != nil {
 		log.Errorf("failed to pull up wgInterface [%s]: %s", e.wgInterface.Name(), err.Error())
+		e.close()
 		return fmt.Errorf("up wg interface: %w", err)
 	}
 
@@ -613,7 +574,9 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 		e.acl = acl.NewDefaultManager(e.firewall)
 	}
 
-	if err := e.dnsServer.Initialize(); err != nil {
+	err = e.dnsServer.Initialize()
+	if err != nil {
+		e.close()
 		return fmt.Errorf("initialize dns server: %w", err)
 	}
 
@@ -625,9 +588,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	e.srWatcher = guard.NewSRWatcher(e.signal, e.relayManager, e.mobileDep.IFaceDiscover, iceCfg)
 	e.srWatcher.Start(peer.IsForceRelayed())
 
-	if err = e.receiveSignalEvents(); err != nil {
-		return err
-	}
+	e.receiveSignalEvents()
 	e.receiveManagementEvents()
 	e.receiveJobEvents()
 
@@ -679,6 +640,7 @@ func (e *Engine) createFirewall() error {
 
 func (e *Engine) initFirewall() error {
 	if err := e.routeManager.SetFirewall(e.firewall); err != nil {
+		e.close()
 		return fmt.Errorf("set firewall: %w", err)
 	}
 
@@ -904,16 +866,6 @@ func (e *Engine) handleAutoUpdateVersion(autoUpdateSettings *mgmProto.AutoUpdate
 	e.updateManager.SetVersion(autoUpdateSettings.Version, autoUpdateSettings.AlwaysUpdate)
 }
 
-// phase times a sync sub-phase: it returns a function that records the elapsed
-// duration when called. Starting the timer at the call site keeps inter-phase
-// glue code out of the measurement.
-func (e *Engine) phase(name string) func() {
-	start := time.Now()
-	return func() {
-		e.clientMetrics.RecordSyncPhase(e.ctx, name, time.Since(start))
-	}
-}
-
 func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 	started := time.Now()
 	defer func() {
@@ -933,10 +885,7 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 		e.handleAutoUpdateVersion(update.NetworkMap.PeerConfig.AutoUpdate)
 	}
 
-	done := e.phase("netbird_config")
-	err := e.updateNetbirdConfig(update.GetNetbirdConfig())
-	done()
-	if err != nil {
+	if err := e.updateNetbirdConfig(update.GetNetbirdConfig()); err != nil {
 		return err
 	}
 
@@ -950,16 +899,11 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 		return nil
 	}
 
-	done = e.phase("checks")
-	err = e.updateChecksIfNew(update.Checks)
-	done()
-	if err != nil {
+	if err := e.updateChecksIfNew(update.Checks); err != nil {
 		return err
 	}
 
-	done = e.phase("persist")
 	e.persistSyncResponse(update)
-	done()
 
 	// only apply new changes and ignore old ones
 	if err := e.updateNetworkMap(nm); err != nil {
@@ -1000,8 +944,6 @@ func (e *Engine) updateNetbirdConfig(wCfg *mgmProto.NetbirdConfig) error {
 		return fmt.Errorf("handle the flow configuration: %w", err)
 	}
 
-	e.handleMetricsUpdate(wCfg.GetMetrics())
-
 	if err := e.PopulateNetbirdConfig(wCfg, nil); err != nil {
 		log.Warnf("Failed to update DNS server config: %v", err)
 	}
@@ -1071,14 +1013,6 @@ func (e *Engine) handleFlowUpdate(config *mgmProto.FlowConfig) error {
 	return e.flowManager.Update(flowConfig)
 }
 
-func (e *Engine) handleMetricsUpdate(config *mgmProto.MetricsConfig) {
-	if config == nil {
-		return
-	}
-	log.Infof("received metrics configuration from management: enabled=%v", config.GetEnabled())
-	e.clientMetrics.UpdatePushFromMgm(e.metricsCtx, config.GetEnabled())
-}
-
 func toFlowLoggerConfig(config *mgmProto.FlowConfig) (*nftypes.FlowConfig, error) {
 	if config.GetInterval() == nil {
 		return nil, errors.New("flow interval is nil")
@@ -1103,22 +1037,11 @@ func (e *Engine) updateChecksIfNew(checks []*mgmProto.Checks) error {
 	}
 	e.checks = checks
 
-	info, ok := system.GetInfoWithChecksTimeout(e.ctx, systemInfoTimeout, checks, e.overlayAddresses()...)
-	if !ok {
-		// Gathering timed out; skip the meta sync this cycle rather than blocking the
-		// sync loop (and syncMsgMux) on a stuck system call. A later sync will retry.
-		return nil
+	info, err := system.GetInfoWithChecks(e.ctx, checks)
+	if err != nil {
+		log.Warnf("failed to get system info with checks: %v", err)
+		info = system.GetInfo(e.ctx)
 	}
-	e.applyInfoFlags(info)
-
-	if err := e.mgmClient.SyncMeta(info); err != nil {
-		return fmt.Errorf("could not sync meta: error %s", err)
-	}
-	return nil
-}
-
-// applyInfoFlags sets the engine's config-derived feature flags on the gathered system info.
-func (e *Engine) applyInfoFlags(info *system.Info) {
 	info.SetFlags(
 		e.config.RosenpassEnabled,
 		e.config.RosenpassPermissive,
@@ -1137,20 +1060,12 @@ func (e *Engine) applyInfoFlags(info *system.Info) {
 		e.config.EnableSSHRemotePortForwarding,
 		e.config.DisableSSHAuth,
 	)
-}
 
-// overlayAddresses returns our own WireGuard overlay address (v4 and v6) so it
-// can be excluded from the reported network addresses; the interface coming and
-// going otherwise churns the peer meta on the management server.
-func (e *Engine) overlayAddresses() []netip.Addr {
-	var ips []netip.Addr
-	if e.config.WgAddr.IP.IsValid() {
-		ips = append(ips, e.config.WgAddr.IP)
+	if err := e.mgmClient.SyncMeta(info); err != nil {
+		log.Errorf("could not sync meta: error %s", err)
+		return err
 	}
-	if e.config.WgAddr.HasIPv6() {
-		ips = append(ips, e.config.WgAddr.IPv6)
-	}
-	return ips
+	return nil
 }
 
 func (e *Engine) updateConfig(conf *mgmProto.PeerConfig) error {
@@ -1212,6 +1127,20 @@ func (e *Engine) hasIPv6Changed(conf *mgmProto.PeerConfig) bool {
 	return !current.HasIPv6() || current.IPv6 != prefix.Addr() || current.IPv6Net != prefix.Masked()
 }
 
+// wrapDisconnectError classifies a receive-loop failure before the run is torn
+// down. An auth rejection (PermissionDenied/Unauthenticated) means the session
+// needs re-login and retrying is futile, so mark it terminal (NeedsLogin) — run()
+// then exits on its own instead of spinning the backoff. Any other failure is a
+// recoverable connection reset that the backoff should retry.
+func (e *Engine) wrapDisconnectError(err error) {
+	state := CtxGetState(e.ctx)
+	if s, ok := gstatus.FromError(err); ok && (s.Code() == codes.PermissionDenied || s.Code() == codes.Unauthenticated) {
+		state.Set(StatusNeedsLogin)
+		return
+	}
+	_ = state.Wrap(ErrResetConnection)
+}
+
 func (e *Engine) receiveJobEvents() {
 	e.jobExecutorWG.Add(1)
 	go func() {
@@ -1238,9 +1167,9 @@ func (e *Engine) receiveJobEvents() {
 			}
 		})
 		if err != nil {
-			// happens if management is unavailable for a long time.
-			// We want to cancel the operation of the whole client
-			_ = CtxGetState(e.ctx).Wrap(ErrResetConnection)
+			// happens if management is unavailable for a long time, or rejects
+			// us (auth). wrapDisconnectError decides retry vs needs-login.
+			e.wrapDisconnectError(err)
 			e.clientCancel()
 			return
 		}
@@ -1296,19 +1225,35 @@ func (e *Engine) receiveManagementEvents() {
 	e.shutdownWg.Add(1)
 	go func() {
 		defer e.shutdownWg.Done()
-		info, ok := system.GetInfoWithChecksTimeout(e.ctx, systemInfoTimeout, e.checks, e.overlayAddresses()...)
-		if !ok {
-			// Gathering timed out; connect the stream with base info so management
-			// connectivity still comes up rather than blocking here.
+		info, err := system.GetInfoWithChecks(e.ctx, e.checks)
+		if err != nil {
+			log.Warnf("failed to get system info with checks: %v", err)
 			info = system.GetInfo(e.ctx)
 		}
-		e.applyInfoFlags(info)
+		info.SetFlags(
+			e.config.RosenpassEnabled,
+			e.config.RosenpassPermissive,
+			&e.config.ServerSSHAllowed,
+			e.config.DisableClientRoutes,
+			e.config.DisableServerRoutes,
+			e.config.DisableDNS,
+			e.config.DisableFirewall,
+			e.config.BlockLANAccess,
+			e.config.BlockInbound,
+			e.config.DisableIPv6,
+			e.config.LazyConnectionEnabled,
+			e.config.EnableSSHRoot,
+			e.config.EnableSSHSFTP,
+			e.config.EnableSSHLocalPortForwarding,
+			e.config.EnableSSHRemotePortForwarding,
+			e.config.DisableSSHAuth,
+		)
 
-		err := e.mgmClient.Sync(e.ctx, info, e.handleSync)
+		err = e.mgmClient.Sync(e.ctx, info, e.handleSync)
 		if err != nil {
-			// happens if management is unavailable for a long time.
-			// We want to cancel the operation of the whole client
-			_ = CtxGetState(e.ctx).Wrap(ErrResetConnection)
+			// happens if management is unavailable for a long time, or rejects
+			// us (auth). wrapDisconnectError decides retry vs needs-login.
+			e.wrapDisconnectError(err)
 			e.clientCancel()
 			return
 		}
@@ -1397,16 +1342,13 @@ func (e *Engine) updateNetworkMap(networkMap *mgmProto.NetworkMap) error {
 
 	dnsConfig := toDNSConfig(protoDNSConfig, e.wgInterface.Address())
 
-	done := e.phase("dns_server")
 	if err := e.dnsServer.UpdateDNSServer(serial, dnsConfig); err != nil {
 		log.Errorf("failed to update dns server, err: %v", err)
 	}
-	done()
 
 	e.routeManager.SetDNSForwarderPort(dnsConfig.ForwarderPort)
 
 	// apply routes first, route related actions might depend on routing being enabled
-	done = e.phase("routes_classify")
 	routes := toRoutes(networkMap.GetRoutes())
 	serverRoutes, clientRoutes := e.routeManager.ClassifyRoutes(routes)
 
@@ -1415,60 +1357,29 @@ func (e *Engine) updateNetworkMap(networkMap *mgmProto.NetworkMap) error {
 		e.connMgr.UpdateRouteHAMap(clientRoutes)
 		log.Debugf("updated lazy connection manager with %d HA groups", len(clientRoutes))
 	}
-	done()
 
-	done = e.phase("routes_apply")
 	dnsRouteFeatureFlag := toDNSFeatureFlag(networkMap)
 	if err := e.routeManager.UpdateRoutes(serial, serverRoutes, clientRoutes, dnsRouteFeatureFlag); err != nil {
 		log.Errorf("failed to update routes: %v", err)
 	}
-	done()
 
-	done = e.phase("filtering")
 	if e.acl != nil {
 		e.acl.ApplyFiltering(networkMap, dnsRouteFeatureFlag)
 	}
-	done()
 
-	done = e.phase("dns_forwarder")
 	fwdEntries := toRouteDomains(e.config.WgPrivateKey.PublicKey().String(), routes)
 	e.updateDNSForwarder(dnsRouteFeatureFlag, fwdEntries)
-	done()
 
 	// Ingress forward rules
-	done = e.phase("forward_rules")
 	forwardingRules, err := e.updateForwardRules(networkMap.GetForwardingRules())
 	if err != nil {
 		log.Errorf("failed to update forward rules, err: %v", err)
 	}
-	done()
 
 	log.Debugf("got peers update from Management Service, total peers to connect to = %d", len(networkMap.GetRemotePeers()))
 
-	done = e.phase("offline_peers")
 	e.updateOfflinePeers(networkMap.GetOfflinePeers())
-	done()
 
-	remotePeers, err := e.reconcilePeers(networkMap)
-	if err != nil {
-		return err
-	}
-
-	// must set the exclude list after the peers are added. Without it the manager can not figure out the peers parameters from the store
-	done = e.phase("lazy_exclude")
-	excludedLazyPeers := e.toExcludedLazyPeers(forwardingRules, remotePeers)
-	e.connMgr.SetExcludeList(e.ctx, excludedLazyPeers)
-	done()
-
-	e.networkSerial = serial
-
-	return nil
-}
-
-// reconcilePeers applies the remote peer list from the network map (removing,
-// modifying and adding peers, then updating SSH config) and returns the remote
-// peers with our own peer filtered out, for use by later sync steps.
-func (e *Engine) reconcilePeers(networkMap *mgmProto.NetworkMap) ([]*mgmProto.RemotePeerConfig, error) {
 	// Filter out own peer from the remote peers list
 	localPubKey := e.config.WgPrivateKey.PublicKey().String()
 	remotePeers := make([]*mgmProto.RemotePeerConfig, 0, len(networkMap.GetRemotePeers()))
@@ -1483,43 +1394,42 @@ func (e *Engine) reconcilePeers(networkMap *mgmProto.NetworkMap) ([]*mgmProto.Re
 		err := e.removeAllPeers()
 		e.statusRecorder.FinishPeerListModifications()
 		if err != nil {
-			return nil, err
+			return err
 		}
-		return remotePeers, nil
+	} else {
+		err := e.removePeers(remotePeers)
+		if err != nil {
+			return err
+		}
+
+		err = e.modifyPeers(remotePeers)
+		if err != nil {
+			return err
+		}
+
+		err = e.addNewPeers(remotePeers)
+		if err != nil {
+			return err
+		}
+
+		e.statusRecorder.FinishPeerListModifications()
+
+		e.updatePeerSSHHostKeys(remotePeers)
+
+		if err := e.updateSSHClientConfig(remotePeers); err != nil {
+			log.Warnf("failed to update SSH client config: %v", err)
+		}
+
+		e.updateSSHServerAuth(networkMap.GetSshAuth())
 	}
 
-	done := e.phase("removed_peers")
-	err := e.removePeers(remotePeers)
-	done()
-	if err != nil {
-		return nil, err
-	}
+	// must set the exclude list after the peers are added. Without it the manager can not figure out the peers parameters from the store
+	excludedLazyPeers := e.toExcludedLazyPeers(forwardingRules, remotePeers)
+	e.connMgr.SetExcludeList(e.ctx, excludedLazyPeers)
 
-	done = e.phase("modified_peers")
-	err = e.modifyPeers(remotePeers)
-	done()
-	if err != nil {
-		return nil, err
-	}
+	e.networkSerial = serial
 
-	done = e.phase("added_peers")
-	err = e.addNewPeers(remotePeers)
-	done()
-	if err != nil {
-		return nil, err
-	}
-
-	e.statusRecorder.FinishPeerListModifications()
-
-	e.updatePeerSSHHostKeys(remotePeers)
-
-	if err := e.updateSSHClientConfig(remotePeers); err != nil {
-		log.Warnf("failed to update SSH client config: %v", err)
-	}
-
-	e.updateSSHServerAuth(networkMap.GetSshAuth())
-
-	return remotePeers, nil
+	return nil
 }
 
 func toDNSFeatureFlag(networkMap *mgmProto.NetworkMap) bool {
@@ -1804,7 +1714,7 @@ func (e *Engine) createPeerConn(pubKey string, allowedIPs []netip.Prefix, agentV
 }
 
 // receiveSignalEvents connects to the Signal Service event stream to negotiate connection with remote peers
-func (e *Engine) receiveSignalEvents() error {
+func (e *Engine) receiveSignalEvents() {
 	e.shutdownWg.Add(1)
 	go func() {
 		defer e.shutdownWg.Done()
@@ -1867,20 +1777,15 @@ func (e *Engine) receiveSignalEvents() error {
 			return nil
 		})
 		if err != nil {
-			// happens if signal is unavailable for a long time.
-			// We want to cancel the operation of the whole client
-			_ = CtxGetState(e.ctx).Wrap(ErrResetConnection)
+			// happens if signal is unavailable for a long time, or rejects us
+			// (auth). wrapDisconnectError decides retry vs needs-login.
+			e.wrapDisconnectError(err)
 			e.clientCancel()
 			return
 		}
 	}()
 
-	// todo: consider to remove this blocker. I do not see benefit to block the Start operations
-	e.signal.WaitStreamConnected(e.ctx)
-	if err := e.ctx.Err(); err != nil {
-		return fmt.Errorf("wait for signal stream: %w", err)
-	}
-	return nil
+	e.signal.WaitStreamConnected()
 }
 
