use API

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

.github/workflows/release.yml

@@ -9,10 +9,13 @@ on:
   pull_request:
 
 env:
-  SIGN_PIPE_VER: "v0.1.5"
-  GORELEASER_VER: "v2.14.3"
+  SIGN_PIPE_VER: "v0.1.6"
+  GORELEASER_VER: "v2.16.0"
   PRODUCT_NAME: "NetBird"
   COPYRIGHT: "NetBird GmbH"
+  flags: ""
+  SKIP_PUBLISH: "true"
+  SKIP_DOCKER_PUSH: "false"
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}-${{ github.head_ref || github.actor_id }}
@@ -130,8 +133,6 @@ jobs:
       windows_packages_artifact_url: ${{ steps.upload_windows_packages.outputs.artifact-url }}
       macos_packages_artifact_url: ${{ steps.upload_macos_packages.outputs.artifact-url }}
       ghcr_images: ${{ steps.tag_and_push_images.outputs.images_markdown }}
-    env:
-      flags: ""
     steps:
       - name: Checkout
         uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
@@ -143,8 +144,27 @@ jobs:
         id: semver_parser
         uses: netbirdio/shared-actions/actions/parse-semver@be5df6047383da2236e02243cceb857d8567c27e # v0.0.2
 
-      - if: ${{ !startsWith(github.ref, 'refs/tags/v') }}
-        run: echo "flags=--snapshot" >> $GITHUB_ENV
+      - name: Set snapshot flag
+        if: ${{ !startsWith(github.ref, 'refs/tags/v') }}
+        run: | 
+          echo "flags=--snapshot" >> $GITHUB_ENV
+
+      - name: Set build vars
+        if: ${{ startsWith(github.ref, 'refs/tags/v') }}
+        run: |
+          if [[ "x-${{ steps.semver_parser.outputs.prerelease }}" == "x-" && "x-${{ github.repository }}" == "x-netbirdio/netbird" ]]; then
+            echo "x-${{ github.repository }}" 
+            echo "x-${{ steps.semver_parser.outputs.prerelease }}"
+            echo "SKIP_PUBLISH=false" >> $GITHUB_ENV
+          else
+            echo "x-${{ github.repository }}" 
+            echo "x-${{ steps.semver_parser.outputs.prerelease }}"
+          fi
+          
+          if [[ "x-${{ github.repository }}" != "x-netbirdio/netbird" ]]; then
+            echo "SKIP_DOCKER_PUSH=true" >> $GITHUB_ENV
+          fi
+
       - name: Set up Go
         uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
         with:
@@ -212,6 +232,8 @@ jobs:
           UPLOAD_YUM_SECRET: ${{ secrets.PKG_UPLOAD_SECRET }}
           GPG_RPM_KEY_FILE: ${{ env.GPG_RPM_KEY_FILE }}
           NFPM_NETBIRD_RPM_PASSPHRASE: ${{ secrets.GPG_RPM_PASSPHRASE }}
+          SKIP_PUBLISH: ${{ env.SKIP_PUBLISH }}
+          SKIP_DOCKER_PUSH: ${{ env.SKIP_DOCKER_PUSH }}
       - name: Verify RPM signatures
         run: |
           docker run --rm -v $(pwd)/dist:/dist fedora:41 bash -c '
@@ -334,8 +356,22 @@ jobs:
         id: semver_parser
         uses: netbirdio/shared-actions/actions/parse-semver@be5df6047383da2236e02243cceb857d8567c27e # v0.0.2
 
-      - if: ${{ !startsWith(github.ref, 'refs/tags/v') }}
-        run: echo "flags=--snapshot" >> $GITHUB_ENV
+      - name: Set snapshot flag
+        if: ${{ !startsWith(github.ref, 'refs/tags/v') }}
+        run: |
+          echo "flags=--snapshot" >> $GITHUB_ENV
+
+      - name: Set build vars
+        if: ${{ startsWith(github.ref, 'refs/tags/v') }}
+        run: |
+          if [[ "x-${{ steps.semver_parser.outputs.prerelease }}" == "x-" && "x-${{ github.repository }}" == "x-netbirdio/netbird" ]]; then
+            echo "x-${{ github.repository }}" 
+            echo "x-${{ steps.semver_parser.outputs.prerelease }}"
+            echo "SKIP_PUBLISH=false" >> $GITHUB_ENV
+          else
+            echo "x-${{ github.repository }}" 
+            echo "x-${{ steps.semver_parser.outputs.prerelease }}"
+          fi
 