 func (e *Engine) parseNATExternalIPMappings() []string {

client/internal/engine_privileged_test.go

@@ -1,565 +0,0 @@
-//go:build privileged
-
-package internal
-
-import (
-	"context"
-	"fmt"
-	"net"
-	"runtime"
-	"strings"
-	"sync"
-	"testing"
-	"time"
-
-	"github.com/golang/mock/gomock"
-	"github.com/google/uuid"
-	log "github.com/sirupsen/logrus"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-	"go.opentelemetry.io/otel"
-	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
-	"google.golang.org/grpc"
-	"google.golang.org/grpc/keepalive"
-
-	"github.com/netbirdio/netbird/client/iface"
-	"github.com/netbirdio/netbird/client/iface/device"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
-	"github.com/netbirdio/netbird/client/internal/dns"
-	"github.com/netbirdio/netbird/client/internal/peer"
-	nbssh "github.com/netbirdio/netbird/client/ssh"
-	"github.com/netbirdio/netbird/client/system"
-	nbdns "github.com/netbirdio/netbird/dns"
-	"github.com/netbirdio/netbird/management/internals/controllers/network_map/controller"
-	"github.com/netbirdio/netbird/management/internals/controllers/network_map/update_channel"
-	"github.com/netbirdio/netbird/management/internals/modules/peers"
-	"github.com/netbirdio/netbird/management/internals/modules/peers/ephemeral/manager"
-	"github.com/netbirdio/netbird/management/internals/server/config"
-	nbgrpc "github.com/netbirdio/netbird/management/internals/shared/grpc"
-	"github.com/netbirdio/netbird/management/server"
-	"github.com/netbirdio/netbird/management/server/activity"
-	nbcache "github.com/netbirdio/netbird/management/server/cache"
-	"github.com/netbirdio/netbird/management/server/groups"
-	"github.com/netbirdio/netbird/management/server/integrations/integrated_validator/validator"
-	"github.com/netbirdio/netbird/management/server/integrations/port_forwarding"
-	"github.com/netbirdio/netbird/management/server/job"
-	"github.com/netbirdio/netbird/management/server/permissions"
-	"github.com/netbirdio/netbird/management/server/settings"
-	"github.com/netbirdio/netbird/management/server/store"
-	"github.com/netbirdio/netbird/management/server/telemetry"
-	"github.com/netbirdio/netbird/management/server/types"
-	mgmt "github.com/netbirdio/netbird/shared/management/client"
-	mgmtProto "github.com/netbirdio/netbird/shared/management/proto"
-	relayClient "github.com/netbirdio/netbird/shared/relay/client"
-	signal "github.com/netbirdio/netbird/shared/signal/client"
-	"github.com/netbirdio/netbird/shared/signal/proto"
-	signalServer "github.com/netbirdio/netbird/signal/server"
-	"github.com/netbirdio/netbird/util"
-)
-
-func TestEngine_SSH(t *testing.T) {
-	key, err := wgtypes.GeneratePrivateKey()
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-
-	sshKey, err := nbssh.GeneratePrivateKey(nbssh.ED25519)
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-
-	ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
-	defer cancel()
-
-	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
-	engine := NewEngine(
-		ctx, cancel,
-		&EngineConfig{
-			WgIfaceName:      "utun101",
-			WgAddr:           wgaddr.MustParseWGAddress("100.64.0.1/24"),
-			WgPrivateKey:     key,
-			WgPort:           33100,
-			ServerSSHAllowed: true,
-			MTU:              iface.DefaultMTU,
-			SSHKey:           sshKey,
-		},
-		EngineServices{
-			SignalClient:   &signal.MockClient{},
-			MgmClient:      &mgmt.MockClient{},
-			RelayManager:   relayMgr,
-			StatusRecorder: peer.NewRecorder("https://mgm"),
-		},
-		MobileDependency{},
-	)
-
-	engine.dnsServer = &dns.MockServer{
-		UpdateDNSServerFunc: func(serial uint64, update nbdns.Config) error { return nil },
-	}
-
-	err = engine.Start(nil, nil)
-	require.NoError(t, err)
-
-	defer func() {
-		err := engine.Stop()
-		if err != nil {
-			return
-		}
-	}()
-
-	peerWithSSH := &mgmtProto.RemotePeerConfig{
-		WgPubKey:   "MNHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
-		AllowedIps: []string{"100.64.0.21/24"},
-		SshConfig: &mgmtProto.SSHConfig{
-			SshPubKey: []byte("ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFATYCqaQw/9id1Qkq3n16JYhDhXraI6Pc1fgB8ynEfQ"),
-		},
-	}
-
-	// SSH server is not enabled so SSH config of a remote peer should be ignored
-	networkMap := &mgmtProto.NetworkMap{
-		Serial:             6,
-		PeerConfig:         nil,
-		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
-		RemotePeersIsEmpty: false,
-	}
-
-	err = engine.updateNetworkMap(networkMap)
-	require.NoError(t, err)
-
-	assert.Nil(t, engine.sshServer)
-
-	// SSH server is enabled, therefore SSH config should be applied
-	networkMap = &mgmtProto.NetworkMap{
-		Serial: 7,
-		PeerConfig: &mgmtProto.PeerConfig{Address: "100.64.0.1/24",
-			SshConfig: &mgmtProto.SSHConfig{
-				SshEnabled: true,
-				JwtConfig: &mgmtProto.JWTConfig{
-					Issuer:       "test-issuer",
-					Audience:     "test-audience",
-					KeysLocation: "test-keys",
-					MaxTokenAge:  3600,
-				},
-			}},
-		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
-		RemotePeersIsEmpty: false,
-	}
-
-	err = engine.updateNetworkMap(networkMap)
-	require.NoError(t, err)
-
-	time.Sleep(250 * time.Millisecond)
-	assert.NotNil(t, engine.sshServer)
-
-	// now remove peer
-	networkMap = &mgmtProto.NetworkMap{
-		Serial:             8,
-		RemotePeers:        []*mgmtProto.RemotePeerConfig{},
-		RemotePeersIsEmpty: false,
-	}
-
-	err = engine.updateNetworkMap(networkMap)
-	require.NoError(t, err)
-
-	// time.Sleep(250 * time.Millisecond)
-	assert.NotNil(t, engine.sshServer)
-
-	// now disable SSH server
-	networkMap = &mgmtProto.NetworkMap{
-		Serial: 9,
-		PeerConfig: &mgmtProto.PeerConfig{Address: "100.64.0.1/24",
-			SshConfig: &mgmtProto.SSHConfig{SshEnabled: false}},
-		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
-		RemotePeersIsEmpty: false,
-	}
-
-	err = engine.updateNetworkMap(networkMap)
-	require.NoError(t, err)
-
-	assert.Nil(t, engine.sshServer)
-}
-
-func TestEngine_Sync(t *testing.T) {
-	key, err := wgtypes.GeneratePrivateKey()
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-
-	ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
-	defer cancel()
-
-	// feed updates to Engine via mocked Management client
-	updates := make(chan *mgmtProto.SyncResponse)
-	defer close(updates)
-	syncFunc := func(ctx context.Context, info *system.Info, msgHandler func(msg *mgmtProto.SyncResponse) error) error {
-		for msg := range updates {
-			err := msgHandler(msg)
-			if err != nil {
-				t.Fatal(err)
-			}
-		}
-		return nil
-	}
-	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
-	engine := NewEngine(ctx, cancel, &EngineConfig{
-		WgIfaceName:  "utun103",
-		WgAddr:       wgaddr.MustParseWGAddress("100.64.0.1/24"),
-		WgPrivateKey: key,
-		WgPort:       33100,
-		MTU:          iface.DefaultMTU,
-	}, EngineServices{
-		SignalClient:   &signal.MockClient{},
-		MgmClient:      &mgmt.MockClient{SyncFunc: syncFunc},
-		RelayManager:   relayMgr,
-		StatusRecorder: peer.NewRecorder("https://mgm"),
-	}, MobileDependency{})
-	engine.ctx = ctx
-
-	engine.dnsServer = &dns.MockServer{
-		UpdateDNSServerFunc: func(serial uint64, update nbdns.Config) error { return nil },
-	}
-
-	defer func() {
-		err := engine.Stop()
-		if err != nil {
-			return
-		}
-	}()
-
-	err = engine.Start(nil, nil)
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-
-	peer1 := &mgmtProto.RemotePeerConfig{
-		WgPubKey:   "RRHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
-		AllowedIps: []string{"100.64.0.10/24"},
-	}
-	peer2 := &mgmtProto.RemotePeerConfig{
-		WgPubKey:   "LLHf3Ma6z6mdLbriAJbqhX9+nM/B71lgw2+91q3LlhU=",
-		AllowedIps: []string{"100.64.0.11/24"},
-	}
-	peer3 := &mgmtProto.RemotePeerConfig{
-		WgPubKey:   "GGHf3Ma6z6mdLbriAJbqhX9+nM/B71lgw2+91q3LlhU=",
-		AllowedIps: []string{"100.64.0.12/24"},
-	}
-	// 1st update with just 1 peer and serial larger than the current serial of the engine => apply update
-	updates <- &mgmtProto.SyncResponse{
-		NetworkMap: &mgmtProto.NetworkMap{
-			Serial:             10,
-			PeerConfig:         nil,
-			RemotePeers:        []*mgmtProto.RemotePeerConfig{peer1, peer2, peer3},
-			RemotePeersIsEmpty: false,
-		},
-	}
-
-	timeout := time.After(time.Second * 2)
-	for {
-		select {
-		case <-timeout:
-			t.Fatalf("timeout while waiting for test to finish")
-			return
-		default:
-		}
-
-		if getPeers(engine) == 3 && engine.networkSerial == 10 {
-			break
-		}
-	}
-}
-
-func TestEngine_MultiplePeers(t *testing.T) {
-	// log.SetLevel(log.DebugLevel)
-
-	ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
-	defer cancel()
-
-	sigServer, signalAddr, err := startSignal(t)
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-	defer sigServer.Stop()
-	mgmtServer, mgmtAddr, err := startManagement(t, t.TempDir(), "../testdata/store.sql")
-	if err != nil {
-		t.Fatal(err)
-		return
-	}
-	defer mgmtServer.GracefulStop()
-
-	setupKey := "A2C8E62B-38F5-4553-B31E-DD66C696CEBB"
-
-	mu := sync.Mutex{}
-	engines := []*Engine{}
-	numPeers := 10
-	wg := sync.WaitGroup{}
-	wg.Add(numPeers)
-	// create and start peers
-	for i := 0; i < numPeers; i++ {
-		j := i
-		go func() {
-			engine, err := createEngine(ctx, cancel, setupKey, j, mgmtAddr, signalAddr)
-			if err != nil {
-				wg.Done()
-				t.Errorf("unable to create the engine for peer %d with error %v", j, err)
-				return
-			}
-			engine.dnsServer = &dns.MockServer{}
-			mu.Lock()
-			defer mu.Unlock()
-			guid := fmt.Sprintf("{%s}", uuid.New().String())
-			device.CustomWindowsGUIDString = strings.ToLower(guid)
-			err = engine.Start(nil, nil)
-			if err != nil {
-				t.Errorf("unable to start engine for peer %d with error %v", j, err)
-				wg.Done()
-				return
-			}
-			engines = append(engines, engine)
-			wg.Done()
-		}()
-	}
-
-	// wait until all have been created and started
-	wg.Wait()
-	if len(engines) != numPeers {
-		t.Fatal("not all peers were started")
-	}
-	// check whether all the peer have expected peers connected
-
-	expectedConnected := numPeers * (numPeers - 1)
-
-	// adjust according to timeouts
-	timeout := 50 * time.Second
-	timeoutChan := time.After(timeout)
-	ticker := time.NewTicker(time.Second)
-	defer ticker.Stop()
-loop:
-	for {
-		select {
-		case <-timeoutChan:
-			t.Fatalf("waiting for expected connections timeout after %s", timeout.String())
-			break loop
-		case <-ticker.C:
-			totalConnected := 0
-			for _, engine := range engines {
-				totalConnected += getConnectedPeers(engine)
-			}
-			if totalConnected == expectedConnected {
-				log.Infof("total connected=%d", totalConnected)
-				break loop
-			}
-			log.Infof("total connected=%d", totalConnected)
-		}
-	}
-	// cleanup test
-	for n, peerEngine := range engines {
-		t.Logf("stopping peer with interface %s from multipeer test, loopIndex %d", peerEngine.wgInterface.Name(), n)
-		errStop := peerEngine.mgmClient.Close()
-		if errStop != nil {
-			log.Infoln("got error trying to close management clients from engine: ", errStop)
-		}
-		errStop = peerEngine.Stop()
-		if errStop != nil {
-			log.Infoln("got error trying to close testing peers engine: ", errStop)
-		}
-	}
-}
-
-var (
-	kaep = keepalive.EnforcementPolicy{
-		MinTime:             15 * time.Second,
-		PermitWithoutStream: true,
-	}
-
-	kasp = keepalive.ServerParameters{
-		MaxConnectionIdle:     15 * time.Second,
-		MaxConnectionAgeGrace: 5 * time.Second,
-		Time:                  5 * time.Second,
-		Timeout:               2 * time.Second,
-	}
-)
-
-func createEngine(ctx context.Context, cancel context.CancelFunc, setupKey string, i int, mgmtAddr string, signalAddr string) (*Engine, error) {
-	key, err := wgtypes.GeneratePrivateKey()
-	if err != nil {
-		return nil, err
-	}
-	mgmtClient, err := mgmt.NewClient(ctx, mgmtAddr, key, false)
-	if err != nil {
-		return nil, err
-	}
-	signalClient, err := signal.NewClient(ctx, signalAddr, key, false)
-	if err != nil {
-		return nil, err
-	}
-
-	info := system.GetInfo(ctx)
-	resp, err := mgmtClient.Register(setupKey, "", info, nil, nil)
-	if err != nil {
-		return nil, err
-	}
-
-	var ifaceName string
-	if runtime.GOOS == "darwin" {
-		ifaceName = fmt.Sprintf("utun1%d", i)
-	} else {
-		ifaceName = fmt.Sprintf("wt%d", i)
-	}
-
-	wgPort := 33100 + i
-	conf := &EngineConfig{
-		WgIfaceName:  ifaceName,
-		WgAddr:       wgaddr.MustParseWGAddress(resp.PeerConfig.Address),
-		WgPrivateKey: key,
-		WgPort:       wgPort,
-		MTU:          iface.DefaultMTU,
-	}
-
-	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
-	e, err := NewEngine(ctx, cancel, conf, EngineServices{
-		SignalClient:   signalClient,
-		MgmClient:      mgmtClient,
-		RelayManager:   relayMgr,
-		StatusRecorder: peer.NewRecorder("https://mgm"),
-	}, MobileDependency{}), nil
-	e.ctx = ctx
-	return e, err
-}
-
-func startSignal(t *testing.T) (*grpc.Server, string, error) {
-	t.Helper()
-
-	s := grpc.NewServer(grpc.KeepaliveEnforcementPolicy(kaep), grpc.KeepaliveParams(kasp))
-
-	lis, err := net.Listen("tcp", "localhost:0")
-	if err != nil {
-		log.Fatalf("failed to listen: %v", err)
-	}
-
-	srv, err := signalServer.NewServer(context.Background(), otel.Meter(""))
-	require.NoError(t, err)
-	proto.RegisterSignalExchangeServer(s, srv)
-
-	go func() {
-		if err = s.Serve(lis); err != nil {
-			log.Fatalf("failed to serve: %v", err)
-		}
-	}()
-
-	return s, lis.Addr().String(), nil
-}
-
-func startManagement(t *testing.T, dataDir, testFile string) (*grpc.Server, string, error) {
-	t.Helper()
-
-	config := &config.Config{
-		Stuns:      []*config.Host{},
-		TURNConfig: &config.TURNConfig{},
-		Relay: &config.Relay{
-			Addresses:      []string{"127.0.0.1:1234"},
-			CredentialsTTL: util.Duration{Duration: time.Hour},
-			Secret:         "222222222222222222",
-		},
-		Signal: &config.Host{
-			Proto: "http",
-			URI:   "localhost:10000",
-		},
-		Datadir:    dataDir,
-		HttpConfig: nil,
-	}
-
-	lis, err := net.Listen("tcp", "localhost:0")
-	if err != nil {
-		return nil, "", err
-	}
-	s := grpc.NewServer(grpc.KeepaliveEnforcementPolicy(kaep), grpc.KeepaliveParams(kasp))
-
-	store, cleanUp, err := store.NewTestStoreFromSQL(context.Background(), testFile, config.Datadir)
-	if err != nil {
-		return nil, "", err
-	}
-	t.Cleanup(cleanUp)
-
-	eventStore := &activity.InMemoryEventStore{}
-	if err != nil {
-		return nil, "", err
-	}
-
-	permissionsManager := permissions.NewManager(store)
-	peersManager := peers.NewManager(store, permissionsManager)
-	jobManager := job.NewJobManager(nil, store, peersManager)
-
-	cacheStore, err := nbcache.NewStore(context.Background(), 100*time.Millisecond, 300*time.Millisecond, 100)
-	if err != nil {
-		return nil, "", err
-	}
-
-	ia, _ := validator.NewIntegratedValidator(context.Background(), peersManager, nil, eventStore, cacheStore)
-
-	metrics, err := telemetry.NewDefaultAppMetrics(context.Background())
-	require.NoError(t, err)
-
-	ctrl := gomock.NewController(t)
-	t.Cleanup(ctrl.Finish)
-	settingsMockManager := settings.NewMockManager(ctrl)
-	settingsMockManager.EXPECT().
-		GetSettings(gomock.Any(), gomock.Any(), gomock.Any()).
-		Return(&types.Settings{}, nil).
-		AnyTimes()
-	settingsMockManager.EXPECT().
-		GetExtraSettings(gomock.Any(), gomock.Any()).
-		Return(&types.ExtraSettings{}, nil).
-		AnyTimes()
-
-	groupsManager := groups.NewManagerMock()
-
-	updateManager := update_channel.NewPeersUpdateManager(metrics)
-	requestBuffer := server.NewAccountRequestBuffer(context.Background(), store)
-	networkMapController := controller.NewController(context.Background(), store, metrics, updateManager, requestBuffer, server.MockIntegratedValidator{}, settingsMockManager, "netbird.selfhosted", port_forwarding.NewControllerMock(), manager.NewEphemeralManager(store, peersManager), config)
-	accountManager, err := server.BuildManager(context.Background(), config, store, networkMapController, jobManager, nil, "", eventStore, nil, false, ia, metrics, port_forwarding.NewControllerMock(), settingsMockManager, permissionsManager, false, cacheStore)
-	if err != nil {
-		return nil, "", err
-	}
-
-	secretsManager, err := nbgrpc.NewTimeBasedAuthSecretsManager(updateManager, config.TURNConfig, config.Relay, settingsMockManager, groupsManager)
-	if err != nil {
-		return nil, "", err
-	}
-	mgmtServer, err := nbgrpc.NewServer(config, accountManager, settingsMockManager, jobManager, secretsManager, nil, nil, &server.MockIntegratedValidator{}, networkMapController, nil, nil)
-	if err != nil {
-		return nil, "", err
-	}
-	mgmtProto.RegisterManagementServiceServer(s, mgmtServer)
-	go func() {
-		if err = s.Serve(lis); err != nil {
-			log.Fatalf("failed to serve: %v", err)
-		}
-	}()
-
-	return s, lis.Addr().String(), nil
-}
-
-// getConnectedPeers returns a connection Status or nil if peer connection wasn't found
-func getConnectedPeers(e *Engine) int {
-	e.syncMsgMux.Lock()
-	defer e.syncMsgMux.Unlock()
-	i := 0
-	for _, id := range e.peerStore.PeersPubKey() {
-		conn, _ := e.peerStore.PeerConn(id)
-		if conn.IsConnected() {
-			i++
-		}
-	}
-	return i
-}
-
-func getPeers(e *Engine) int {
-	e.syncMsgMux.Lock()
-	defer e.syncMsgMux.Unlock()
-
-	return len(e.peerStore.PeersPubKey())
-}

client/internal/engine_test.go

@@ -6,18 +6,37 @@ import (
 	"net"
 	"net/netip"
 	"os"
+	"runtime"
 	"strings"
 	"sync"
 	"testing"
 	"time"
 
+	"github.com/golang/mock/gomock"
+	"github.com/google/uuid"
+	log "github.com/sirupsen/logrus"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
+	"go.opentelemetry.io/otel"
 	wgdevice "golang.zx2c4.com/wireguard/device"
 	"golang.zx2c4.com/wireguard/tun/netstack"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
+	"google.golang.org/grpc"
+	"google.golang.org/grpc/keepalive"
 
 	"github.com/netbirdio/netbird/client/internal/stdnet"
+	"github.com/netbirdio/netbird/management/server/job"
+
+	"github.com/netbirdio/netbird/management/server/integrations/integrated_validator/validator"
+
+	"github.com/netbirdio/netbird/management/internals/controllers/network_map/controller"
+	"github.com/netbirdio/netbird/management/internals/controllers/network_map/update_channel"
+	"github.com/netbirdio/netbird/management/internals/modules/peers"
+	"github.com/netbirdio/netbird/management/internals/modules/peers/ephemeral/manager"
+	nbgrpc "github.com/netbirdio/netbird/management/internals/shared/grpc"
+
+	"github.com/netbirdio/netbird/management/internals/server/config"
+	"github.com/netbirdio/netbird/management/server/groups"
 
 	"github.com/netbirdio/netbird/client/iface"
 	"github.com/netbirdio/netbird/client/iface/configurer"
@@ -31,7 +50,18 @@ import (
 	icemaker "github.com/netbirdio/netbird/client/internal/peer/ice"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	"github.com/netbirdio/netbird/client/internal/routemanager"
+	nbssh "github.com/netbirdio/netbird/client/ssh"
+	"github.com/netbirdio/netbird/client/system"
 	nbdns "github.com/netbirdio/netbird/dns"
+	"github.com/netbirdio/netbird/management/server"
+	"github.com/netbirdio/netbird/management/server/activity"
+	nbcache "github.com/netbirdio/netbird/management/server/cache"
+	"github.com/netbirdio/netbird/management/server/integrations/port_forwarding"
+	"github.com/netbirdio/netbird/management/server/permissions"
+	"github.com/netbirdio/netbird/management/server/settings"
+	"github.com/netbirdio/netbird/management/server/store"
+	"github.com/netbirdio/netbird/management/server/telemetry"
+	"github.com/netbirdio/netbird/management/server/types"
 	"github.com/netbirdio/netbird/monotime"
 	"github.com/netbirdio/netbird/route"
 	mgmt "github.com/netbirdio/netbird/shared/management/client"
@@ -39,9 +69,25 @@ import (
 	"github.com/netbirdio/netbird/shared/netiputil"
 	relayClient "github.com/netbirdio/netbird/shared/relay/client"
 	signal "github.com/netbirdio/netbird/shared/signal/client"
+	"github.com/netbirdio/netbird/shared/signal/proto"
+	signalServer "github.com/netbirdio/netbird/signal/server"
 	"github.com/netbirdio/netbird/util"
 )
 
+var (
+	kaep = keepalive.EnforcementPolicy{
+		MinTime:             15 * time.Second,
+		PermitWithoutStream: true,
+	}
+
+	kasp = keepalive.ServerParameters{
+		MaxConnectionIdle:     15 * time.Second,
+		MaxConnectionAgeGrace: 5 * time.Second,
+		Time:                  5 * time.Second,
+		Timeout:               2 * time.Second,
+	}
+)
+
 type MockWGIface struct {
 	CreateFunc                 func() error
 	CreateOnAndroidFunc        func(routeRange []string, ip string, domains []string) error
@@ -178,10 +224,6 @@ func (m *MockWGIface) LastActivities() map[string]monotime.Time {
 	return nil
 }
 
-func (m *MockWGIface) MTU() uint16 {
-	return 1280
-}
-
 func (m *MockWGIface) SetPresharedKey(peerKey string, psk wgtypes.Key, updateOnly bool) error {
 	return nil
 }
@@ -192,6 +234,129 @@ func TestMain(m *testing.M) {
 	os.Exit(code)
 }
 