       - name: Set up Go
         uses: actions/setup-go@4b73464bb391d4059bd26b0524d20df3927bd417 # v6.3.0
@@ -395,6 +431,7 @@ jobs:
           UPLOAD_YUM_SECRET: ${{ secrets.PKG_UPLOAD_SECRET }}
           GPG_RPM_KEY_FILE: ${{ env.GPG_RPM_KEY_FILE }}
           NFPM_NETBIRD_UI_RPM_PASSPHRASE: ${{ secrets.GPG_RPM_PASSPHRASE }}
+          SKIP_PUBLISH: ${{ env.SKIP_PUBLISH }}
       - name: Verify RPM signatures
         run: |
           docker run --rm -v $(pwd)/dist:/dist fedora:41 bash -c '

.goreleaser.yaml

@@ -1,5 +1,7 @@
 version: 2
-
+env:
+  - SKIP_PUBLISH={{ if index .Env "SKIP_PUBLISH" }}{{ .Env.SKIP_PUBLISH }}{{ else }}true{{ end }}
+  - SKIP_DOCKER_PUSH={{ if index .Env "SKIP_DOCKER_PUSH" }}{{ .Env.SKIP_DOCKER_PUSH }}{{ else }}false{{ end }}
 project_name: netbird
 builds:
   - id: netbird-wasm
@@ -74,6 +76,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -88,6 +92,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -102,6 +108,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -122,6 +130,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -136,6 +146,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -150,6 +162,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X main.Version={{.Version}} -X main.Commit={{.Commit}} -X main.BuildDate={{.CommitDate}}
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -170,6 +184,8 @@ builds:
       - amd64
       - arm64
       - arm
+    goarm:
+      - 7
     ldflags:
       - -s -w -X github.com/netbirdio/netbird/version.version={{.Version}} -X main.commit={{.Commit}} -X main.date={{.CommitDate}} -X main.builtBy=goreleaser
     mod_timestamp: "{{ .CommitTimestamp }}"
@@ -222,670 +238,192 @@ nfpms:
     rpm:
       signature:
         key_file: '{{ if index .Env "GPG_RPM_KEY_FILE" }}{{ .Env.GPG_RPM_KEY_FILE }}{{ end }}'
-dockers:
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-amd64
-    ids:
-      - netbird
-    goarch: amd64
-    use: buildx
-    dockerfile: client/Dockerfile
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm64v8
-    ids:
-      - netbird
-    goarch: arm64
-    use: buildx
-    dockerfile: client/Dockerfile
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm
-    ids:
-      - netbird
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: client/Dockerfile
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-rootless-amd64
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-amd64
-    ids:
-      - netbird
-    goarch: amd64
-    use: buildx
-    dockerfile: client/Dockerfile-rootless
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-    ids:
-      - netbird
-    goarch: arm64
-    use: buildx
-    dockerfile: client/Dockerfile-rootless
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird:{{ .Version }}-rootless-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm
-    ids:
-      - netbird
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: client/Dockerfile-rootless
-    extra_files:
-      - client/netbird-entrypoint.sh
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-
-  - image_templates:
-      - netbirdio/relay:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/relay:{{ .Version }}-amd64
-    ids:
-      - netbird-relay
-    goarch: amd64
-    use: buildx
-    dockerfile: relay/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/relay:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm64v8
-    ids:
-      - netbird-relay
-    goarch: arm64
-    use: buildx
-    dockerfile: relay/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/relay:{{ .Version }}-arm
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm
-    ids:
-      - netbird-relay
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: relay/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/signal:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/signal:{{ .Version }}-amd64
-    ids:
-      - netbird-signal
-    goarch: amd64
-    use: buildx
-    dockerfile: signal/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/signal:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm64v8
-    ids:
-      - netbird-signal
-    goarch: arm64
-    use: buildx
-    dockerfile: signal/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/signal:{{ .Version }}-arm
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm
-    ids:
-      - netbird-signal
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: signal/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/management:{{ .Version }}-amd64
-    ids:
-      - netbird-mgmt
-    goarch: amd64
-    use: buildx
-    dockerfile: management/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm64v8
-    ids:
-      - netbird-mgmt
-    goarch: arm64
-    use: buildx
-    dockerfile: management/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-arm
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm
-    ids:
-      - netbird-mgmt
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: management/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-debug-amd64
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-amd64
-    ids:
-      - netbird-mgmt
-    goarch: amd64
-    use: buildx
-    dockerfile: management/Dockerfile.debug
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-debug-arm64v8
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-arm64v8
-    ids:
-      - netbird-mgmt
-    goarch: arm64
-    use: buildx
-    dockerfile: management/Dockerfile.debug
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-
-  - image_templates:
-      - netbirdio/management:{{ .Version }}-debug-arm
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-arm
-    ids:
-      - netbird-mgmt
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: management/Dockerfile.debug
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/upload:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/upload:{{ .Version }}-amd64
-    ids:
-      - netbird-upload
-    goarch: amd64
-    use: buildx
-    dockerfile: upload-server/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/upload:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm64v8