+func TestEngine_SSH(t *testing.T) {
+	key, err := wgtypes.GeneratePrivateKey()
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+
+	sshKey, err := nbssh.GeneratePrivateKey(nbssh.ED25519)
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
+	engine := NewEngine(
+		ctx, cancel,
+		&EngineConfig{
+			WgIfaceName:      "utun101",
+			WgAddr:           wgaddr.MustParseWGAddress("100.64.0.1/24"),
+			WgPrivateKey:     key,
+			WgPort:           33100,
+			ServerSSHAllowed: true,
+			MTU:              iface.DefaultMTU,
+			SSHKey:           sshKey,
+		},
+		EngineServices{
+			SignalClient:   &signal.MockClient{},
+			MgmClient:      &mgmt.MockClient{},
+			RelayManager:   relayMgr,
+			StatusRecorder: peer.NewRecorder("https://mgm"),
+		},
+		MobileDependency{},
+	)
+
+	engine.dnsServer = &dns.MockServer{
+		UpdateDNSServerFunc: func(serial uint64, update nbdns.Config) error { return nil },
+	}
+
+	err = engine.Start(nil, nil)
+	require.NoError(t, err)
+
+	defer func() {
+		err := engine.Stop()
+		if err != nil {
+			return
+		}
+	}()
+
+	peerWithSSH := &mgmtProto.RemotePeerConfig{
+		WgPubKey:   "MNHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
+		AllowedIps: []string{"100.64.0.21/24"},
+		SshConfig: &mgmtProto.SSHConfig{
+			SshPubKey: []byte("ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFATYCqaQw/9id1Qkq3n16JYhDhXraI6Pc1fgB8ynEfQ"),
+		},
+	}
+
+	// SSH server is not enabled so SSH config of a remote peer should be ignored
+	networkMap := &mgmtProto.NetworkMap{
+		Serial:             6,
+		PeerConfig:         nil,
+		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
+		RemotePeersIsEmpty: false,
+	}
+
+	err = engine.updateNetworkMap(networkMap)
+	require.NoError(t, err)
+
+	assert.Nil(t, engine.sshServer)
+
+	// SSH server is enabled, therefore SSH config should be applied
+	networkMap = &mgmtProto.NetworkMap{
+		Serial: 7,
+		PeerConfig: &mgmtProto.PeerConfig{Address: "100.64.0.1/24",
+			SshConfig: &mgmtProto.SSHConfig{
+				SshEnabled: true,
+				JwtConfig: &mgmtProto.JWTConfig{
+					Issuer:       "test-issuer",
+					Audience:     "test-audience",
+					KeysLocation: "test-keys",
+					MaxTokenAge:  3600,
+				},
+			}},
+		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
+		RemotePeersIsEmpty: false,
+	}
+
+	err = engine.updateNetworkMap(networkMap)
+	require.NoError(t, err)
+
+	time.Sleep(250 * time.Millisecond)
+	assert.NotNil(t, engine.sshServer)
+
+	// now remove peer
+	networkMap = &mgmtProto.NetworkMap{
+		Serial:             8,
+		RemotePeers:        []*mgmtProto.RemotePeerConfig{},
+		RemotePeersIsEmpty: false,
+	}
+
+	err = engine.updateNetworkMap(networkMap)
+	require.NoError(t, err)
+
+	// time.Sleep(250 * time.Millisecond)
+	assert.NotNil(t, engine.sshServer)
+
+	// now disable SSH server
+	networkMap = &mgmtProto.NetworkMap{
+		Serial: 9,
+		PeerConfig: &mgmtProto.PeerConfig{Address: "100.64.0.1/24",
+			SshConfig: &mgmtProto.SSHConfig{SshEnabled: false}},
+		RemotePeers:        []*mgmtProto.RemotePeerConfig{peerWithSSH},
+		RemotePeersIsEmpty: false,
+	}
+
+	err = engine.updateNetworkMap(networkMap)
+	require.NoError(t, err)
+
+	assert.Nil(t, engine.sshServer)
+}
+
 func TestEngine_SSHUpdateLogic(t *testing.T) {
 	// Test that SSH server start/stop logic works based on config
 	engine := &Engine{
@@ -261,7 +426,7 @@ func TestEngine_UpdateNetworkMap(t *testing.T) {
 		return
 	}
 
-	ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
+	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
 	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
@@ -466,6 +631,97 @@ func TestEngine_UpdateNetworkMap(t *testing.T) {
 	}
 }
 
+func TestEngine_Sync(t *testing.T) {
+	key, err := wgtypes.GeneratePrivateKey()
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+
+	// feed updates to Engine via mocked Management client
+	updates := make(chan *mgmtProto.SyncResponse)
+	defer close(updates)
+	syncFunc := func(ctx context.Context, info *system.Info, msgHandler func(msg *mgmtProto.SyncResponse) error) error {
+		for msg := range updates {
+			err := msgHandler(msg)
+			if err != nil {
+				t.Fatal(err)
+			}
+		}
+		return nil
+	}
+	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
+	engine := NewEngine(ctx, cancel, &EngineConfig{
+		WgIfaceName:  "utun103",
+		WgAddr:       wgaddr.MustParseWGAddress("100.64.0.1/24"),
+		WgPrivateKey: key,
+		WgPort:       33100,
+		MTU:          iface.DefaultMTU,
+	}, EngineServices{
+		SignalClient:   &signal.MockClient{},
+		MgmClient:      &mgmt.MockClient{SyncFunc: syncFunc},
+		RelayManager:   relayMgr,
+		StatusRecorder: peer.NewRecorder("https://mgm"),
+	}, MobileDependency{})
+	engine.ctx = ctx
+
+	engine.dnsServer = &dns.MockServer{
+		UpdateDNSServerFunc: func(serial uint64, update nbdns.Config) error { return nil },
+	}
+
+	defer func() {
+		err := engine.Stop()
+		if err != nil {
+			return
+		}
+	}()
+
+	err = engine.Start(nil, nil)
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+
+	peer1 := &mgmtProto.RemotePeerConfig{
+		WgPubKey:   "RRHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
+		AllowedIps: []string{"100.64.0.10/24"},
+	}
+	peer2 := &mgmtProto.RemotePeerConfig{
+		WgPubKey:   "LLHf3Ma6z6mdLbriAJbqhX9+nM/B71lgw2+91q3LlhU=",
+		AllowedIps: []string{"100.64.0.11/24"},
+	}
+	peer3 := &mgmtProto.RemotePeerConfig{
+		WgPubKey:   "GGHf3Ma6z6mdLbriAJbqhX9+nM/B71lgw2+91q3LlhU=",
+		AllowedIps: []string{"100.64.0.12/24"},
+	}
+	// 1st update with just 1 peer and serial larger than the current serial of the engine => apply update
+	updates <- &mgmtProto.SyncResponse{
+		NetworkMap: &mgmtProto.NetworkMap{
+			Serial:             10,
+			PeerConfig:         nil,
+			RemotePeers:        []*mgmtProto.RemotePeerConfig{peer1, peer2, peer3},
+			RemotePeersIsEmpty: false,
+		},
+	}
+
+	timeout := time.After(time.Second * 2)
+	for {
+		select {
+		case <-timeout:
+			t.Fatalf("timeout while waiting for test to finish")
+			return
+		default:
+		}
+
+		if getPeers(engine) == 3 && engine.networkSerial == 10 {
+			break
+		}
+	}
+}
+
 func TestEngine_UpdateNetworkMapWithRoutes(t *testing.T) {
 	testCases := []struct {
 		name                 string
@@ -561,7 +817,7 @@ func TestEngine_UpdateNetworkMapWithRoutes(t *testing.T) {
 				return
 			}
 
-			ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
+			ctx, cancel := context.WithCancel(context.Background())
 			defer cancel()
 
 			wgIfaceName := fmt.Sprintf("utun%d", 104+n)
@@ -768,7 +1024,7 @@ func TestEngine_UpdateNetworkMapWithDNSUpdate(t *testing.T) {
 				return
 			}
 
-			ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
+			ctx, cancel := context.WithCancel(context.Background())
 			defer cancel()
 
 			wgIfaceName := fmt.Sprintf("utun%d", 104+n)
@@ -849,6 +1105,104 @@ func TestEngine_UpdateNetworkMapWithDNSUpdate(t *testing.T) {
 	}
 }
 
+func TestEngine_MultiplePeers(t *testing.T) {
+	// log.SetLevel(log.DebugLevel)
+
+	ctx, cancel := context.WithCancel(CtxInitState(context.Background()))
+	defer cancel()
+
+	sigServer, signalAddr, err := startSignal(t)
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+	defer sigServer.Stop()
+	mgmtServer, mgmtAddr, err := startManagement(t, t.TempDir(), "../testdata/store.sql")
+	if err != nil {
+		t.Fatal(err)
+		return
+	}
+	defer mgmtServer.GracefulStop()
+
+	setupKey := "A2C8E62B-38F5-4553-B31E-DD66C696CEBB"
+
+	mu := sync.Mutex{}
+	engines := []*Engine{}
+	numPeers := 10
+	wg := sync.WaitGroup{}
+	wg.Add(numPeers)
+	// create and start peers
+	for i := 0; i < numPeers; i++ {
+		j := i
+		go func() {
+			engine, err := createEngine(ctx, cancel, setupKey, j, mgmtAddr, signalAddr)
+			if err != nil {
+				wg.Done()
+				t.Errorf("unable to create the engine for peer %d with error %v", j, err)
+				return
+			}
+			engine.dnsServer = &dns.MockServer{}
+			mu.Lock()
+			defer mu.Unlock()
+			guid := fmt.Sprintf("{%s}", uuid.New().String())
+			device.CustomWindowsGUIDString = strings.ToLower(guid)
+			err = engine.Start(nil, nil)
+			if err != nil {
+				t.Errorf("unable to start engine for peer %d with error %v", j, err)
+				wg.Done()
+				return
+			}
+			engines = append(engines, engine)
+			wg.Done()
+		}()
+	}
+
+	// wait until all have been created and started
+	wg.Wait()
+	if len(engines) != numPeers {
+		t.Fatal("not all peers was started")
+	}
+	// check whether all the peer have expected peers connected
+
+	expectedConnected := numPeers * (numPeers - 1)
+
+	// adjust according to timeouts
+	timeout := 50 * time.Second
+	timeoutChan := time.After(timeout)
+	ticker := time.NewTicker(time.Second)
+	defer ticker.Stop()
+loop:
+	for {
+		select {
+		case <-timeoutChan:
+			t.Fatalf("waiting for expected connections timeout after %s", timeout.String())
+			break loop
+		case <-ticker.C:
+			totalConnected := 0
+			for _, engine := range engines {
+				totalConnected += getConnectedPeers(engine)
+			}
+			if totalConnected == expectedConnected {
+				log.Infof("total connected=%d", totalConnected)
+				break loop
+			}
+			log.Infof("total connected=%d", totalConnected)
+		}
+	}
+	// cleanup test
+	for n, peerEngine := range engines {
+		t.Logf("stopping peer with interface %s from multipeer test, loopIndex %d", peerEngine.wgInterface.Name(), n)
+		errStop := peerEngine.mgmClient.Close()
+		if errStop != nil {
+			log.Infoln("got error trying to close management clients from engine: ", errStop)
+		}
+		errStop = peerEngine.Stop()
+		if errStop != nil {
+			log.Infoln("got error trying to close testing peers engine: ", errStop)
+		}
+	}
+}
+
 func Test_ParseNATExternalIPMappings(t *testing.T) {
 	ifaceList, err := net.Interfaces()
 	if err != nil {
@@ -1172,6 +1526,187 @@ func TestCompareNetIPLists(t *testing.T) {
 	}
 }
 
+func createEngine(ctx context.Context, cancel context.CancelFunc, setupKey string, i int, mgmtAddr string, signalAddr string) (*Engine, error) {
+	key, err := wgtypes.GeneratePrivateKey()
+	if err != nil {
+		return nil, err
+	}
+	mgmtClient, err := mgmt.NewClient(ctx, mgmtAddr, key, false)
+	if err != nil {
+		return nil, err
+	}
+	signalClient, err := signal.NewClient(ctx, signalAddr, key, false)
+	if err != nil {
+		return nil, err
+	}
+
+	info := system.GetInfo(ctx)
+	resp, err := mgmtClient.Register(setupKey, "", info, nil, nil)
+	if err != nil {
+		return nil, err
+	}
+
+	var ifaceName string
+	if runtime.GOOS == "darwin" {
+		ifaceName = fmt.Sprintf("utun1%d", i)
+	} else {
+		ifaceName = fmt.Sprintf("wt%d", i)
+	}
+
+	wgPort := 33100 + i
+	conf := &EngineConfig{
+		WgIfaceName:  ifaceName,
+		WgAddr:       wgaddr.MustParseWGAddress(resp.PeerConfig.Address),
+		WgPrivateKey: key,
+		WgPort:       wgPort,
+		MTU:          iface.DefaultMTU,
+	}
+
+	relayMgr := relayClient.NewManager(ctx, nil, key.PublicKey().String(), iface.DefaultMTU)
+	e, err := NewEngine(ctx, cancel, conf, EngineServices{
+		SignalClient:   signalClient,
+		MgmClient:      mgmtClient,
+		RelayManager:   relayMgr,
+		StatusRecorder: peer.NewRecorder("https://mgm"),
+	}, MobileDependency{}), nil
+	e.ctx = ctx
+	return e, err
+}
+
+func startSignal(t *testing.T) (*grpc.Server, string, error) {
+	t.Helper()
+
+	s := grpc.NewServer(grpc.KeepaliveEnforcementPolicy(kaep), grpc.KeepaliveParams(kasp))
+
+	lis, err := net.Listen("tcp", "localhost:0")
+	if err != nil {
+		log.Fatalf("failed to listen: %v", err)
+	}
+
+	srv, err := signalServer.NewServer(context.Background(), otel.Meter(""))
+	require.NoError(t, err)
+	proto.RegisterSignalExchangeServer(s, srv)
+
+	go func() {
+		if err = s.Serve(lis); err != nil {
+			log.Fatalf("failed to serve: %v", err)
+		}
+	}()
+
+	return s, lis.Addr().String(), nil
+}
+
+func startManagement(t *testing.T, dataDir, testFile string) (*grpc.Server, string, error) {
+	t.Helper()
+
+	config := &config.Config{
+		Stuns:      []*config.Host{},
+		TURNConfig: &config.TURNConfig{},
+		Relay: &config.Relay{
+			Addresses:      []string{"127.0.0.1:1234"},
+			CredentialsTTL: util.Duration{Duration: time.Hour},
+			Secret:         "222222222222222222",
+		},
+		Signal: &config.Host{
+			Proto: "http",
+			URI:   "localhost:10000",
+		},
+		Datadir:    dataDir,
+		HttpConfig: nil,
+	}
+
+	lis, err := net.Listen("tcp", "localhost:0")
+	if err != nil {
+		return nil, "", err
+	}
+	s := grpc.NewServer(grpc.KeepaliveEnforcementPolicy(kaep), grpc.KeepaliveParams(kasp))
+
+	store, cleanUp, err := store.NewTestStoreFromSQL(context.Background(), testFile, config.Datadir)
+	if err != nil {
+		return nil, "", err
+	}
+	t.Cleanup(cleanUp)
+
+	eventStore := &activity.InMemoryEventStore{}
+	if err != nil {
+		return nil, "", err
+	}
+
+	permissionsManager := permissions.NewManager(store)
+	peersManager := peers.NewManager(store, permissionsManager)
+	jobManager := job.NewJobManager(nil, store, peersManager)
+
+	cacheStore, err := nbcache.NewStore(context.Background(), 100*time.Millisecond, 300*time.Millisecond, 100)
+	if err != nil {
+		return nil, "", err
+	}
+
+	ia, _ := validator.NewIntegratedValidator(context.Background(), peersManager, nil, eventStore, cacheStore)
+
+	metrics, err := telemetry.NewDefaultAppMetrics(context.Background())
+	require.NoError(t, err)
+
+	ctrl := gomock.NewController(t)
+	t.Cleanup(ctrl.Finish)
+	settingsMockManager := settings.NewMockManager(ctrl)
+	settingsMockManager.EXPECT().
+		GetSettings(gomock.Any(), gomock.Any(), gomock.Any()).
+		Return(&types.Settings{}, nil).
+		AnyTimes()
+	settingsMockManager.EXPECT().
+		GetExtraSettings(gomock.Any(), gomock.Any()).
+		Return(&types.ExtraSettings{}, nil).
+		AnyTimes()
+
+	groupsManager := groups.NewManagerMock()
+
+	updateManager := update_channel.NewPeersUpdateManager(metrics)
+	requestBuffer := server.NewAccountRequestBuffer(context.Background(), store)
+	networkMapController := controller.NewController(context.Background(), store, metrics, updateManager, requestBuffer, server.MockIntegratedValidator{}, settingsMockManager, "netbird.selfhosted", port_forwarding.NewControllerMock(), manager.NewEphemeralManager(store, peersManager), config)
+	accountManager, err := server.BuildManager(context.Background(), config, store, networkMapController, jobManager, nil, "", eventStore, nil, false, ia, metrics, port_forwarding.NewControllerMock(), settingsMockManager, permissionsManager, false, cacheStore)
+	if err != nil {
+		return nil, "", err
+	}
+
+	secretsManager, err := nbgrpc.NewTimeBasedAuthSecretsManager(updateManager, config.TURNConfig, config.Relay, settingsMockManager, groupsManager)
+	if err != nil {
+		return nil, "", err
+	}
+	mgmtServer, err := nbgrpc.NewServer(config, accountManager, settingsMockManager, jobManager, secretsManager, nil, nil, &server.MockIntegratedValidator{}, networkMapController, nil, nil)
+	if err != nil {
+		return nil, "", err
+	}
+	mgmtProto.RegisterManagementServiceServer(s, mgmtServer)
+	go func() {
+		if err = s.Serve(lis); err != nil {
+			log.Fatalf("failed to serve: %v", err)
+		}
+	}()
+
+	return s, lis.Addr().String(), nil
+}
+
+// getConnectedPeers returns a connection Status or nil if peer connection wasn't found
+func getConnectedPeers(e *Engine) int {
+	e.syncMsgMux.Lock()
+	defer e.syncMsgMux.Unlock()
+	i := 0
+	for _, id := range e.peerStore.PeersPubKey() {
+		conn, _ := e.peerStore.PeerConn(id)
+		if conn.IsConnected() {
+			i++
+		}
+	}
+	return i
+}
+
+func getPeers(e *Engine) int {
+	e.syncMsgMux.Lock()
+	defer e.syncMsgMux.Unlock()
+
+	return len(e.peerStore.PeersPubKey())
+}
+
 func mustEncodePrefix(t *testing.T, p netip.Prefix) []byte {
 	t.Helper()
 	b, err := netiputil.EncodePrefix(p)

client/internal/iface_common.go

@@ -44,5 +44,4 @@ type wgIfaceBase interface {
 	FullStats() (*configurer.Stats, error)
 	LastActivities() map[string]monotime.Time
 	SetPresharedKey(peerKey string, psk wgtypes.Key, updateOnly bool) error
-	MTU() uint16
 }

client/internal/lazyconn/activity/listener_bind.go

@@ -119,16 +119,15 @@ func (d *BindListener) ReadPackets() {
 	}
 
 	d.peerCfg.Log.Debugf("removing lazy endpoint for peer %s", d.peerCfg.PublicKey)
+	if err := d.wgIface.RemovePeer(d.peerCfg.PublicKey); err != nil {
+		d.peerCfg.Log.Errorf("failed to remove endpoint: %s", err)
+	}
+
 	_ = d.lazyConn.Close()
 	d.bind.RemoveEndpoint(d.fakeIP)
 	d.done.Done()
 }
 
-// CapturedPacket is unused in userspace bind mode: first-packet reinjection is kernel-only.
-func (d *BindListener) CapturedPacket() []byte {
-	return nil
-}
-
 // Close stops the listener and cleans up resources.
 func (d *BindListener) Close() {
 	d.peerCfg.Log.Infof("closing activity listener (LazyConn)")

client/internal/lazyconn/activity/listener_bind_test.go

@@ -45,6 +45,10 @@ type MockWGIfaceBind struct {
 	endpointMgr *mockEndpointManager
 }
 
+func (m *MockWGIfaceBind) RemovePeer(string) error {
+	return nil
+}
+
 func (m *MockWGIfaceBind) UpdatePeer(string, []netip.Prefix, time.Duration, *net.UDPAddr, *wgtypes.Key) error {
 	return nil
 }
@@ -64,10 +68,6 @@ func (m *MockWGIfaceBind) GetBind() device.EndpointManager {
 	return m.endpointMgr
 }
 
-func (m *MockWGIfaceBind) MTU() uint16 {
-	return 1280
-}
-
 func TestBindListener_Creation(t *testing.T) {
 	mockEndpointMgr := newMockEndpointManager()
 	mockIface := &MockWGIfaceBind{endpointMgr: mockEndpointMgr}
@@ -207,9 +207,8 @@ func TestManager_BindMode(t *testing.T) {
 	require.NoError(t, err)
 
 	select {
-	case ev := <-mgr.OnActivityChan:
-		assert.Equal(t, cfg.PeerConnID, ev.PeerConnID, "Received peer connection ID should match")
-		assert.Nil(t, ev.FirstPacket, "Bind mode does not capture packets: reinjection is kernel-only")
+	case peerConnID := <-mgr.OnActivityChan:
+		assert.Equal(t, cfg.PeerConnID, peerConnID, "Received peer connection ID should match")
 	case <-time.After(2 * time.Second):
 		t.Fatal("timeout waiting for activity notification")
 	}
@@ -267,8 +266,8 @@ func TestManager_BindMode_MultiplePeers(t *testing.T) {
 	receivedPeers := make(map[peerid.ConnID]bool)
 	for i := 0; i < 2; i++ {
 		select {
-		case ev := <-mgr.OnActivityChan:
-			receivedPeers[ev.PeerConnID] = true
+		case peerConnID := <-mgr.OnActivityChan:
+			receivedPeers[peerConnID] = true
 		case <-time.After(2 * time.Second):
 			t.Fatal("timeout waiting for activity notifications")
 		}

client/internal/lazyconn/activity/listener_udp.go

@@ -3,13 +3,11 @@ package activity
 import (
 	"fmt"
 	"net"
-	"slices"
 	"sync"
 	"sync/atomic"
 
 	log "github.com/sirupsen/logrus"
 
-	"github.com/netbirdio/netbird/client/iface/bufsize"
 	"github.com/netbirdio/netbird/client/internal/lazyconn"
 )
 
@@ -22,8 +20,6 @@ type UDPListener struct {
 	done     sync.Mutex
 
 	isClosed atomic.Bool
-
-	capturedPacket []byte
 }
 
 // NewUDPListener creates a listener that detects activity via UDP socket reads.
@@ -50,13 +46,9 @@ func NewUDPListener(wgIface WgInterface, cfg lazyconn.PeerConfig) (*UDPListener,
 }
 
 // ReadPackets blocks reading from the UDP socket until activity is detected or the listener is closed.
-// The first packet that triggers activity is captured so it can be reinjected through the real
-// transport once it is established. Without this, kernel WireGuard's handshake initiation would be
-// dropped and WG would only retry after REKEY_TIMEOUT.
 func (d *UDPListener) ReadPackets() {
 	for {
-		buf := make([]byte, int(d.wgIface.MTU())+bufsize.WGBufferOverhead)
-		n, remoteAddr, err := d.conn.ReadFromUDP(buf)
+		n, remoteAddr, err := d.conn.ReadFromUDP(make([]byte, 1))
 		if err != nil {
 			if d.isClosed.Load() {
 				d.peerCfg.Log.Infof("exit from activity listener")
@@ -70,24 +62,20 @@ func (d *UDPListener) ReadPackets() {
 			d.peerCfg.Log.Warnf("received %d bytes from %s, too short", n, remoteAddr)
 			continue
 		}
-		d.capturedPacket = slices.Clone(buf[:n])
-		d.peerCfg.Log.Infof("activity detected, captured %d bytes for reinjection", n)
+		d.peerCfg.Log.Infof("activity detected")
 		break
 	}
 
-	// Leave the peer in place. ConfigureWGEndpoint will UpdatePeer with the real endpoint;
-	// removing the peer here wipes kernel WG's staged queue and drops the user packet that
-	// triggered activation.
+	d.peerCfg.Log.Debugf("removing lazy endpoint: %s", d.endpoint.String())
+	if err := d.wgIface.RemovePeer(d.peerCfg.PublicKey); err != nil {
+		d.peerCfg.Log.Errorf("failed to remove endpoint: %s", err)
+	}
+
+	// Ignore close error as it may return "use of closed network connection" if already closed.
 	_ = d.conn.Close()
 	d.done.Unlock()
 }
 
-// CapturedPacket returns the first packet that triggered activity, or nil if none was captured.
-// Safe to call after ReadPackets returns.
-func (d *UDPListener) CapturedPacket() []byte {
-	return d.capturedPacket
-}
-
 // Close stops the listener and cleans up resources.
 func (d *UDPListener) Close() {
 	d.peerCfg.Log.Infof("closing activity listener: %s", d.conn.LocalAddr().String())

client/internal/lazyconn/activity/manager.go

@@ -19,25 +19,17 @@ import (
 type listener interface {
 	ReadPackets()
 	Close()
-	CapturedPacket() []byte
-}
-
-// Event reports activity on a managed peer. FirstPacket is the bytes that triggered activation,
-// captured for reinjection through the real transport.
-type Event struct {
-	PeerConnID  peerid.ConnID
-	FirstPacket []byte
 }
 
 type WgInterface interface {
+	RemovePeer(peerKey string) error
 	UpdatePeer(peerKey string, allowedIps []netip.Prefix, keepAlive time.Duration, endpoint *net.UDPAddr, preSharedKey *wgtypes.Key) error
 	IsUserspaceBind() bool
 	Address() wgaddr.Address
-	MTU() uint16
 }
 
 type Manager struct {
-	OnActivityChan chan Event
+	OnActivityChan chan peerid.ConnID
 
 	wgIface WgInterface
 
@@ -49,7 +41,7 @@ type Manager struct {
 
 func NewManager(wgIface WgInterface) *Manager {
 	m := &Manager{
-		OnActivityChan: make(chan Event, 1),
+		OnActivityChan: make(chan peerid.ConnID, 1),
 		wgIface:        wgIface,
 		peers:          make(map[peerid.ConnID]listener),
 		done:           make(chan struct{}),
@@ -124,12 +116,12 @@ func (m *Manager) waitForTraffic(l listener, peerConnID peerid.ConnID) {
 	delete(m.peers, peerConnID)
 	m.mu.Unlock()
 
-	m.notify(Event{PeerConnID: peerConnID, FirstPacket: l.CapturedPacket()})
+	m.notify(peerConnID)
 }
 
-func (m *Manager) notify(ev Event) {
+func (m *Manager) notify(peerConnID peerid.ConnID) {
 	select {
 	case <-m.done:
-	case m.OnActivityChan <- ev:
+	case m.OnActivityChan <- peerConnID:
 	}
 }

client/internal/lazyconn/activity/manager_test.go

@@ -1,7 +1,6 @@
 package activity
 
 import (
-	"bytes"
 	"net"
 	"net/netip"
 	"testing"
@@ -26,6 +25,10 @@ func (m *MocPeer) ConnID() peerid.ConnID {
 type MocWGIface struct {
 }
 
+func (m MocWGIface) RemovePeer(string) error {
+	return nil
+}
+
 func (m MocWGIface) UpdatePeer(string, []netip.Prefix, time.Duration, *net.UDPAddr, *wgtypes.Key) error {
 	return nil
 }
@@ -41,10 +44,6 @@ func (m MocWGIface) Address() wgaddr.Address {
 	}
 }
 
-func (m MocWGIface) MTU() uint16 {
-	return 1280
-}
-
 // GetPeerListener is a test helper to access listeners
 func (m *Manager) GetPeerListener(peerConnID peerid.ConnID) (listener, bool) {
 	m.mu.Lock()
@@ -87,15 +86,11 @@ func TestManager_MonitorPeerActivity(t *testing.T) {
 	}
 
 	select {
-	case ev := <-mgr.OnActivityChan:
-		if ev.PeerConnID != peerCfg1.PeerConnID {
-			t.Fatalf("unexpected peerConnID: %v", ev.PeerConnID)
-		}
-		if !bytes.Equal(ev.FirstPacket, []byte{0x01, 0x02, 0x03, 0x04, 0x05}) {
-			t.Fatalf("unexpected first packet: %v", ev.FirstPacket)
+	case peerConnID := <-mgr.OnActivityChan:
+		if peerConnID != peerCfg1.PeerConnID {
+			t.Fatalf("unexpected peerConnID: %v", peerConnID)
 		}
 	case <-time.After(1 * time.Second):
-		t.Fatal("timed out waiting for activity")
 	}
 }
 

client/internal/lazyconn/manager/manager.go

@@ -130,8 +130,8 @@ func (m *Manager) Start(ctx context.Context) {
 		select {
 		case <-ctx.Done():
 			return
-		case ev := <-m.activityManager.OnActivityChan:
-			m.onPeerActivity(ev)
+		case peerConnID := <-m.activityManager.OnActivityChan:
+			m.onPeerActivity(peerConnID)
 		case peerIDs := <-m.inactivityManager.InactivePeersChan():
 			m.onPeerInactivityTimedOut(peerIDs)
 		}
@@ -513,13 +513,13 @@ func (m *Manager) checkHaGroupActivity(haGroup route.HAUniqueID, peerID string,
 	return false
 }
 
-func (m *Manager) onPeerActivity(ev activity.Event) {
+func (m *Manager) onPeerActivity(peerConnID peerid.ConnID) {
 	m.managedPeersMu.Lock()
 	defer m.managedPeersMu.Unlock()
 
-	mp, ok := m.managedPeersByConnID[ev.PeerConnID]
+	mp, ok := m.managedPeersByConnID[peerConnID]
 	if !ok {
-		log.Errorf("peer not found by conn id: %v", ev.PeerConnID)
+		log.Errorf("peer not found by conn id: %v", peerConnID)
 		return
 	}
 
@@ -536,7 +536,7 @@ func (m *Manager) onPeerActivity(ev activity.Event) {
 
 	m.activateHAGroupPeers(mp.peerCfg)
 
-	m.peerStore.PeerConnOpenWithFirstPacket(m.engineCtx, mp.peerCfg.PublicKey, ev.FirstPacket)
+	m.peerStore.PeerConnOpen(m.engineCtx, mp.peerCfg.PublicKey)
 }
 
 func (m *Manager) onPeerInactivityTimedOut(peerIDs map[string]struct{}) {

client/internal/lazyconn/wgiface.go

@@ -17,5 +17,4 @@ type WGIface interface {
 	IsUserspaceBind() bool
 	Address() wgaddr.Address
 	LastActivities() map[string]monotime.Time
-	MTU() uint16
 }

client/internal/metrics/env.go

@@ -60,13 +60,6 @@ func getMetricsInterval() time.Duration {
 	return interval
 }
 
-// isMetricsPushEnvSet returns true if NB_METRICS_PUSH_ENABLED is explicitly set (to any value).
-// When set, the env var takes full precedence over management server configuration.
-func isMetricsPushEnvSet() bool {
-	_, set := os.LookupEnv(EnvMetricsPushEnabled)
-	return set
-}
-
 func isForceSending() bool {
 	force, _ := strconv.ParseBool(os.Getenv(EnvMetricsForceSending))
 	return force

client/internal/metrics/influxdb.go

@@ -120,30 +120,6 @@ func (m *influxDBMetrics) RecordSyncDuration(_ context.Context, agentInfo AgentI
 	m.trimLocked()
 }
 
-func (m *influxDBMetrics) RecordSyncPhase(_ context.Context, agentInfo AgentInfo, phase string, duration time.Duration) {
-	tags := fmt.Sprintf("deployment_type=%s,version=%s,os=%s,arch=%s,peer_id=%s,phase=%s",
-		agentInfo.DeploymentType.String(),
-		agentInfo.Version,
-		agentInfo.OS,
-		agentInfo.Arch,
-		agentInfo.peerID,
-		phase,
-	)
-
-	m.mu.Lock()
-	defer m.mu.Unlock()
-
-	m.samples = append(m.samples, influxSample{
-		measurement: "netbird_sync_phase",
-		tags:        tags,
-		fields: map[string]float64{
-			"duration_seconds": duration.Seconds(),
-		},
-		timestamp: time.Now(),
-	})
-	m.trimLocked()
-}
-
 func (m *influxDBMetrics) RecordLoginDuration(_ context.Context, agentInfo AgentInfo, duration time.Duration, success bool) {
 	result := "success"
 	if !success {

client/internal/metrics/infra/README.md

@@ -78,25 +78,6 @@ Tags:
 - `os`: Operating system (linux, darwin, windows, android, ios, etc.)
 - `arch`: CPU architecture (amd64, arm64, etc.)
 
-### Sync Phase Timing
-
-Measurement: `netbird_sync_phase`
-
-Breaks down where time goes inside a single sync, so the total `netbird_sync` duration can be attributed to the sub-step that dominates.
-
-| Field | Description |
-|-------|-------------|
-| `duration_seconds` | Time spent in one sub-phase of sync processing |
-
-Tags:
-- `phase`: the sub-phase — `netbird_config`, `checks`, `persist`, `dns_server`, `routes_classify`, `routes_apply`, `filtering`, `dns_forwarder`, `forward_rules`, `offline_peers`, `removed_peers`, `modified_peers`, `added_peers`, `lazy_exclude`
-- `deployment_type`: "cloud" | "selfhosted" | "unknown"
-- `version`: NetBird version string
-- `os`: Operating system (linux, darwin, windows, android, ios, etc.)
-- `arch`: CPU architecture (amd64, arm64, etc.)
-
-**Note:** this is wall-time per phase — it includes both CPU work and time spent waiting on locks. A slow phase points to *where* the time goes, not *why*; pair it with lock-wait metrics to tell contention apart from real work.
-
 ### Login Duration
 
 Measurement: `netbird_login`
@@ -210,52 +191,4 @@ docker compose exec influxdb influx query \
 
 # Check ingest server health
 curl http://localhost:8087/health
-```
-
-## Analyzing a Debug Bundle
-
-Metrics collection is always on, so every debug bundle ships a `metrics.txt` in InfluxDB line protocol — a timestamped time series of all recorded events (sync durations, sync phases, connection stages, login). You can replay it into the local stack and graph it, without a running client.
-
-The bundle's `metrics.txt` is a rolling window (capped at 5 days / ~20k samples, see [Buffer Limits](#buffer-limits)). For a connection incident the relevant window is short (connection setup is seconds), so a bundle captured during the issue is enough.
-
-### 1. Start the stack
-
-```bash
-# From this directory (client/internal/metrics/infra)
-INFLUXDB_ADMIN_TOKEN=admin123 INFLUXDB_ADMIN_PASSWORD=admin123 GRAFANA_ADMIN_PASSWORD=admin123 \
-  docker compose up -d
-```
-
-(`admin123` are throwaway local credentials — fine for offline analysis.)
-
-### 2. Clear any previous data
-
-So you only see this bundle:
-
-```bash
-docker exec influxdb influx delete --org netbird --bucket metrics --token admin123 \
-  --start 1970-01-01T00:00:00Z --stop 2100-01-01T00:00:00Z
-```
-
-### 3. Import the bundle's metrics.txt
-
-InfluxDB is not exposed on the host, so import inside the container:
-
-```bash
-docker cp /path/to/bundle/metrics.txt influxdb:/tmp/m.txt
-docker exec influxdb influx write --org netbird --bucket metrics --precision ns \
-  --token admin123 --file /tmp/m.txt
-```
-
-Re-importing the same file is idempotent (same measurement+tags+timestamp overwrites).
-
-### 4. View the dashboards
-
-Grafana on http://localhost:3001 (login `admin` / `admin123`), datasource pre-provisioned:
-
-- **Where sync time goes:** http://localhost:3001/d/netbird-sync-phases/netbird-sync-phases-where-time-goes
-- **General client metrics:** http://localhost:3001/d/netbird-influxdb-metrics
-
-**Set the time range** to cover the bundle's timestamps (e.g. "Last 7 days" or an absolute range matching when the bundle was taken) — with the default short range the panels look empty.
-
-Bundles are distinguishable by the `version` tag; add a tag at import time (e.g. `sed 's/^netbird_\([a-z_]*\),/netbird_\1,bundle=mycase,/' metrics.txt`) if you want to compare several side by side.
+```

client/internal/metrics/infra/grafana/provisioning/dashboards/json/netbird-sync-phases.json

@@ -1,259 +0,0 @@
-{
-  "annotations": {
-    "list": []
-  },
-  "editable": true,
-  "fiscalYearStartMonth": 0,
-  "graphTooltip": 1,
-  "links": [],
-  "refresh": "",
-  "schemaVersion": 39,
-  "tags": [
-    "netbird",
-    "sync"
-  ],
-  "templating": {
-    "list": [
-      {
-        "current": {
-          "text": "All",
-          "value": "$__all"
-        },
-        "datasource": {
-          "type": "influxdb",
-          "uid": "influxdb"
-        },
-        "definition": "import \"influxdata/influxdb/schema\"\nschema.tagValues(bucket: \"metrics\", tag: \"version\")",
-        "includeAll": true,
-        "label": "version",
-        "multi": true,
-        "name": "version",
-        "query": "import \"influxdata/influxdb/schema\"\nschema.tagValues(bucket: \"metrics\", tag: \"version\")",
-        "refresh": 2,
-        "type": "query",
-        "allValue": ".*"
-      }
-    ]
-  },
-  "time": {
-    "from": "now-2d",
-    "to": "now"
-  },
-  "timepicker": {},
-  "timezone": "",
-  "title": "NetBird Sync Phases (where time goes)",
-  "uid": "netbird-sync-phases",
-  "version": 1,
-  "panels": [
-    {
-      "id": 1,
-      "title": "Time per phase over time (stacked, ms)",
-      "type": "timeseries",
-      "datasource": {
-        "type": "influxdb",
-        "uid": "influxdb"
-      },
-      "gridPos": {
-        "h": 10,
-        "w": 24,
-        "x": 0,
-        "y": 0
-      },
-      "fieldConfig": {
-        "defaults": {
-          "unit": "ms",
-          "custom": {
-            "drawStyle": "bars",
-            "stacking": {
-              "mode": "normal",
-              "group": "A"
-            },
-            "fillOpacity": 80,
-            "lineWidth": 0
-          }
-        },
-        "overrides": []
-      },
-      "options": {
-        "legend": {
-          "displayMode": "table",
-          "placement": "right",
-          "calcs": [
-            "max",
-            "mean"
-          ]
-        },
-        "tooltip": {
-          "mode": "multi",
-          "sort": "desc"
-        }
-      },
-      "targets": [
-        {
-          "refId": "A",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync_phase\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> keep(columns: [\"_time\", \"_value\", \"phase\"])\n  |> group(columns: [\"phase\"])"
-        }
-      ]
-    },
-    {
-      "id": 2,
-      "title": "p95 per phase (ms)",
-      "type": "bargauge",
-      "datasource": {
-        "type": "influxdb",
-        "uid": "influxdb"
-      },
-      "gridPos": {
-        "h": 11,
-        "w": 12,
-        "x": 0,
-        "y": 10
-      },
-      "fieldConfig": {
-        "defaults": {
-          "unit": "ms",
-          "color": {
-            "mode": "continuous-GrYlRd"
-          }
-        },
-        "overrides": []
-      },
-      "options": {
-        "displayMode": "gradient",
-        "orientation": "horizontal",
-        "reduceOptions": {
-          "calcs": [
-            "lastNotNull"
-          ],
-          "fields": "",
-          "values": false
-        },
-        "showUnfilled": true
-      },
-      "targets": [
-        {
-          "refId": "A",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync_phase\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> group(columns: [\"phase\"])\n  |> quantile(q: 0.95)\n  |> group()\n  |> sort(columns: [\"_value\"], desc: true)"
-        }
-      ]
-    },
-    {
-      "id": 3,
-      "title": "Per-phase stats (ms): mean / p95 / max",
-      "type": "table",
-      "datasource": {
-        "type": "influxdb",
-        "uid": "influxdb"
-      },
-      "gridPos": {
-        "h": 11,
-        "w": 12,
-        "x": 12,
-        "y": 10
-      },
-      "fieldConfig": {
-        "defaults": {
-          "unit": "ms"
-        },
-        "overrides": []
-      },
-      "options": {
-        "showHeader": true,
-        "sortBy": [
-          {
-            "displayName": "max",
-            "desc": true
-          }
-        ]
-      },
-      "transformations": [
-        {
-          "id": "merge",
-          "options": {}
-        }
-      ],
-      "targets": [
-        {
-          "refId": "mean",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync_phase\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> group(columns: [\"phase\"])\n  |> mean()\n  |> group()\n  |> keep(columns: [\"phase\", \"_value\"])\n  |> rename(columns: {_value: \"mean\"})"
-        },
-        {
-          "refId": "p95",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync_phase\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> group(columns: [\"phase\"])\n  |> quantile(q: 0.95)\n  |> group()\n  |> keep(columns: [\"phase\", \"_value\"])\n  |> rename(columns: {_value: \"p95\"})"
-        },
-        {
-          "refId": "max",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync_phase\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> group(columns: [\"phase\"])\n  |> max()\n  |> group()\n  |> keep(columns: [\"phase\", \"_value\"])\n  |> rename(columns: {_value: \"max\"})"
-        }
-      ]
-    },
-    {
-      "id": 4,
-      "title": "Total sync duration (netbird_sync, ms) \u2014 reference",
-      "type": "timeseries",
-      "datasource": {
-        "type": "influxdb",
-        "uid": "influxdb"
-      },
-      "gridPos": {
-        "h": 8,
-        "w": 24,
-        "x": 0,
-        "y": 21
-      },
-      "fieldConfig": {
-        "defaults": {
-          "unit": "ms",
-          "custom": {
-            "drawStyle": "points",
-            "pointSize": 5
-          }
-        },
-        "overrides": []
-      },
-      "options": {
-        "legend": {
-          "displayMode": "table",
-          "placement": "right",
-          "calcs": [
-            "max",
-            "mean"
-          ]
-        },
-        "tooltip": {
-          "mode": "single"
-        }
-      },
-      "targets": [
-        {
-          "refId": "A",
-          "datasource": {
-            "type": "influxdb",
-            "uid": "influxdb"
-          },
-          "query": "from(bucket: \"metrics\")\n  |> range(start: v.timeRangeStart, stop: v.timeRangeStop)\n  |> filter(fn: (r) => r._measurement == \"netbird_sync\" and r._field == \"duration_seconds\")\n  |> filter(fn: (r) => r.version =~ /${version:regex}/)\n  |> map(fn: (r) => ({ r with _value: r._value * 1000.0 }))\n  |> keep(columns: [\"_time\", \"_value\", \"version\"])\n  |> group(columns: [\"version\"])"
-        }
-      ]
-    }
-  ]
-}