-    ids:
-      - netbird-upload
-    goarch: arm64
-    use: buildx
-    dockerfile: upload-server/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/upload:{{ .Version }}-arm
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm
-    ids:
-      - netbird-upload
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: upload-server/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird-server:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-amd64
-    ids:
-      - netbird-server
-    goarch: amd64
-    use: buildx
-    dockerfile: combined/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird-server:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm64v8
-    ids:
-      - netbird-server
-    goarch: arm64
-    use: buildx
-    dockerfile: combined/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/netbird-server:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm
-    ids:
-      - netbird-server
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: combined/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/reverse-proxy:{{ .Version }}-amd64
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-amd64
-    ids:
-      - netbird-proxy
-    goarch: amd64
-    use: buildx
-    dockerfile: proxy/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/amd64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-    ids:
-      - netbird-proxy
-    goarch: arm64
-    use: buildx
-    dockerfile: proxy/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm64"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-  - image_templates:
-      - netbirdio/reverse-proxy:{{ .Version }}-arm
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm
-    ids:
-      - netbird-proxy
-    goarch: arm
-    goarm: 6
-    use: buildx
-    dockerfile: proxy/Dockerfile
-    build_flag_templates:
-      - "--platform=linux/arm"
-      - "--label=org.opencontainers.image.created={{.Date}}"
-      - "--label=org.opencontainers.image.title={{.ProjectName}}"
-      - "--label=org.opencontainers.image.version={{.Version}}"
-      - "--label=org.opencontainers.image.revision={{.FullCommit}}"
-      - "--label=org.opencontainers.image.source=https://github.com/netbirdio/{{.ProjectName}}"
-      - "--label=maintainer=dev@netbird.io"
-docker_manifests:
-  - name_template: netbirdio/netbird:{{ .Version }}
-    image_templates:
-      - netbirdio/netbird:{{ .Version }}-arm64v8
-      - netbirdio/netbird:{{ .Version }}-arm
-      - netbirdio/netbird:{{ .Version }}-amd64
-
-  - name_template: netbirdio/netbird:latest
-    image_templates:
-      - netbirdio/netbird:{{ .Version }}-arm64v8
-      - netbirdio/netbird:{{ .Version }}-arm
-      - netbirdio/netbird:{{ .Version }}-amd64
-
-  - name_template: netbirdio/netbird:{{ .Version }}-rootless
-    image_templates:
-      - netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-      - netbirdio/netbird:{{ .Version }}-rootless-arm
-      - netbirdio/netbird:{{ .Version }}-rootless-amd64
-
-  - name_template: netbirdio/netbird:rootless-latest
-    image_templates:
-      - netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-      - netbirdio/netbird:{{ .Version }}-rootless-arm
-      - netbirdio/netbird:{{ .Version }}-rootless-amd64
-
-  - name_template: netbirdio/relay:{{ .Version }}
-    image_templates:
-      - netbirdio/relay:{{ .Version }}-arm64v8
-      - netbirdio/relay:{{ .Version }}-arm
-      - netbirdio/relay:{{ .Version }}-amd64
-
-  - name_template: netbirdio/relay:latest
-    image_templates:
-      - netbirdio/relay:{{ .Version }}-arm64v8
-      - netbirdio/relay:{{ .Version }}-arm
-      - netbirdio/relay:{{ .Version }}-amd64
-
-  - name_template: netbirdio/signal:{{ .Version }}
-    image_templates:
-      - netbirdio/signal:{{ .Version }}-arm64v8
-      - netbirdio/signal:{{ .Version }}-arm
-      - netbirdio/signal:{{ .Version }}-amd64
-
-  - name_template: netbirdio/signal:latest
-    image_templates:
-      - netbirdio/signal:{{ .Version }}-arm64v8
-      - netbirdio/signal:{{ .Version }}-arm
-      - netbirdio/signal:{{ .Version }}-amd64
-
-  - name_template: netbirdio/management:{{ .Version }}
-    image_templates:
-      - netbirdio/management:{{ .Version }}-arm64v8
-      - netbirdio/management:{{ .Version }}-arm
-      - netbirdio/management:{{ .Version }}-amd64
-
-  - name_template: netbirdio/management:latest
-    image_templates:
-      - netbirdio/management:{{ .Version }}-arm64v8
-      - netbirdio/management:{{ .Version }}-arm
-      - netbirdio/management:{{ .Version }}-amd64
-
-  - name_template: netbirdio/management:debug-latest
-    image_templates:
-      - netbirdio/management:{{ .Version }}-debug-arm64v8
-      - netbirdio/management:{{ .Version }}-debug-arm
-      - netbirdio/management:{{ .Version }}-debug-amd64
-  - name_template: netbirdio/upload:{{ .Version }}
-    image_templates:
-      - netbirdio/upload:{{ .Version }}-arm64v8
-      - netbirdio/upload:{{ .Version }}-arm
-      - netbirdio/upload:{{ .Version }}-amd64
-
-  - name_template: netbirdio/upload:latest
-    image_templates:
-      - netbirdio/upload:{{ .Version }}-arm64v8
-      - netbirdio/upload:{{ .Version }}-arm
-      - netbirdio/upload:{{ .Version }}-amd64
-
-  - name_template: netbirdio/netbird-server:{{ .Version }}
-    image_templates:
-      - netbirdio/netbird-server:{{ .Version }}-arm64v8
-      - netbirdio/netbird-server:{{ .Version }}-arm
-      - netbirdio/netbird-server:{{ .Version }}-amd64
-
-  - name_template: netbirdio/netbird-server:latest
-    image_templates:
-      - netbirdio/netbird-server:{{ .Version }}-arm64v8
-      - netbirdio/netbird-server:{{ .Version }}-arm
-      - netbirdio/netbird-server:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird:latest
-    image_templates:
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird:{{ .Version }}-rootless
-    image_templates:
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird:rootless-latest
-    image_templates:
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm64v8
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-arm
-      - ghcr.io/netbirdio/netbird:{{ .Version }}-rootless-amd64
-
-  - name_template: ghcr.io/netbirdio/relay:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm
-      - ghcr.io/netbirdio/relay:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/relay:latest
-    image_templates:
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/relay:{{ .Version }}-arm
-      - ghcr.io/netbirdio/relay:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/signal:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm
-      - ghcr.io/netbirdio/signal:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/signal:latest
-    image_templates:
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/signal:{{ .Version }}-arm
-      - ghcr.io/netbirdio/signal:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/management:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm
-      - ghcr.io/netbirdio/management:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/management:latest
-    image_templates:
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/management:{{ .Version }}-arm
-      - ghcr.io/netbirdio/management:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/management:debug-latest
-    image_templates:
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-arm64v8
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-arm
-      - ghcr.io/netbirdio/management:{{ .Version }}-debug-amd64
-
-  - name_template: ghcr.io/netbirdio/upload:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm
-      - ghcr.io/netbirdio/upload:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/upload:latest
-    image_templates:
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/upload:{{ .Version }}-arm
-      - ghcr.io/netbirdio/upload:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird-server:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/netbird-server:latest
-    image_templates:
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-arm
-      - ghcr.io/netbirdio/netbird-server:{{ .Version }}-amd64
-
-  - name_template: netbirdio/reverse-proxy:{{ .Version }}
-    image_templates:
-      - netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-      - netbirdio/reverse-proxy:{{ .Version }}-arm
-      - netbirdio/reverse-proxy:{{ .Version }}-amd64
-
-  - name_template: netbirdio/reverse-proxy:latest
-    image_templates:
-      - netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-      - netbirdio/reverse-proxy:{{ .Version }}-arm
-      - netbirdio/reverse-proxy:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/reverse-proxy:{{ .Version }}
-    image_templates:
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-amd64
-
-  - name_template: ghcr.io/netbirdio/reverse-proxy:latest
-    image_templates:
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm64v8
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-arm
-      - ghcr.io/netbirdio/reverse-proxy:{{ .Version }}-amd64
+dockers_v2:
+   - id: netbird
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird
+     images:
+       - netbirdio/netbird
+       - ghcr.io/netbirdio/netbird
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: client/Dockerfile
+     extra_files:
+       - client/netbird-entrypoint.sh
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm/6
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: netbird-rootless
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird
+     images:
+       - netbirdio/netbird
+       - ghcr.io/netbirdio/netbird
+     tags:
+       - "v{{ .Version }}-rootless"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: client/Dockerfile-rootless
+     extra_files:
+       - client/netbird-entrypoint.sh
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm/6
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: relay
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-relay
+     images:
+       - netbirdio/relay
+       - ghcr.io/netbirdio/relay
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: relay/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: signal
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-signal
+     images:
+       - netbirdio/signal
+       - ghcr.io/netbirdio/signal
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: signal/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: management
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-mgmt
+     images:
+       - netbirdio/management
+       - ghcr.io/netbirdio/management
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: management/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: upload
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-upload
+     images:
+       - netbirdio/upload
+       - ghcr.io/netbirdio/upload
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: upload-server/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: netbird-server
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-server
+     images:
+       - netbirdio/netbird-server
+       - ghcr.io/netbirdio/netbird-server
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: combined/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
+   - id: netbird-proxy
+     disable: "{{ .Env.SKIP_DOCKER_PUSH }}"
+     ids:
+       - netbird-proxy
+     images:
+       - netbirdio/reverse-proxy
+       - ghcr.io/netbirdio/reverse-proxy
+     tags:
+       - "v{{ .Version }}"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
+     dockerfile: proxy/Dockerfile
+     platforms:
+       - linux/amd64
+       - linux/arm64
+       - linux/arm
+     annotations:
+       "org.opencontainers.image.created": "{{.Date}}"
+       "org.opencontainers.image.title": "{{.ProjectName}}"
+       "org.opencontainers.image.version": "{{.Version}}"
+       "org.opencontainers.image.revision": "{{.FullCommit}}"
+       "org.opencontainers.image.source": "{{.GitURL}}"
+       "maintainer": "dev@netbird.io"
 