client/internal/metrics/infra/ingest/main.go

@@ -19,7 +19,7 @@ const (
 	defaultListenAddr  = ":8087"
 	defaultInfluxDBURL = "http://influxdb:8086/api/v2/write?org=netbird&bucket=metrics&precision=ns"
 	maxBodySize        = 50 * 1024 * 1024 // 50 MB max request body
-	maxDurationSeconds = 86400.0          // reject any duration field > 24 hours
+	maxDurationSeconds = 300.0            // reject any duration field > 5 minutes
 	peerIDLength       = 16               // truncated SHA-256: 8 bytes = 16 hex chars
 	maxTagValueLength  = 64               // reject tag values longer than this
 )
@@ -59,19 +59,6 @@ var allowedMeasurements = map[string]measurementSpec{
 			"peer_id":         true,
 		},
 	},
-	"netbird_sync_phase": {
-		allowedFields: map[string]bool{
-			"duration_seconds": true,
-		},
-		allowedTags: map[string]bool{
-			"deployment_type": true,
-			"version":         true,
-			"os":              true,
-			"arch":            true,
-			"peer_id":         true,
-			"phase":           true,
-		},
-	},
 	"netbird_login": {
 		allowedFields: map[string]bool{
 			"duration_seconds": true,

client/internal/metrics/infra/ingest/main_test.go

@@ -53,14 +53,14 @@ func TestValidateLine_NegativeValue(t *testing.T) {
 }
 
 func TestValidateLine_DurationTooLarge(t *testing.T) {
-	line := `netbird_sync,deployment_type=cloud,version=1.0.0,os=linux,arch=amd64,peer_id=abc duration_seconds=100000 1234567890`
+	line := `netbird_sync,deployment_type=cloud,version=1.0.0,os=linux,arch=amd64,peer_id=abc duration_seconds=999 1234567890`
 	err := validateLine(line)
 	require.Error(t, err)
 	assert.Contains(t, err.Error(), "too large")
 }
 
 func TestValidateLine_TotalSecondsTooLarge(t *testing.T) {
-	line := `netbird_peer_connection,deployment_type=cloud,connection_type=ice,attempt_type=initial,version=1.0.0,os=linux,arch=amd64,peer_id=abc,connection_pair_id=pair total_seconds=100000 1234567890`
+	line := `netbird_peer_connection,deployment_type=cloud,connection_type=ice,attempt_type=initial,version=1.0.0,os=linux,arch=amd64,peer_id=abc,connection_pair_id=pair total_seconds=500 1234567890`
 	err := validateLine(line)
 	require.Error(t, err)
 	assert.Contains(t, err.Error(), "too large")

client/internal/metrics/metrics.go

@@ -7,7 +7,6 @@ import (
 	"fmt"
 	"io"
 	"sync"
-	"sync/atomic"
 	"time"
 
 	log "github.com/sirupsen/logrus"
@@ -57,9 +56,6 @@ type metricsImplementation interface {
 	// RecordSyncDuration records how long it took to process a sync message
 	RecordSyncDuration(ctx context.Context, agentInfo AgentInfo, duration time.Duration)
 
-	// RecordSyncPhase records how long a single sub-phase of sync processing took
-	RecordSyncPhase(ctx context.Context, agentInfo AgentInfo, phase string, duration time.Duration)
-
 	// RecordLoginDuration records how long the login to management took
 	RecordLoginDuration(ctx context.Context, agentInfo AgentInfo, duration time.Duration, success bool)
 
@@ -76,7 +72,7 @@ type ClientMetrics struct {
 	agentInfo AgentInfo
 	mu        sync.RWMutex
 
-	push       atomic.Pointer[Push]
+	push       *Push
 	pushMu     sync.Mutex
 	wg         sync.WaitGroup
 	pushCancel context.CancelFunc
@@ -131,18 +127,6 @@ func (c *ClientMetrics) RecordSyncDuration(ctx context.Context, duration time.Du
 	c.impl.RecordSyncDuration(ctx, agentInfo, duration)
 }
 
-// RecordSyncPhase records the duration of a single sub-phase of sync processing
-func (c *ClientMetrics) RecordSyncPhase(ctx context.Context, phase string, duration time.Duration) {
-	if c == nil {
-		return
-	}
-	c.mu.RLock()
-	agentInfo := c.agentInfo
-	c.mu.RUnlock()
-
-	c.impl.RecordSyncPhase(ctx, agentInfo, phase, duration)
-}
-
 // RecordLoginDuration records how long the login to management server took
 func (c *ClientMetrics) RecordLoginDuration(ctx context.Context, duration time.Duration, success bool) {
 	if c == nil {
@@ -168,7 +152,10 @@ func (c *ClientMetrics) UpdateAgentInfo(agentInfo AgentInfo, publicKey string) {
 	c.agentInfo = agentInfo
 	c.mu.Unlock()
 
-	if push := c.push.Load(); push != nil {
+	c.pushMu.Lock()
+	push := c.push
+	c.pushMu.Unlock()
+	if push != nil {
 		push.SetPeerID(agentInfo.peerID)
 	}
 }
@@ -182,7 +169,7 @@ func (c *ClientMetrics) Export(w io.Writer) error {
 	return c.impl.Export(w)
 }
 
-// StartPush starts periodic pushing of metrics with the given configuration.
+// StartPush starts periodic pushing of metrics with the given configuration
 // Precedence: PushConfig.ServerAddress > remote config server_url
 func (c *ClientMetrics) StartPush(ctx context.Context, config PushConfig) {
 	if c == nil {
@@ -192,58 +179,11 @@ func (c *ClientMetrics) StartPush(ctx context.Context, config PushConfig) {
 	c.pushMu.Lock()
 	defer c.pushMu.Unlock()
 
-	if c.push.Load() != nil {
+	if c.push != nil {
 		log.Warnf("metrics push already running")
 		return
 	}
 
-	c.startPushLocked(ctx, config)
-}
-
-// StopPush stops the periodic metrics push.
-func (c *ClientMetrics) StopPush() {
-	if c == nil {
-		return
-	}
-	c.pushMu.Lock()
-	defer c.pushMu.Unlock()
-
-	c.stopPushLocked()
-}
-
-// UpdatePushFromMgm updates metrics push based on management server configuration.
-// If NB_METRICS_PUSH_ENABLED is explicitly set (true or false), management config is ignored.
-// When unset, management controls whether push is enabled.
-func (c *ClientMetrics) UpdatePushFromMgm(ctx context.Context, enabled bool) {
-	if c == nil {
-		return
-	}
-
-	if isMetricsPushEnvSet() {
-		log.Debugf("ignoring management config, env var is explicitly set: %s", EnvMetricsPushEnabled)
-		return
-	}
-
-	c.pushMu.Lock()
-	defer c.pushMu.Unlock()
-
-	if enabled {
-		if c.push.Load() != nil {
-			return
-		}
-		log.Infof("enabled metrics push by management")
-		c.startPushLocked(ctx, PushConfigFromEnv())
-	} else {
-		if c.push.Load() == nil {
-			return
-		}
-		log.Infof("disabled metrics push by management")
-		c.stopPushLocked()
-	}
-}
-
-// startPushLocked starts push. Caller must hold pushMu.
-func (c *ClientMetrics) startPushLocked(ctx context.Context, config PushConfig) {
 	c.mu.RLock()
 	agentVersion := c.agentInfo.Version
 	peerID := c.agentInfo.peerID
@@ -259,23 +199,26 @@ func (c *ClientMetrics) startPushLocked(ctx context.Context, config PushConfig)
 
 	ctx, cancel := context.WithCancel(ctx)
 	c.pushCancel = cancel
-	c.push.Store(push)
 
 	c.wg.Add(1)
 	go func() {
 		defer c.wg.Done()
 		push.Start(ctx)
-		c.push.CompareAndSwap(push, nil)
 	}()
+	c.push = push
 }
 
-// stopPushLocked stops push. Caller must hold pushMu.
-func (c *ClientMetrics) stopPushLocked() {
-	if c.push.Load() == nil {
+func (c *ClientMetrics) StopPush() {
+	if c == nil {
+		return
+	}
+	c.pushMu.Lock()
+	defer c.pushMu.Unlock()
+	if c.push == nil {
 		return
 	}
 
 	c.pushCancel()
 	c.wg.Wait()
-	c.push.Store(nil)
+	c.push = nil
 }

client/internal/metrics/push_test.go

@@ -70,9 +70,6 @@ func (m *mockMetrics) RecordConnectionStages(_ context.Context, _ AgentInfo, _ s
 func (m *mockMetrics) RecordSyncDuration(_ context.Context, _ AgentInfo, _ time.Duration) {
 }
 
-func (m *mockMetrics) RecordSyncPhase(_ context.Context, _ AgentInfo, _ string, _ time.Duration) {
-}
-
 func (m *mockMetrics) RecordLoginDuration(_ context.Context, _ AgentInfo, _ time.Duration, _ bool) {
 }
 

client/internal/peer/conn.go

@@ -6,7 +6,6 @@ import (
 	"net"
 	"net/netip"
 	"runtime"
-	"slices"
 	"sync"
 	"time"
 
@@ -137,39 +136,6 @@ type Conn struct {
 	// Connection stage timestamps for metrics
 	metricsRecorder MetricsRecorder
 	metricsStages   *MetricsStages
-
-	// pendingFirstPacket is the lazyconn-captured handshake init, replayed once the real
-	// transport is up.
-	pendingFirstPacket []byte
-}
-
-// injectPendingFirstPacket replays the captured handshake through the proxy if present, else
-// directly through the ICE conn. The packet is cleared only after a successful write, so a failed
-// or transport-less attempt leaves it available for a later reinjection. Caller must hold conn.mu.
-func (conn *Conn) injectPendingFirstPacket(proxy wgproxy.Proxy, directConn net.Conn) {
-	pkt := conn.pendingFirstPacket
-	if len(pkt) == 0 {
-		return
-	}
-
-	switch {
-	case proxy != nil:
-		if err := proxy.InjectPacket(pkt); err != nil {
-			conn.Log.Debugf("failed to reinject captured first packet via proxy: %v", err)
-			return
-		}
-	case directConn != nil:
-		if _, err := directConn.Write(pkt); err != nil {
-			conn.Log.Debugf("failed to reinject captured first packet via direct conn: %v", err)
-			return
-		}
-	default:
-		conn.Log.Debugf("no transport available to reinject captured first packet")
-		return
-	}
-
-	conn.pendingFirstPacket = nil
-	conn.Log.Debugf("reinjected captured first packet (%d bytes)", len(pkt))
 }
 
 // NewConn creates a new not opened Conn to the remote peer.
@@ -206,16 +172,6 @@ func NewConn(config ConnConfig, services ServiceDependencies) (*Conn, error) {
 // It will try to establish a connection using ICE and in parallel with relay. The higher priority connection type will
 // be used.
 func (conn *Conn) Open(engineCtx context.Context) error {
-	return conn.open(engineCtx, nil)
-}
-
-// OpenWithFirstPacket opens the connection like Open and stashes firstPacket to be replayed once
-// the real transport is established. The packet is retained only on a successful open.
-func (conn *Conn) OpenWithFirstPacket(engineCtx context.Context, firstPacket []byte) error {
-	return conn.open(engineCtx, firstPacket)
-}
-
-func (conn *Conn) open(engineCtx context.Context, firstPacket []byte) error {
 	conn.mu.Lock()
 	defer conn.mu.Unlock()
 
@@ -271,9 +227,6 @@ func (conn *Conn) open(engineCtx context.Context, firstPacket []byte) error {
 		defer conn.wg.Done()
 		conn.guard.Start(conn.ctx, conn.onGuardEvent)
 	}()
-	if len(firstPacket) > 0 {
-		conn.pendingFirstPacket = slices.Clone(firstPacket)
-	}
 	conn.opened = true
 	return nil
 }
@@ -470,8 +423,6 @@ func (conn *Conn) onICEConnectionIsReady(priority conntype.ConnPriority, iceConn
 		conn.wgProxyRelay.RedirectAs(ep)
 	}
 
-	conn.injectPendingFirstPacket(wgProxy, iceConnInfo.RemoteConn)
-
 	conn.currentConnPriority = priority
 	conn.statusICE.SetConnected()
 	conn.updateIceState(iceConnInfo, updateTime)
@@ -595,8 +546,6 @@ func (conn *Conn) onRelayConnectionIsReady(rci RelayConnInfo) {
 
 	wgConfigWorkaround()
 
-	conn.injectPendingFirstPacket(wgProxy, nil)
-
 	conn.rosenpassRemoteKey = rci.rosenpassPubKey
 	conn.currentConnPriority = conntype.Relay
 	conn.statusRelay.SetConnected()
@@ -803,17 +752,15 @@ func (conn *Conn) isConnectedOnAllWay() (status guard.ConnStatus) {
 }
 
 func (conn *Conn) enableWgWatcherIfNeeded(enabledTime time.Time) {
-	if !conn.wgWatcher.PrepareInitialHandshake() {
-		return
+	if !conn.wgWatcher.IsEnabled() {
+		wgWatcherCtx, wgWatcherCancel := context.WithCancel(conn.ctx)
+		conn.wgWatcherCancel = wgWatcherCancel
+		conn.wgWatcherWg.Add(1)
+		go func() {
+			defer conn.wgWatcherWg.Done()
+			conn.wgWatcher.EnableWgWatcher(wgWatcherCtx, enabledTime, conn.onWGDisconnected, conn.onWGHandshakeSuccess)
+		}()
 	}
-
-	wgWatcherCtx, wgWatcherCancel := context.WithCancel(conn.ctx)
-	conn.wgWatcherCancel = wgWatcherCancel
-	conn.wgWatcherWg.Add(1)
-	go func() {
-		defer conn.wgWatcherWg.Done()
-		conn.wgWatcher.EnableWgWatcher(wgWatcherCtx, enabledTime, conn.onWGDisconnected, conn.onWGHandshakeSuccess)
-	}()
 }
 
 func (conn *Conn) disableWgWatcherIfNeeded() {

client/internal/peer/handshaker.go

@@ -195,14 +195,14 @@ func (h *Handshaker) sendOffer() error {
 	}
 
 	offer := h.buildOfferAnswer()
-	h.log.Debugf("sending offer with serial: %s", offer.SessionIDString())
+	h.log.Infof("sending offer with serial: %s", offer.SessionIDString())
 
 	return h.signaler.SignalOffer(offer, h.config.Key)
 }
 
 func (h *Handshaker) sendAnswer() error {
 	answer := h.buildOfferAnswer()
-	h.log.Debugf("sending answer with serial: %s", answer.SessionIDString())
+	h.log.Infof("sending answer with serial: %s", answer.SessionIDString())
 
 	return h.signaler.SignalAnswer(answer, h.config.Key)
 }

client/internal/peer/status.go

@@ -192,7 +192,6 @@ func (s *StatusChangeSubscription) Events() chan map[string]RouterState {
 // Pure read methods take RLock; anything that mutates state takes Lock.
 type Status struct {
 	mux                   sync.RWMutex
-	muxRelays             sync.RWMutex
 	peers                 map[string]State
 	ipToKey               map[string]string
 	changeNotify          map[string]map[string]*StatusChangeSubscription // map[peerID]map[subscriptionID]*StatusChangeSubscription
@@ -245,8 +244,8 @@ func NewRecorder(mgmAddress string) *Status {
 }
 
 func (d *Status) SetRelayMgr(manager *relayClient.Manager) {
-	d.muxRelays.Lock()
-	defer d.muxRelays.Unlock()
+	d.mux.Lock()
+	defer d.mux.Unlock()
 	d.relayMgr = manager
 }
 
@@ -907,8 +906,8 @@ func (d *Status) MarkSignalConnected() {
 }
 
 func (d *Status) UpdateRelayStates(relayResults []relay.ProbeResult) {
-	d.muxRelays.Lock()
-	defer d.muxRelays.Unlock()
+	d.mux.Lock()
+	defer d.mux.Unlock()
 	d.relayStates = relayResults
 }
 
@@ -1019,26 +1018,24 @@ func (d *Status) GetSignalState() SignalState {
 
 // GetRelayStates returns the stun/turn/permanent relay states
 func (d *Status) GetRelayStates() []relay.ProbeResult {
-	d.muxRelays.RLock()
+	d.mux.RLock()
+	defer d.mux.RUnlock()
 	if d.relayMgr == nil {
-		defer d.muxRelays.RUnlock()
-		return slices.Clone(d.relayStates)
+		return d.relayStates
 	}
 
-	relayMgr := d.relayMgr
 	// extend the list of stun, turn servers with the relay server connections
 	relayStates := slices.Clone(d.relayStates)
-	d.muxRelays.RUnlock()
 
-	states := relayMgr.RelayStates()
+	states := d.relayMgr.RelayStates()
 	if len(states) == 0 {
 		// no relay connection tracked yet; surface configured servers as
 		// unavailable with the real reconnect error when known
 		err := relayClient.ErrRelayClientNotConnected
-		if connErr := relayMgr.RelayConnectError(); connErr != nil {
+		if connErr := d.relayMgr.RelayConnectError(); connErr != nil {
 			err = connErr
 		}
-		for _, r := range relayMgr.ServerURLs() {
+		for _, r := range d.relayMgr.ServerURLs() {
 			relayStates = append(relayStates, relay.ProbeResult{
 				URI: r,
 				Err: err,

client/internal/peer/wg_watcher.go

@@ -31,9 +31,7 @@ type WGWatcher struct {
 	stateDump     *stateDump
 
 	enabled   bool
-	muEnabled sync.Mutex
-	// initialHandshake is not thread-safe; never call PrepareInitialHandshake and EnableWgWatcher concurrently.
-	initialHandshake time.Time
+	muEnabled sync.RWMutex
 
 	resetCh chan struct{}
 }
@@ -48,38 +46,38 @@ func NewWGWatcher(log *log.Entry, wgIfaceStater WGInterfaceStater, peerKey strin
 	}
 }
 
-// PrepareInitialHandshake reserves the watcher and reads the peer's current WireGuard
-// handshake time. It must be called before the peer is (re)configured on the WireGuard
-// interface, so the captured baseline reflects the state prior to this connection attempt
-// instead of racing with that configuration. Returns ok=false if the watcher is already
-// running, in which case EnableWgWatcher must not be called.
-func (w *WGWatcher) PrepareInitialHandshake() (ok bool) {
+// EnableWgWatcher starts the WireGuard watcher. If it is already enabled, it will return immediately and do nothing.
+// The watcher runs until ctx is cancelled. Caller is responsible for context lifecycle management.
+func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time)) {
 	w.muEnabled.Lock()
 	if w.enabled {
 		w.muEnabled.Unlock()
-		return false
+		return
 	}
 
 	w.log.Debugf("enable WireGuard watcher")
 	w.enabled = true
 	w.muEnabled.Unlock()
 
-	handshake, _ := w.wgState()
-	w.initialHandshake = handshake
-	return true
-}
+	initialHandshake, err := w.wgState()
+	if err != nil {
+		w.log.Warnf("failed to read initial wg stats: %v", err)
+	}
 
-// EnableWgWatcher runs the WireGuard watcher loop using the handshake baseline captured by
-// PrepareInitialHandshake. The watcher runs until ctx is cancelled. Caller is responsible
-// for context lifecycle management.
-func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time)) {
-	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, enabledTime, w.initialHandshake)
+	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, enabledTime, initialHandshake)
 
 	w.muEnabled.Lock()
 	w.enabled = false
 	w.muEnabled.Unlock()
 }
 
+// IsEnabled returns true if the WireGuard watcher is currently enabled
+func (w *WGWatcher) IsEnabled() bool {
+	w.muEnabled.RLock()
+	defer w.muEnabled.RUnlock()
+	return w.enabled
+}
+
 // Reset signals the watcher that the WireGuard peer has been reset and a new
 // handshake is expected. This restarts the handshake timeout from scratch.
 func (w *WGWatcher) Reset() {
@@ -103,16 +101,13 @@ func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn
 		case <-timer.C:
 			handshake, ok := w.handshakeCheck(lastHandshake)
 			if !ok {
-				if ctx.Err() != nil {
-					return
-				}
 				onDisconnectedFn()
 				return
 			}
 			if lastHandshake.IsZero() {
 				elapsed := calcElapsed(enabledTime, *handshake)
 				w.log.Infof("first wg handshake detected within: %.2fsec, (%s)", elapsed, handshake)
-				if onHandshakeSuccessFn != nil && ctx.Err() == nil {
+				if onHandshakeSuccessFn != nil {
 					onHandshakeSuccessFn(*handshake)
 				}
 			}

client/internal/peer/wg_watcher_test.go

@@ -7,7 +7,6 @@ import (
 	"time"
 
 	log "github.com/sirupsen/logrus"
-	"github.com/stretchr/testify/require"
 
 	"github.com/netbirdio/netbird/client/iface/configurer"
 )
@@ -35,9 +34,6 @@ func TestWGWatcher_EnableWgWatcher(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
-	ok := watcher.PrepareInitialHandshake()
-	require.True(t, ok, "watcher should not be enabled yet")
-
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
 		mlog.Infof("onDisconnectedFn")
@@ -66,9 +62,6 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	watcher := NewWGWatcher(mlog, mocWgIface, "", newStateDump("peer", mlog, &Status{}))
 
 	ctx, cancel := context.WithCancel(context.Background())
-	ok := watcher.PrepareInitialHandshake()
-	require.True(t, ok, "watcher should not be enabled yet")
-
 	wg := &sync.WaitGroup{}
 	wg.Add(1)
 	go func() {
@@ -83,9 +76,6 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	ctx, cancel = context.WithCancel(context.Background())
 	defer cancel()
 
-	ok = watcher.PrepareInitialHandshake()
-	require.True(t, ok, "watcher should be re-enabled after the previous run stopped")
-
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
 		onDisconnected <- struct{}{}

client/internal/peerstore/store.go

@@ -88,24 +88,11 @@ func (s *Store) PeerConnOpen(ctx context.Context, pubKey string) {
 	if !ok {
 		return
 	}
+	// this can be blocked because of the connect open limiter semaphore
 	if err := p.Open(ctx); err != nil {
 		p.Log.Errorf("failed to open peer connection: %v", err)
 	}
-}
 
-// PeerConnOpenWithFirstPacket opens the peer connection and stashes a first packet to be
-// reinjected once the real transport is established.
-func (s *Store) PeerConnOpenWithFirstPacket(ctx context.Context, pubKey string, firstPacket []byte) {
-	s.peerConnsMu.RLock()
-	defer s.peerConnsMu.RUnlock()
-
-	p, ok := s.peerConns[pubKey]
-	if !ok {
-		return
-	}
-	if err := p.OpenWithFirstPacket(ctx, firstPacket); err != nil {
-		p.Log.Errorf("failed to open peer connection: %v", err)
-	}
 }
 
 func (s *Store) PeerConnIdle(pubKey string) {

client/internal/profilemanager/config.go

@@ -386,7 +386,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.NetworkMonitor != nil && (config.NetworkMonitor == nil || *input.NetworkMonitor != *config.NetworkMonitor) {
+	if input.NetworkMonitor != nil && input.NetworkMonitor != config.NetworkMonitor {
 		log.Infof("switching Network Monitor to %t", *input.NetworkMonitor)
 		config.NetworkMonitor = input.NetworkMonitor
 		updated = true
@@ -433,7 +433,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.ServerSSHAllowed != nil && (config.ServerSSHAllowed == nil || *input.ServerSSHAllowed != *config.ServerSSHAllowed) {
+	if input.ServerSSHAllowed != nil && *input.ServerSSHAllowed != *config.ServerSSHAllowed {
 		if *input.ServerSSHAllowed {
 			log.Infof("enabling SSH server")
 		} else {
@@ -454,7 +454,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.EnableSSHRoot != nil && (config.EnableSSHRoot == nil || *input.EnableSSHRoot != *config.EnableSSHRoot) {
+	if input.EnableSSHRoot != nil && input.EnableSSHRoot != config.EnableSSHRoot {
 		if *input.EnableSSHRoot {
 			log.Infof("enabling SSH root login")
 		} else {
@@ -464,7 +464,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.EnableSSHSFTP != nil && (config.EnableSSHSFTP == nil || *input.EnableSSHSFTP != *config.EnableSSHSFTP) {
+	if input.EnableSSHSFTP != nil && input.EnableSSHSFTP != config.EnableSSHSFTP {
 		if *input.EnableSSHSFTP {
 			log.Infof("enabling SSH SFTP subsystem")
 		} else {
@@ -474,7 +474,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.EnableSSHLocalPortForwarding != nil && (config.EnableSSHLocalPortForwarding == nil || *input.EnableSSHLocalPortForwarding != *config.EnableSSHLocalPortForwarding) {
+	if input.EnableSSHLocalPortForwarding != nil && input.EnableSSHLocalPortForwarding != config.EnableSSHLocalPortForwarding {
 		if *input.EnableSSHLocalPortForwarding {
 			log.Infof("enabling SSH local port forwarding")
 		} else {
@@ -484,7 +484,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.EnableSSHRemotePortForwarding != nil && (config.EnableSSHRemotePortForwarding == nil || *input.EnableSSHRemotePortForwarding != *config.EnableSSHRemotePortForwarding) {
+	if input.EnableSSHRemotePortForwarding != nil && input.EnableSSHRemotePortForwarding != config.EnableSSHRemotePortForwarding {
 		if *input.EnableSSHRemotePortForwarding {
 			log.Infof("enabling SSH remote port forwarding")
 		} else {
@@ -494,7 +494,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.DisableSSHAuth != nil && (config.DisableSSHAuth == nil || *input.DisableSSHAuth != *config.DisableSSHAuth) {
+	if input.DisableSSHAuth != nil && input.DisableSSHAuth != config.DisableSSHAuth {
 		if *input.DisableSSHAuth {
 			log.Infof("disabling SSH authentication")
 		} else {
@@ -504,7 +504,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.SSHJWTCacheTTL != nil && (config.SSHJWTCacheTTL == nil || *input.SSHJWTCacheTTL != *config.SSHJWTCacheTTL) {
+	if input.SSHJWTCacheTTL != nil && input.SSHJWTCacheTTL != config.SSHJWTCacheTTL {
 		log.Infof("updating SSH JWT cache TTL to %d seconds", *input.SSHJWTCacheTTL)
 		config.SSHJWTCacheTTL = input.SSHJWTCacheTTL
 		updated = true
@@ -587,7 +587,7 @@ func (config *Config) apply(input ConfigInput) (updated bool, err error) {
 		updated = true
 	}
 
-	if input.DisableNotifications != nil && (config.DisableNotifications == nil || *input.DisableNotifications != *config.DisableNotifications) {
+	if input.DisableNotifications != nil && input.DisableNotifications != config.DisableNotifications {
 		if *input.DisableNotifications {
 			log.Infof("disabling notifications")
 		} else {

client/internal/profilemanager/config_test.go

@@ -242,35 +242,6 @@ func TestWireguardPortDefaultVsExplicit(t *testing.T) {
 	}
 }
 
-func TestUpdateConfigServerSSHAllowedNotSet(t *testing.T) {
-	// Configs written before ServerSSHAllowed was introduced lack the field and
-	// unmarshal to nil. Supplying the SSH server flag on top of such a config must
-	// apply the value instead of panicking on a nil pointer dereference.
-	tests := []struct {
-		name  string
-		input *bool
-		want  bool
-	}{
-		{"enable", util.True(), true},
-		{"disable", util.False(), false},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			configPath := filepath.Join(t.TempDir(), "config.json")
-			require.NoError(t, os.WriteFile(configPath, []byte("{}"), 0600))
-
-			config, err := UpdateConfig(ConfigInput{
-				ConfigPath:       configPath,
-				ServerSSHAllowed: tt.input,
-			})
-			require.NoError(t, err)
-			require.NotNil(t, config.ServerSSHAllowed, "ServerSSHAllowed should be set from input")
-			assert.Equal(t, tt.want, *config.ServerSSHAllowed)
-		})
-	}
-}
-
 func TestUpdateOldManagementURL(t *testing.T) {
 	origProber := newMgmProber
 	newMgmProber = func(_ context.Context, _ string, _ wgtypes.Key, _ bool) (mgmProber, error) {

client/internal/routemanager/dnsinterceptor/handler.go

@@ -226,11 +226,12 @@ func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 		return
 	}
 
-	// All query types for an intercepted domain are forwarded to the peer's
-	// DNS forwarder, which owns the name. Falling through to the system
-	// resolver would let it answer NXDOMAIN for a name it isn't authoritative
-	// for, poisoning the whole name (including the A/AAAA records the route
-	// does serve). The forwarder answers NODATA for types it cannot resolve.
+	// pass if non A/AAAA query
+	if r.Question[0].Qtype != dns.TypeA && r.Question[0].Qtype != dns.TypeAAAA {
+		d.continueToNextHandler(w, r, logger, "non A/AAAA query")
+		return
+	}
+
 	d.mu.RLock()
 	peerKey := d.currentPeerKey
 	d.mu.RUnlock()
@@ -250,14 +251,6 @@ func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 		r.MsgHdr.AuthenticatedData = true
 	}
 
-	// Advertise EDNS0 to the forwarder so it may return an Extended DNS Error
-	// describing why a lookup failed. The OPT is stripped from the reply when
-	// the original client did not request EDNS0.
-	hadEdns := r.IsEdns0() != nil
-	if !hadEdns {
-		r.SetEdns0(dns.DefaultMsgSize, false)
-	}
-
 	upstream := net.JoinHostPort(upstreamIP.String(), strconv.FormatUint(uint64(d.forwarderPort.Load()), 10))
 	ctx, cancel := context.WithTimeout(context.Background(), dnsTimeout)
 	defer cancel()
@@ -267,13 +260,6 @@ func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 		return
 	}
 
-	if ede, ok := resutil.ExtractEDE(reply); ok {
-		resutil.SetMeta(w, "ede", fmt.Sprintf("%d %s", ede.InfoCode, ede.ExtraText))
-	}
-	if !hadEdns {
-		resutil.StripOPT(reply)
-	}
-
 	resutil.SetMeta(w, "peer", peerKey)
 
 	reply.Id = r.Id
@@ -292,6 +278,19 @@ func (d *DnsInterceptor) writeDNSError(w dns.ResponseWriter, r *dns.Msg, logger
 	}
 }
 
+// continueToNextHandler signals the handler chain to try the next handler
+func (d *DnsInterceptor) continueToNextHandler(w dns.ResponseWriter, r *dns.Msg, logger *log.Entry, reason string) {
+	logger.Tracef("continuing to next handler for domain=%s reason=%s", r.Question[0].Name, reason)
+
+	resp := new(dns.Msg)
+	resp.SetRcode(r, dns.RcodeNameError)
+	// Set Zero bit to signal handler chain to continue
+	resp.MsgHdr.Zero = true
+	if err := w.WriteMsg(resp); err != nil {
+		logger.Errorf("failed writing DNS continue response: %v", err)
+	}
+}
+
 func (d *DnsInterceptor) getUpstreamIP(peerKey string) (netip.Addr, error) {
 	peerAllowedIP, exists := d.peerStore.AllowedIP(peerKey)
 	if !exists {

client/internal/routemanager/exit_node_selection_test.go

@@ -1,191 +0,0 @@
-package routemanager
-
-import (
-	"net/netip"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-
-	"github.com/netbirdio/netbird/client/internal/routeselector"
-	"github.com/netbirdio/netbird/route"
-)
-
-func newExitNodeTestManager() *DefaultManager {
-	return &DefaultManager{routeSelector: routeselector.NewRouteSelector()}
-}
-
-func exitRoute(netID, peer string, skipAutoApply bool) *route.Route {
-	return &route.Route{
-		NetID:         route.NetID(netID),
-		Network:       netip.MustParsePrefix("0.0.0.0/0"),
-		Peer:          peer,
-		SkipAutoApply: skipAutoApply,
-	}
-}
-
-func TestPickPreferredExitNode(t *testing.T) {
-	tests := []struct {
-		name string
-		info exitNodeInfo
-		want route.NetID
-	}{
-		{
-			name: "persisted user selection wins over management",
-			info: exitNodeInfo{
-				allIDs:               []route.NetID{"a", "b", "c"},
-				userSelected:         []route.NetID{"b"},
-				selectedByManagement: []route.NetID{"a"},
-			},
-			want: "b",
-		},
-		{
-			name: "multiple user-selected self-heal to deterministic min",
-			info: exitNodeInfo{
-				allIDs:       []route.NetID{"a", "b", "c"},
-				userSelected: []route.NetID{"c", "a"},
-			},
-			want: "a",
-		},
-		{
-			name: "explicit opt-out keeps none",
-			info: exitNodeInfo{
-				allIDs:         []route.NetID{"a", "b"},
-				userDeselected: []route.NetID{"a", "b"},
-			},
-			want: "",
-		},
-		{
-			name: "fresh defaults to management auto-apply pick",
-			info: exitNodeInfo{
-				allIDs:               []route.NetID{"a", "b", "c"},
-				selectedByManagement: []route.NetID{"b"},
-			},
-			want: "b",
-		},
-		{
-			name: "no user pick and no management auto-apply selects none",
-			info: exitNodeInfo{
-				allIDs: []route.NetID{"c", "a", "b"},
-			},
-			want: "",
-		},
-		{
-			name: "user-deselect does not block a management auto-apply sibling",
-			info: exitNodeInfo{
-				allIDs:               []route.NetID{"a", "b"},
-				userDeselected:       []route.NetID{"a"},
-				selectedByManagement: []route.NetID{"b"},
-			},
-			want: "b",
-		},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			assert.Equal(t, tt.want, pickPreferredExitNode(tt.info), "preferred exit node")
-		})
-	}
-}
-
-func TestEnforceSingleExitNode(t *testing.T) {
-	m := newExitNodeTestManager()
-	all := []route.NetID{"a", "b", "c"}
-
-	m.enforceSingleExitNode("b", all)
-	assert.False(t, m.routeSelector.IsSelected("a"), "a should be deselected")
-	assert.True(t, m.routeSelector.IsSelected("b"), "b should be the only selected exit node")
-	assert.False(t, m.routeSelector.IsSelected("c"), "c should be deselected")
-
-	// Switching the preferred node moves the single selection.
-	m.enforceSingleExitNode("c", all)
-	assert.False(t, m.routeSelector.IsSelected("a"), "a stays deselected")
-	assert.False(t, m.routeSelector.IsSelected("b"), "b should now be deselected")
-	assert.True(t, m.routeSelector.IsSelected("c"), "c should now be selected")
-
-	// Empty preferred turns every exit node off.
-	m.enforceSingleExitNode("", all)
-	for _, id := range all {
-		assert.False(t, m.routeSelector.IsSelected(id), "no exit node should be selected")
-	}
-}
-
-func TestEnforceSingleExitNode_RespectsDeselectAll(t *testing.T) {
-	m := newExitNodeTestManager()
-	m.routeSelector.DeselectAllRoutes()
-
-	m.enforceSingleExitNode("b", []route.NetID{"a", "b"})
-
-	assert.True(t, m.routeSelector.IsDeselectAll(), "global deselect-all must stay in effect")
-	assert.False(t, m.routeSelector.IsSelected("b"), "no exit node should be forced on while deselect-all is set")
-}
-
-func TestUpdateRouteSelectorFromManagement_FreshSelectsOne(t *testing.T) {
-	m := newExitNodeTestManager()
-	routes := route.HAMap{
-		"exitA|0.0.0.0/0":    {exitRoute("exitA", "p1", false)},
-		"exitB|0.0.0.0/0":    {exitRoute("exitB", "p2", false)},
-		"lan|192.168.1.0/24": {{NetID: "lan", Network: netip.MustParsePrefix("192.168.1.0/24"), Peer: "p3"}},
-		"exitC|0.0.0.0/0":    {exitRoute("exitC", "p4", false)},
-	}
-
-	m.updateRouteSelectorFromManagement(routes)
-
-	// Exactly one exit node (the deterministic first) is selected.
-	assert.True(t, m.routeSelector.IsSelected("exitA"), "exitA is the deterministic default")
-	assert.False(t, m.routeSelector.IsSelected("exitB"), "exitB must not also be selected")
-	assert.False(t, m.routeSelector.IsSelected("exitC"), "exitC must not also be selected")
-	// Non-exit routes are left at their default-on state.
-	assert.True(t, m.routeSelector.IsSelected("lan"), "non-exit route selection is untouched")
-}
-
-func TestUpdateRouteSelectorFromManagement_HonorsPersistedPick(t *testing.T) {
-	m := newExitNodeTestManager()
-	routes := route.HAMap{
-		"exitA|0.0.0.0/0": {exitRoute("exitA", "p1", false)},
-		"exitB|0.0.0.0/0": {exitRoute("exitB", "p2", false)},
-	}
-	all := []route.NetID{"exitA", "exitB"}
-
-	// Simulate the state the runtime select path leaves behind: exactly one
-	// exit node explicitly selected, its sibling deselected.
-	require.NoError(t, m.routeSelector.SelectRoutes([]route.NetID{"exitB"}, true, all))
-	require.NoError(t, m.routeSelector.DeselectRoutes([]route.NetID{"exitA"}, all))
-
-	m.updateRouteSelectorFromManagement(routes)
-
-	assert.True(t, m.routeSelector.IsSelected("exitB"), "persisted pick must stay selected")
-	assert.False(t, m.routeSelector.IsSelected("exitA"), "the other exit node stays deselected")
-}
-
-func TestUpdateRouteSelectorFromManagement_OptOutKeepsNone(t *testing.T) {
-	m := newExitNodeTestManager()
-	routes := route.HAMap{
-		"exitA|0.0.0.0/0": {exitRoute("exitA", "p1", false)},
-		"exitB|0.0.0.0/0": {exitRoute("exitB", "p2", false)},
-	}
-	all := []route.NetID{"exitA", "exitB"}
-
-	// User deselected exit nodes and selected none.
-	require.NoError(t, m.routeSelector.DeselectRoutes(all, all))
-
-	m.updateRouteSelectorFromManagement(routes)
-
-	assert.False(t, m.routeSelector.IsSelected("exitA"), "opt-out keeps exitA off")
-	assert.False(t, m.routeSelector.IsSelected("exitB"), "opt-out keeps exitB off")
-}
-
-func TestUpdateRouteSelectorFromManagement_NoAutoApplySelectsNone(t *testing.T) {
-	m := newExitNodeTestManager()
-	// SkipAutoApply=true: management offers the exit nodes but doesn't request
-	// auto-activation, so none should be selected until the user picks one.
-	routes := route.HAMap{
-		"exitA|0.0.0.0/0": {exitRoute("exitA", "p1", true)},
-		"exitB|0.0.0.0/0": {exitRoute("exitB", "p2", true)},
-	}
-
-	m.updateRouteSelectorFromManagement(routes)
-
-	assert.False(t, m.routeSelector.IsSelected("exitA"), "no auto-apply keeps exitA off")
-	assert.False(t, m.routeSelector.IsSelected("exitB"), "no auto-apply keeps exitB off")
-}

client/internal/routemanager/manager.go

@@ -701,13 +701,7 @@ func resolveURLsToIPs(urls []string) []net.IP {
 	return ips
 }
 
-// updateRouteSelectorFromManagement reconciles exit-node selection on every
-// network map: it keeps at most one exit node selected — the user's persisted
-// pick, else whatever management marks for auto-apply (SkipAutoApply=false),
-// else none. We never auto-activate an exit node the map doesn't request; it
-// stays off until the user picks it. Exit nodes are mutually exclusive, but the
-// RouteSelector stores routes with default-on semantics, so without this every
-// available exit node would report selected at once.
+// updateRouteSelectorFromManagement updates the route selector based on the isSelected status from the management server
 func (m *DefaultManager) updateRouteSelectorFromManagement(clientRoutes route.HAMap) {
 	m.mirrorV6ExitPairSelections(clientRoutes)
 
@@ -718,14 +712,13 @@ func (m *DefaultManager) updateRouteSelectorFromManagement(clientRoutes route.HA
 		return
 	}
 
-	info := m.collectExitNodeInfo(clientRoutes)
-	if len(info.allIDs) == 0 {
+	exitNodeInfo := m.collectExitNodeInfo(clientRoutes)
+	if len(exitNodeInfo.allIDs) == 0 {
 		return
 	}
 
-	preferred := pickPreferredExitNode(info)
-	m.enforceSingleExitNode(preferred, info.allIDs)
-	m.logExitNodeUpdate(info, preferred)
+	m.updateExitNodeSelections(exitNodeInfo)
+	m.logExitNodeUpdate(exitNodeInfo)
 }
 
 // mirrorV6ExitPairSelections keeps every synthesized "-v6" exit route's selection
@@ -753,10 +746,6 @@ type exitNodeInfo struct {
 	userDeselected       []route.NetID
 }
 
-// collectExitNodeInfo categorises the available exit nodes by their persisted
-// selection state. It keys on the base (v4) NetID and skips the synthesized
-// "-v6" partner, which inherits its base's selection through the RouteSelector
-// — counting it separately would double-count the pair.
 func (m *DefaultManager) collectExitNodeInfo(clientRoutes route.HAMap) exitNodeInfo {
 	var info exitNodeInfo
 
@@ -766,9 +755,6 @@ func (m *DefaultManager) collectExitNodeInfo(clientRoutes route.HAMap) exitNodeI
 		}
 
 		netID := haID.NetID()
-		if strings.HasSuffix(string(netID), route.V6ExitSuffix) {
-			continue
-		}
 		info.allIDs = append(info.allIDs, netID)
 
 		if m.routeSelector.HasUserSelectionForRoute(netID) {
@@ -805,52 +791,45 @@ func (m *DefaultManager) checkManagementSelection(routes []*route.Route, netID r
 	}
 }
 
-// pickPreferredExitNode chooses the single exit node to keep selected. In order:
-//   - a persisted user selection wins (deterministic if several survive from
-//     legacy state, so the set self-heals down to one);
-//   - otherwise activate only what management marks for auto-apply
-//     (SkipAutoApply=false); the lexicographically first if it marks several.
-//
-// Returns "" when neither holds — we never force an arbitrary exit node on. A
-// route the map doesn't auto-apply stays off until the user selects it.
-// info.userDeselected is informational only: an explicit deselect simply keeps
-// that route out of both lists above, so it can't be picked.
-func pickPreferredExitNode(info exitNodeInfo) route.NetID {
-	if len(info.userSelected) > 0 {
-		return minNetID(info.userSelected)
-	}
-	if len(info.selectedByManagement) > 0 {
-		return minNetID(info.selectedByManagement)
-	}
-	return ""
+func (m *DefaultManager) updateExitNodeSelections(info exitNodeInfo) {
+	routesToDeselect := m.getRoutesToDeselect(info.allIDs)
+	m.deselectExitNodes(routesToDeselect)
+	m.selectExitNodesByManagement(info.selectedByManagement, info.allIDs)
 }
 
-// enforceSingleExitNode makes preferred the only selected exit node: every other
-// available exit node is deselected and preferred (if any) is selected, without
-// disturbing non-exit route selections. The whole reconciliation runs under a
-// single RouteSelector lock (SetExclusiveExitNode) so a concurrent deselect-all
-// cannot interleave and get undone; a global deselect-all is left untouched so
-// the user's "all off" stays in effect.
-func (m *DefaultManager) enforceSingleExitNode(preferred route.NetID, allIDs []route.NetID) {
-	m.routeSelector.SetExclusiveExitNode(preferred, allIDs)
-}
-
-func (m *DefaultManager) logExitNodeUpdate(info exitNodeInfo, preferred route.NetID) {
-	log.Debugf("Exit node selection: %d available, preferred=%q (%d user-selected, %d user-deselected, %d management-selected)",
-		len(info.allIDs), preferred, len(info.userSelected), len(info.userDeselected), len(info.selectedByManagement))
-}
-
-// minNetID returns the lexicographically smallest NetID, for a deterministic
-// default pick that stays stable across restarts.
-func minNetID(ids []route.NetID) route.NetID {
-	if len(ids) == 0 {
-		return ""
-	}
-	best := ids[0]
-	for _, id := range ids[1:] {
-		if id < best {
-			best = id
+func (m *DefaultManager) getRoutesToDeselect(allIDs []route.NetID) []route.NetID {
+	var routesToDeselect []route.NetID
+	for _, netID := range allIDs {
+		if !m.routeSelector.HasUserSelectionForRoute(netID) {
+			routesToDeselect = append(routesToDeselect, netID)
 		}
 	}
-	return best
+	return routesToDeselect
+}
+
+func (m *DefaultManager) deselectExitNodes(routesToDeselect []route.NetID) {
+	if len(routesToDeselect) == 0 {
+		return
+	}
+
+	err := m.routeSelector.DeselectRoutes(routesToDeselect, routesToDeselect)
+	if err != nil {
+		log.Warnf("Failed to deselect exit nodes: %v", err)
+	}
+}
+
+func (m *DefaultManager) selectExitNodesByManagement(selectedByManagement []route.NetID, allIDs []route.NetID) {
+	if len(selectedByManagement) == 0 {
+		return
+	}
+
+	err := m.routeSelector.SelectRoutes(selectedByManagement, true, allIDs)
+	if err != nil {
+		log.Warnf("Failed to select exit nodes: %v", err)
+	}
+}
+
+func (m *DefaultManager) logExitNodeUpdate(info exitNodeInfo) {
+	log.Debugf("Updated route selector: %d exit nodes available, %d selected by management, %d user-selected, %d user-deselected",
+		len(info.allIDs), len(info.selectedByManagement), len(info.userSelected), len(info.userDeselected))
 }

client/internal/routemanager/manager_test.go

@@ -1,5 +1,3 @@
-//go:build privileged
-
 package routemanager
 
 import (

client/internal/routemanager/systemops/rt_tables_linux_test.go

@@ -1,69 +0,0 @@
-//go:build linux && !android
-
-package systemops
-
-import (
-	"fmt"
-	"os"
-	"strings"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-func TestEntryExists(t *testing.T) {
-	tempDir := t.TempDir()
-	tempFilePath := fmt.Sprintf("%s/rt_tables", tempDir)
-
-	content := []string{
-		"1000 reserved",
-		fmt.Sprintf("%d %s", NetbirdVPNTableID, NetbirdVPNTableName),
-		"9999 other_table",
-	}
-	require.NoError(t, os.WriteFile(tempFilePath, []byte(strings.Join(content, "\n")), 0644))
-
-	file, err := os.Open(tempFilePath)
-	require.NoError(t, err)
-	defer func() {
-		assert.NoError(t, file.Close())
-	}()
-
-	tests := []struct {
-		name        string
-		id          int
-		shouldExist bool
-		err         error
-	}{
-		{
-			name:        "ExistsWithNetbirdPrefix",
-			id:          7120,
-			shouldExist: true,
-			err:         nil,
-		},
-		{
-			name:        "ExistsWithDifferentName",
-			id:          1000,
-			shouldExist: true,
-			err:         ErrTableIDExists,
-		},
-		{
-			name:        "DoesNotExist",
-			id:          1234,
-			shouldExist: false,
-			err:         nil,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			exists, err := entryExists(file, tc.id)
-			if tc.err != nil {
-				assert.ErrorIs(t, err, tc.err)
-			} else {
-				assert.NoError(t, err)
-			}
-			assert.Equal(t, tc.shouldExist, exists)
-		})
-	}
-}

client/internal/routemanager/systemops/systemops_bsd_privileged_test.go

@@ -1,191 +0,0 @@
-//go:build (darwin || dragonfly || freebsd || netbsd || openbsd) && privileged
-
-package systemops
-
-import (
-	"fmt"
-	"net"
-	"net/netip"
-	"os/exec"
-	"regexp"
-	"runtime"
-	"strings"
-	"sync"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-)
-
-func init() {
-	testCases = append(testCases, []testCase{
-		{
-			name:              "To more specific route without custom dialer via vpn",
-			expectedInterface: expectedVPNint,
-			dialer:            &net.Dialer{},
-			expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "10.10.0.2", 53),
-		},
-	}...)
-}
-
-func TestConcurrentRoutes(t *testing.T) {
-	baseIP := netip.MustParseAddr("192.0.2.0")
-
-	var intf *net.Interface
-	var nexthop Nexthop
-
-	_, intf = setupDummyInterface(t)
-	nexthop = Nexthop{netip.Addr{}, intf}
-
-	r := New(nil, nil)
-
-	var wg sync.WaitGroup
-	for i := 0; i < 1024; i++ {
-		wg.Add(1)
-		go func(ip netip.Addr) {
-			defer wg.Done()
-			prefix := netip.PrefixFrom(ip, 32)
-			if err := r.addToRouteTable(prefix, nexthop); err != nil {
-				t.Errorf("Failed to add route for %s: %v", prefix, err)
-			}
-		}(baseIP)
-		baseIP = baseIP.Next()
-	}
-
-	wg.Wait()
-
-	baseIP = netip.MustParseAddr("192.0.2.0")
-
-	for i := 0; i < 1024; i++ {
-		wg.Add(1)
-		go func(ip netip.Addr) {
-			defer wg.Done()
-			prefix := netip.PrefixFrom(ip, 32)
-			if err := r.removeFromRouteTable(prefix, nexthop); err != nil {
-				t.Errorf("Failed to remove route for %s: %v", prefix, err)
-			}
-		}(baseIP)
-		baseIP = baseIP.Next()
-	}
-
-	wg.Wait()
-}
-
-func createAndSetupDummyInterface(t *testing.T, intf string, ipAddressCIDR string) string {
-	t.Helper()
-
-	if runtime.GOOS == "darwin" {
-		err := exec.Command("ifconfig", intf, "alias", ipAddressCIDR).Run()
-		require.NoError(t, err, "Failed to create loopback alias")
-
-		t.Cleanup(func() {
-			err := exec.Command("ifconfig", intf, ipAddressCIDR, "-alias").Run()
-			assert.NoError(t, err, "Failed to remove loopback alias")
-		})
-
-		return intf
-	}
-
-	prefix, err := netip.ParsePrefix(ipAddressCIDR)
-	require.NoError(t, err, "Failed to parse prefix")
-
-	netIntf, err := net.InterfaceByName(intf)
-	require.NoError(t, err, "Failed to get interface by name")
-
-	nexthop := Nexthop{netip.Addr{}, netIntf}
-
-	r := New(nil, nil)
-	err = r.addToRouteTable(prefix, nexthop)
-	require.NoError(t, err, "Failed to add route to table")
-
-	t.Cleanup(func() {
-		err := r.removeFromRouteTable(prefix, nexthop)
-		assert.NoError(t, err, "Failed to remove route from table")
-	})
-
-	return intf
-}
-
-func addDummyRoute(t *testing.T, dstCIDR string, gw netip.Addr, _ string) {
-	t.Helper()
-
-	var originalNexthop net.IP
-	if dstCIDR == "0.0.0.0/0" {
-		var err error
-		originalNexthop, err = fetchOriginalGateway()
-		if err != nil {
-			t.Logf("Failed to fetch original gateway: %v", err)
-		}
-
-		if output, err := exec.Command("route", "delete", "-net", dstCIDR).CombinedOutput(); err != nil {
-			t.Logf("Failed to delete route: %v, output: %s", err, output)
-		}
-	}
-
-	t.Cleanup(func() {
-		if originalNexthop != nil {
-			err := exec.Command("route", "add", "-net", dstCIDR, originalNexthop.String()).Run()
-			assert.NoError(t, err, "Failed to restore original route")
-		}
-	})
-
-	err := exec.Command("route", "add", "-net", dstCIDR, gw.String()).Run()
-	require.NoError(t, err, "Failed to add route")
-
-	t.Cleanup(func() {
-		err := exec.Command("route", "delete", "-net", dstCIDR).Run()
-		assert.NoError(t, err, "Failed to remove route")
-	})
-}
-
-func fetchOriginalGateway() (net.IP, error) {
-	output, err := exec.Command("route", "-n", "get", "default").CombinedOutput()
-	if err != nil {
-		return nil, err
-	}
-
-	matches := regexp.MustCompile(`gateway: (\S+)`).FindStringSubmatch(string(output))
-	if len(matches) == 0 {
-		return nil, fmt.Errorf("gateway not found")
-	}
-
-	return net.ParseIP(matches[1]), nil
-}
-
-// setupDummyInterface creates a dummy tun interface for FreeBSD route testing
-func setupDummyInterface(t *testing.T) (netip.Addr, *net.Interface) {
-	t.Helper()
-
-	if runtime.GOOS == "darwin" {
-		return netip.AddrFrom4([4]byte{192, 168, 1, 2}), &net.Interface{Name: "lo0"}
-	}
-
-	output, err := exec.Command("ifconfig", "tun", "create").CombinedOutput()
-	require.NoError(t, err, "Failed to create tun interface: %s", string(output))
-
-	tunName := strings.TrimSpace(string(output))
-
-	output, err = exec.Command("ifconfig", tunName, "192.168.1.1", "netmask", "255.255.0.0", "192.168.1.2", "up").CombinedOutput()
-	require.NoError(t, err, "Failed to configure tun interface: %s", string(output))
-
-	intf, err := net.InterfaceByName(tunName)
-	require.NoError(t, err, "Failed to get interface by name")
-
-	t.Cleanup(func() {
-		if err := exec.Command("ifconfig", tunName, "destroy").Run(); err != nil {
-			t.Logf("Failed to destroy tun interface %s: %v", tunName, err)
-		}
-	})
-
-	return netip.AddrFrom4([4]byte{192, 168, 1, 2}), intf
-}
-
-func setupDummyInterfacesAndRoutes(t *testing.T) {
-	t.Helper()
-
-	defaultDummy := createAndSetupDummyInterface(t, expectedExternalInt, "192.168.0.1/24")
-	addDummyRoute(t, "0.0.0.0/0", netip.AddrFrom4([4]byte{192, 168, 0, 1}), defaultDummy)
-
-	otherDummy := createAndSetupDummyInterface(t, expectedInternalInt, "192.168.1.1/24")
-	addDummyRoute(t, "10.0.0.0/8", netip.AddrFrom4([4]byte{192, 168, 1, 1}), otherDummy)
-}

client/internal/routemanager/systemops/systemops_bsd_test.go

@@ -3,24 +3,79 @@
 package systemops
 
 import (
+	"fmt"
+	"net"
+	"net/netip"
+	"os/exec"
+	"regexp"
+	"runtime"
+	"strings"
+	"sync"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 	"golang.org/x/net/route"
 )
 
-// Interface names used by the shared routing test fixtures. Kept untagged (no
-// privileged build tag) so the non-privileged test files in this package compile.
-//
-//nolint:unused // consumed by the privileged-tagged routing tests
 var expectedVPNint = "utun100"
-
-//nolint:unused // consumed by the privileged-tagged routing tests
 var expectedExternalInt = "lo0"
-
-//nolint:unused // consumed by the privileged-tagged routing tests
 var expectedInternalInt = "lo0"
 
+func init() {
+	testCases = append(testCases, []testCase{
+		{
+			name:              "To more specific route without custom dialer via vpn",
+			expectedInterface: expectedVPNint,
+			dialer:            &net.Dialer{},
+			expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "10.10.0.2", 53),
+		},
+	}...)
+}
+
+func TestConcurrentRoutes(t *testing.T) {
+	baseIP := netip.MustParseAddr("192.0.2.0")
+
+	var intf *net.Interface
+	var nexthop Nexthop
+
+	_, intf = setupDummyInterface(t)
+	nexthop = Nexthop{netip.Addr{}, intf}
+
+	r := New(nil, nil)
+
+	var wg sync.WaitGroup
+	for i := 0; i < 1024; i++ {
+		wg.Add(1)
+		go func(ip netip.Addr) {
+			defer wg.Done()
+			prefix := netip.PrefixFrom(ip, 32)
+			if err := r.addToRouteTable(prefix, nexthop); err != nil {
+				t.Errorf("Failed to add route for %s: %v", prefix, err)
+			}
+		}(baseIP)
+		baseIP = baseIP.Next()
+	}
+
+	wg.Wait()
+
+	baseIP = netip.MustParseAddr("192.0.2.0")
+
+	for i := 0; i < 1024; i++ {
+		wg.Add(1)
+		go func(ip netip.Addr) {
+			defer wg.Done()
+			prefix := netip.PrefixFrom(ip, 32)
+			if err := r.removeFromRouteTable(prefix, nexthop); err != nil {
+				t.Errorf("Failed to remove route for %s: %v", prefix, err)
+			}
+		}(baseIP)
+		baseIP = baseIP.Next()
+	}
+
+	wg.Wait()
+}
+
 func TestBits(t *testing.T) {
 	tests := []struct {
 		name    string
@@ -67,3 +122,122 @@ func TestBits(t *testing.T) {
 		})
 	}
 }
+
+func createAndSetupDummyInterface(t *testing.T, intf string, ipAddressCIDR string) string {
+	t.Helper()
+
+	if runtime.GOOS == "darwin" {
+		err := exec.Command("ifconfig", intf, "alias", ipAddressCIDR).Run()
+		require.NoError(t, err, "Failed to create loopback alias")
+
+		t.Cleanup(func() {
+			err := exec.Command("ifconfig", intf, ipAddressCIDR, "-alias").Run()
+			assert.NoError(t, err, "Failed to remove loopback alias")
+		})
+
+		return intf
+	}
+
+	prefix, err := netip.ParsePrefix(ipAddressCIDR)
+	require.NoError(t, err, "Failed to parse prefix")
+
+	netIntf, err := net.InterfaceByName(intf)
+	require.NoError(t, err, "Failed to get interface by name")
+
+	nexthop := Nexthop{netip.Addr{}, netIntf}
+
+	r := New(nil, nil)
+	err = r.addToRouteTable(prefix, nexthop)
+	require.NoError(t, err, "Failed to add route to table")
+
+	t.Cleanup(func() {
+		err := r.removeFromRouteTable(prefix, nexthop)
+		assert.NoError(t, err, "Failed to remove route from table")
+	})
+
+	return intf
+}
+
+func addDummyRoute(t *testing.T, dstCIDR string, gw netip.Addr, _ string) {
+	t.Helper()
+
+	var originalNexthop net.IP
+	if dstCIDR == "0.0.0.0/0" {
+		var err error
+		originalNexthop, err = fetchOriginalGateway()
+		if err != nil {
+			t.Logf("Failed to fetch original gateway: %v", err)
+		}
+
+		if output, err := exec.Command("route", "delete", "-net", dstCIDR).CombinedOutput(); err != nil {
+			t.Logf("Failed to delete route: %v, output: %s", err, output)
+		}
+	}
+
+	t.Cleanup(func() {
+		if originalNexthop != nil {
+			err := exec.Command("route", "add", "-net", dstCIDR, originalNexthop.String()).Run()
+			assert.NoError(t, err, "Failed to restore original route")
+		}
+	})
+
+	err := exec.Command("route", "add", "-net", dstCIDR, gw.String()).Run()
+	require.NoError(t, err, "Failed to add route")
+
+	t.Cleanup(func() {
+		err := exec.Command("route", "delete", "-net", dstCIDR).Run()
+		assert.NoError(t, err, "Failed to remove route")
+	})
+}
+
+func fetchOriginalGateway() (net.IP, error) {
+	output, err := exec.Command("route", "-n", "get", "default").CombinedOutput()
+	if err != nil {
+		return nil, err
+	}
+
+	matches := regexp.MustCompile(`gateway: (\S+)`).FindStringSubmatch(string(output))
+	if len(matches) == 0 {
+		return nil, fmt.Errorf("gateway not found")
+	}
+
+	return net.ParseIP(matches[1]), nil
+}
+
+// setupDummyInterface creates a dummy tun interface for FreeBSD route testing
+func setupDummyInterface(t *testing.T) (netip.Addr, *net.Interface) {
+	t.Helper()
+
+	if runtime.GOOS == "darwin" {
+		return netip.AddrFrom4([4]byte{192, 168, 1, 2}), &net.Interface{Name: "lo0"}
+	}
+
+	output, err := exec.Command("ifconfig", "tun", "create").CombinedOutput()
+	require.NoError(t, err, "Failed to create tun interface: %s", string(output))
+
+	tunName := strings.TrimSpace(string(output))
+
+	output, err = exec.Command("ifconfig", tunName, "192.168.1.1", "netmask", "255.255.0.0", "192.168.1.2", "up").CombinedOutput()
+	require.NoError(t, err, "Failed to configure tun interface: %s", string(output))
+
+	intf, err := net.InterfaceByName(tunName)
+	require.NoError(t, err, "Failed to get interface by name")
+
+	t.Cleanup(func() {
+		if err := exec.Command("ifconfig", tunName, "destroy").Run(); err != nil {
+			t.Logf("Failed to destroy tun interface %s: %v", tunName, err)
+		}
+	})
+
+	return netip.AddrFrom4([4]byte{192, 168, 1, 2}), intf
+}
+
+func setupDummyInterfacesAndRoutes(t *testing.T) {
+	t.Helper()
+
+	defaultDummy := createAndSetupDummyInterface(t, expectedExternalInt, "192.168.0.1/24")
+	addDummyRoute(t, "0.0.0.0/0", netip.AddrFrom4([4]byte{192, 168, 0, 1}), defaultDummy)
+
+	otherDummy := createAndSetupDummyInterface(t, expectedInternalInt, "192.168.1.1/24")
+	addDummyRoute(t, "10.0.0.0/8", netip.AddrFrom4([4]byte{192, 168, 1, 1}), otherDummy)
+}

client/internal/routemanager/systemops/systemops_dialer_test.go

@@ -1,17 +0,0 @@
-//go:build !android && !ios
-
-package systemops
-
-import (
-	"context"
-	"net"
-)
-
-// dialer is shared by the per-platform routing test cases. Kept untagged (no
-// privileged build tag) so the non-privileged test files compile on every platform.
-//
-//nolint:unused // consumed by the privileged-tagged routing tests
-type dialer interface {
-	Dial(network, address string) (net.Conn, error)
-	DialContext(ctx context.Context, network, address string) (net.Conn, error)
-}

client/internal/routemanager/systemops/systemops_generic_test.go

@@ -1,4 +1,4 @@
-//go:build !android && !ios && privileged
+//go:build !android && !ios
 
 package systemops
 
@@ -26,6 +26,11 @@ import (
 	nbnet "github.com/netbirdio/netbird/client/net"
 )
 
+type dialer interface {
+	Dial(network, address string) (net.Conn, error)
+	DialContext(ctx context.Context, network, address string) (net.Conn, error)
+}
+
 func TestAddVPNRoute(t *testing.T) {
 	testCases := []struct {
 		name        string
@@ -510,3 +515,125 @@ func setupTestEnv(t *testing.T) {
 	// unique route in vpn table
 	setupRouteAndCleanup(t, r, netip.MustParsePrefix("172.16.0.0/12"), intf)
 }
+
+func TestIsVpnRoute(t *testing.T) {
+	tests := []struct {
+		name           string
+		addr           string
+		vpnRoutes      []string
+		localRoutes    []string
+		expectedVpn    bool
+		expectedPrefix netip.Prefix
+	}{
+		{
+			name:           "Match in VPN routes",
+			addr:           "192.168.1.1",
+			vpnRoutes:      []string{"192.168.1.0/24"},
+			localRoutes:    []string{"10.0.0.0/8"},
+			expectedVpn:    true,
+			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
+		},
+		{
+			name:           "Match in local routes",
+			addr:           "10.1.1.1",
+			vpnRoutes:      []string{"192.168.1.0/24"},
+			localRoutes:    []string{"10.0.0.0/8"},
+			expectedVpn:    false,
+			expectedPrefix: netip.MustParsePrefix("10.0.0.0/8"),
+		},
+		{
+			name:           "No match",
+			addr:           "172.16.0.1",
+			vpnRoutes:      []string{"192.168.1.0/24"},
+			localRoutes:    []string{"10.0.0.0/8"},
+			expectedVpn:    false,
+			expectedPrefix: netip.Prefix{},
+		},
+		{
+			name:           "Default route ignored",
+			addr:           "192.168.1.1",
+			vpnRoutes:      []string{"0.0.0.0/0", "192.168.1.0/24"},
+			localRoutes:    []string{"10.0.0.0/8"},
+			expectedVpn:    true,
+			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
+		},
+		{
+			name:           "Default route matches but ignored",
+			addr:           "172.16.1.1",
+			vpnRoutes:      []string{"0.0.0.0/0", "192.168.1.0/24"},
+			localRoutes:    []string{"10.0.0.0/8"},
+			expectedVpn:    false,
+			expectedPrefix: netip.Prefix{},
+		},
+		{
+			name:           "Longest prefix match local",
+			addr:           "192.168.1.1",
+			vpnRoutes:      []string{"192.168.0.0/16"},
+			localRoutes:    []string{"192.168.1.0/24"},
+			expectedVpn:    false,
+			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
+		},
+		{
+			name:           "Longest prefix match local multiple",
+			addr:           "192.168.0.1",
+			vpnRoutes:      []string{"192.168.0.0/16", "192.168.0.0/25", "192.168.0.0/27"},
+			localRoutes:    []string{"192.168.0.0/24", "192.168.0.0/26", "192.168.0.0/28"},
+			expectedVpn:    false,
+			expectedPrefix: netip.MustParsePrefix("192.168.0.0/28"),
+		},
+		{
+			name:           "Longest prefix match vpn",
+			addr:           "192.168.1.1",
+			vpnRoutes:      []string{"192.168.1.0/24"},
+			localRoutes:    []string{"192.168.0.0/16"},
+			expectedVpn:    true,
+			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
+		},
+		{
+			name:           "Longest prefix match vpn multiple",
+			addr:           "192.168.0.1",
+			vpnRoutes:      []string{"192.168.0.0/16", "192.168.0.0/25", "192.168.0.0/27"},
+			localRoutes:    []string{"192.168.0.0/24", "192.168.0.0/26"},
+			expectedVpn:    true,
+			expectedPrefix: netip.MustParsePrefix("192.168.0.0/27"),
+		},
+		{
+			name:           "Duplicate prefix in both",
+			addr:           "192.168.1.1",
+			vpnRoutes:      []string{"192.168.1.0/24"},
+			localRoutes:    []string{"192.168.1.0/24"},
+			expectedVpn:    false,
+			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			addr, err := netip.ParseAddr(tt.addr)
+			if err != nil {
+				t.Fatalf("Failed to parse address %s: %v", tt.addr, err)
+			}
+
+			var vpnRoutes, localRoutes []netip.Prefix
+			for _, route := range tt.vpnRoutes {
+				prefix, err := netip.ParsePrefix(route)
+				if err != nil {
+					t.Fatalf("Failed to parse VPN route %s: %v", route, err)
+				}
+				vpnRoutes = append(vpnRoutes, prefix)
+			}
+
+			for _, route := range tt.localRoutes {
+				prefix, err := netip.ParsePrefix(route)
+				if err != nil {
+					t.Fatalf("Failed to parse local route %s: %v", route, err)
+				}
+				localRoutes = append(localRoutes, prefix)
+			}
+
+			isVpn, matchedPrefix := isVpnRoute(addr, vpnRoutes, localRoutes)
+			assert.Equal(t, tt.expectedVpn, isVpn, "isVpnRoute should return expectedVpn value")
+			assert.Equal(t, tt.expectedPrefix, matchedPrefix, "isVpnRoute should return expectedVpn prefix")
+		})
+	}
+}

client/internal/routemanager/systemops/systemops_isvpnroute_test.go

@@ -1,132 +0,0 @@
-//go:build !android && !ios
-
-package systemops
-
-import (
-	"net/netip"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-)
-
-func TestIsVpnRoute(t *testing.T) {
-	tests := []struct {
-		name           string
-		addr           string
-		vpnRoutes      []string
-		localRoutes    []string
-		expectedVpn    bool
-		expectedPrefix netip.Prefix
-	}{
-		{
-			name:           "Match in VPN routes",
-			addr:           "192.168.1.1",
-			vpnRoutes:      []string{"192.168.1.0/24"},
-			localRoutes:    []string{"10.0.0.0/8"},
-			expectedVpn:    true,
-			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
-		},
-		{
-			name:           "Match in local routes",
-			addr:           "10.1.1.1",
-			vpnRoutes:      []string{"192.168.1.0/24"},
-			localRoutes:    []string{"10.0.0.0/8"},
-			expectedVpn:    false,
-			expectedPrefix: netip.MustParsePrefix("10.0.0.0/8"),
-		},
-		{
-			name:           "No match",
-			addr:           "172.16.0.1",
-			vpnRoutes:      []string{"192.168.1.0/24"},
-			localRoutes:    []string{"10.0.0.0/8"},
-			expectedVpn:    false,
-			expectedPrefix: netip.Prefix{},
-		},
-		{
-			name:           "Default route ignored",
-			addr:           "192.168.1.1",
-			vpnRoutes:      []string{"0.0.0.0/0", "192.168.1.0/24"},
-			localRoutes:    []string{"10.0.0.0/8"},
-			expectedVpn:    true,
-			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
-		},
-		{
-			name:           "Default route matches but ignored",
-			addr:           "172.16.1.1",
-			vpnRoutes:      []string{"0.0.0.0/0", "192.168.1.0/24"},
-			localRoutes:    []string{"10.0.0.0/8"},
-			expectedVpn:    false,
-			expectedPrefix: netip.Prefix{},
-		},
-		{
-			name:           "Longest prefix match local",
-			addr:           "192.168.1.1",
-			vpnRoutes:      []string{"192.168.0.0/16"},
-			localRoutes:    []string{"192.168.1.0/24"},
-			expectedVpn:    false,
-			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
-		},
-		{
-			name:           "Longest prefix match local multiple",
-			addr:           "192.168.0.1",
-			vpnRoutes:      []string{"192.168.0.0/16", "192.168.0.0/25", "192.168.0.0/27"},
-			localRoutes:    []string{"192.168.0.0/24", "192.168.0.0/26", "192.168.0.0/28"},
-			expectedVpn:    false,
-			expectedPrefix: netip.MustParsePrefix("192.168.0.0/28"),
-		},
-		{
-			name:           "Longest prefix match vpn",
-			addr:           "192.168.1.1",
-			vpnRoutes:      []string{"192.168.1.0/24"},
-			localRoutes:    []string{"192.168.0.0/16"},
-			expectedVpn:    true,
-			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
-		},
-		{
-			name:           "Longest prefix match vpn multiple",
-			addr:           "192.168.0.1",
-			vpnRoutes:      []string{"192.168.0.0/16", "192.168.0.0/25", "192.168.0.0/27"},
-			localRoutes:    []string{"192.168.0.0/24", "192.168.0.0/26"},
-			expectedVpn:    true,
-			expectedPrefix: netip.MustParsePrefix("192.168.0.0/27"),
-		},
-		{
-			name:           "Duplicate prefix in both",
-			addr:           "192.168.1.1",
-			vpnRoutes:      []string{"192.168.1.0/24"},
-			localRoutes:    []string{"192.168.1.0/24"},
-			expectedVpn:    false,
-			expectedPrefix: netip.MustParsePrefix("192.168.1.0/24"),
-		},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			addr, err := netip.ParseAddr(tt.addr)
-			if err != nil {
-				t.Fatalf("Failed to parse address %s: %v", tt.addr, err)
-			}
-
-			var vpnRoutes, localRoutes []netip.Prefix
-			for _, route := range tt.vpnRoutes {
-				prefix, err := netip.ParsePrefix(route)
-				if err != nil {
-					t.Fatalf("Failed to parse VPN route %s: %v", route, err)
-				}
-				vpnRoutes = append(vpnRoutes, prefix)
-			}
-
-			for _, route := range tt.localRoutes {
-				prefix, err := netip.ParsePrefix(route)
-				if err != nil {
-					t.Fatalf("Failed to parse local route %s: %v", route, err)
-				}
-				localRoutes = append(localRoutes, prefix)
-			}
-
-			isVpn, matchedPrefix := isVpnRoute(addr, vpnRoutes, localRoutes)
-			assert.Equal(t, tt.expectedVpn, isVpn, "isVpnRoute should return expectedVpn value")
-			assert.Equal(t, tt.expectedPrefix, matchedPrefix, "isVpnRoute should return expectedVpn prefix")
-		})
-	}
-}

client/internal/routemanager/systemops/systemops_linux_test.go

@@ -1,10 +1,13 @@
-//go:build linux && !android && privileged
+//go:build !android
 
 package systemops
 
 import (
 	"errors"
+	"fmt"
 	"net"
+	"os"
+	"strings"
 	"syscall"
 	"testing"
 
@@ -15,6 +18,10 @@ import (
 	"github.com/netbirdio/netbird/client/internal/routemanager/vars"
 )
 
+var expectedVPNint = "wgtest0"
+var expectedExternalInt = "dummyext0"
+var expectedInternalInt = "dummyint0"
+
 func init() {
 	testCases = append(testCases, []testCase{
 		{
@@ -26,6 +33,62 @@ func init() {
 	}...)
 }
 
+func TestEntryExists(t *testing.T) {
+	tempDir := t.TempDir()
+	tempFilePath := fmt.Sprintf("%s/rt_tables", tempDir)
+
+	content := []string{
+		"1000 reserved",
+		fmt.Sprintf("%d %s", NetbirdVPNTableID, NetbirdVPNTableName),
+		"9999 other_table",
+	}
+	require.NoError(t, os.WriteFile(tempFilePath, []byte(strings.Join(content, "\n")), 0644))
+
+	file, err := os.Open(tempFilePath)
+	require.NoError(t, err)
+	defer func() {
+		assert.NoError(t, file.Close())
+	}()
+
+	tests := []struct {
+		name        string
+		id          int
+		shouldExist bool
+		err         error
+	}{
+		{
+			name:        "ExistsWithNetbirdPrefix",
+			id:          7120,
+			shouldExist: true,
+			err:         nil,
+		},
+		{
+			name:        "ExistsWithDifferentName",
+			id:          1000,
+			shouldExist: true,
+			err:         ErrTableIDExists,
+		},
+		{
+			name:        "DoesNotExist",
+			id:          1234,
+			shouldExist: false,
+			err:         nil,
+		},
+	}
+
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			exists, err := entryExists(file, tc.id)
+			if tc.err != nil {
+				assert.ErrorIs(t, err, tc.err)
+			} else {
+				assert.NoError(t, err)
+			}
+			assert.Equal(t, tc.shouldExist, exists)
+		})
+	}
+}
+
 func createAndSetupDummyInterface(t *testing.T, interfaceName, ipAddressCIDR string) string {
 	t.Helper()
 

client/internal/routemanager/systemops/systemops_routing_data_linux_test.go

@@ -1,15 +0,0 @@
-//go:build linux && !android
-
-package systemops
-
-// Interface names used by the shared routing test fixtures. Kept untagged (no
-// privileged build tag) so the non-privileged test files in this package compile.
-//
-//nolint:unused // consumed by the privileged-tagged routing tests
-var expectedVPNint = "wgtest0"
-
-//nolint:unused // consumed by the privileged-tagged routing tests
-var expectedExternalInt = "dummyext0"
-
-//nolint:unused // consumed by the privileged-tagged routing tests
-var expectedInternalInt = "dummyint0"

client/internal/routemanager/systemops/systemops_routing_data_test.go

@@ -1,83 +0,0 @@
-//go:build (linux && !android) || (darwin && !ios) || freebsd || openbsd || netbsd || dragonfly
-
-package systemops
-
-import (
-	"net"
-
-	nbnet "github.com/netbirdio/netbird/client/net"
-)
-
-// Shared, non-privileged routing test fixtures. The privileged TestRouting (and its
-// per-platform init() appenders) consume these; they live here so the unprivileged
-// BSD/darwin test files compile without the privileged build tag.
-
-type PacketExpectation struct {
-	SrcIP   net.IP
-	DstIP   net.IP
-	SrcPort int
-	DstPort int
-	UDP     bool
-	TCP     bool
-}
-
-//nolint:unused // consumed by the privileged-tagged routing tests
-type testCase struct {
-	name              string
-	expectedInterface string
-	dialer            dialer
-	expectedPacket    PacketExpectation
-}
-
-//nolint:unused // consumed by the privileged-tagged routing tests
-var testCases = []testCase{
-	{
-		name:              "To external host without custom dialer via vpn",
-		expectedInterface: expectedVPNint,
-		dialer:            &net.Dialer{},
-		expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "192.0.2.1", 53),
-	},
-	{
-		name:              "To external host with custom dialer via physical interface",
-		expectedInterface: expectedExternalInt,
-		dialer:            nbnet.NewDialer(),
-		expectedPacket:    createPacketExpectation("192.168.0.1", 12345, "192.0.2.1", 53),
-	},
-
-	{
-		name:              "To duplicate internal route with custom dialer via physical interface",
-		expectedInterface: expectedInternalInt,
-		dialer:            nbnet.NewDialer(),
-		expectedPacket:    createPacketExpectation("192.168.1.1", 12345, "10.0.0.2", 53),
-	},
-	{
-		name:              "To duplicate internal route without custom dialer via physical interface", // local route takes precedence
-		expectedInterface: expectedInternalInt,
-		dialer:            &net.Dialer{},
-		expectedPacket:    createPacketExpectation("192.168.1.1", 12345, "10.0.0.2", 53),
-	},
-
-	{
-		name:              "To unique vpn route with custom dialer via physical interface",
-		expectedInterface: expectedExternalInt,
-		dialer:            nbnet.NewDialer(),
-		expectedPacket:    createPacketExpectation("192.168.0.1", 12345, "172.16.0.2", 53),
-	},
-	{
-		name:              "To unique vpn route without custom dialer via vpn",
-		expectedInterface: expectedVPNint,
-		dialer:            &net.Dialer{},
-		expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "172.16.0.2", 53),
-	},
-}
-
-//nolint:unused // consumed by the privileged-tagged routing tests
-func createPacketExpectation(srcIP string, srcPort int, dstIP string, dstPort int) PacketExpectation {
-	return PacketExpectation{
-		SrcIP:   net.ParseIP(srcIP),
-		DstIP:   net.ParseIP(dstIP),
-		SrcPort: srcPort,
-		DstPort: dstPort,
-		UDP:     true,
-	}
-}

client/internal/routemanager/systemops/systemops_unix_test.go

@@ -1,4 +1,4 @@
-//go:build ((linux && !android) || (darwin && !ios) || freebsd || openbsd || netbsd || dragonfly) && privileged
+//go:build (linux && !android) || (darwin && !ios) || freebsd || openbsd || netbsd || dragonfly
 
 package systemops
 
@@ -20,6 +20,63 @@ import (
 	nbnet "github.com/netbirdio/netbird/client/net"
 )
 
+type PacketExpectation struct {
+	SrcIP   net.IP
+	DstIP   net.IP
+	SrcPort int
+	DstPort int
+	UDP     bool
+	TCP     bool
+}
+
+type testCase struct {
+	name              string
+	expectedInterface string
+	dialer            dialer
+	expectedPacket    PacketExpectation
+}
+
+var testCases = []testCase{
+	{
+		name:              "To external host without custom dialer via vpn",
+		expectedInterface: expectedVPNint,
+		dialer:            &net.Dialer{},
+		expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "192.0.2.1", 53),
+	},
+	{
+		name:              "To external host with custom dialer via physical interface",
+		expectedInterface: expectedExternalInt,
+		dialer:            nbnet.NewDialer(),
+		expectedPacket:    createPacketExpectation("192.168.0.1", 12345, "192.0.2.1", 53),
+	},
+
+	{
+		name:              "To duplicate internal route with custom dialer via physical interface",
+		expectedInterface: expectedInternalInt,
+		dialer:            nbnet.NewDialer(),
+		expectedPacket:    createPacketExpectation("192.168.1.1", 12345, "10.0.0.2", 53),
+	},
+	{
+		name:              "To duplicate internal route without custom dialer via physical interface", // local route takes precedence
+		expectedInterface: expectedInternalInt,
+		dialer:            &net.Dialer{},
+		expectedPacket:    createPacketExpectation("192.168.1.1", 12345, "10.0.0.2", 53),
+	},
+
+	{
+		name:              "To unique vpn route with custom dialer via physical interface",
+		expectedInterface: expectedExternalInt,
+		dialer:            nbnet.NewDialer(),
+		expectedPacket:    createPacketExpectation("192.168.0.1", 12345, "172.16.0.2", 53),
+	},
+	{
+		name:              "To unique vpn route without custom dialer via vpn",
+		expectedInterface: expectedVPNint,
+		dialer:            &net.Dialer{},
+		expectedPacket:    createPacketExpectation("100.64.0.1", 12345, "172.16.0.2", 53),
+	},
+}
+
 func TestRouting(t *testing.T) {
 	nbnet.Init()
 	for _, tc := range testCases {
@@ -45,6 +102,16 @@ func TestRouting(t *testing.T) {
 	}
 }
 
+func createPacketExpectation(srcIP string, srcPort int, dstIP string, dstPort int) PacketExpectation {
+	return PacketExpectation{
+		SrcIP:   net.ParseIP(srcIP),
+		DstIP:   net.ParseIP(dstIP),
+		SrcPort: srcPort,
+		DstPort: dstPort,
+		UDP:     true,
+	}
+}
+
 func startPacketCapture(t *testing.T, intf, filter string) *pcap.Handle {
 	t.Helper()
 

client/internal/routemanager/systemops/systemops_windows_test.go

@@ -1,5 +1,3 @@
-//go:build windows && privileged
-
 package systemops
 
 import (

client/internal/routemanager/systemops/v6route_bsd_test.go

@@ -11,8 +11,6 @@ import (
 // ensureIPv6DefaultRoute installs an IPv6 default route via the loopback
 // interface so route lookups for global IPv6 prefixes resolve in environments
 // without v6 connectivity. If a default already exists it is left alone.
-//
-//nolint:unused // consumed by the privileged-tagged routing tests
 func ensureIPv6DefaultRoute(t *testing.T) {
 	t.Helper()