 brews:
   - ids:
       - default
+    skip_upload: "{{ .Env.SKIP_PUBLISH }}"
     repository:
       owner: netbirdio
       name: homebrew-tap
@@ -902,6 +440,7 @@ brews:
 
 uploads:
   - name: debian
+    skip: "{{ .Env.SKIP_PUBLISH }}"
     ids:
       - netbird_deb
     mode: archive
@@ -910,6 +449,7 @@ uploads:
     method: PUT
 
   - name: yum
+    skip: "{{ .Env.SKIP_PUBLISH }}"
     ids:
       - netbird_rpm
     mode: archive

.goreleaser_ui.yaml

@@ -1,5 +1,6 @@
 version: 2
-
+env:
+  - SKIP_PUBLISH={{ if index .Env "SKIP_PUBLISH" }}{{ .Env.SKIP_PUBLISH }}{{ else }}true{{ end }}
 project_name: netbird-ui
 builds:
   - id: netbird-ui
@@ -101,6 +102,7 @@ nfpms:
 
 uploads:
   - name: debian
+    skip: "{{ .Env.SKIP_PUBLISH }}"
     ids:
       - netbird_ui_deb
     mode: archive
@@ -109,6 +111,7 @@ uploads:
     method: PUT
 
   - name: yum
+    skip: "{{ .Env.SKIP_PUBLISH }}"
     ids:
       - netbird_ui_rpm
     mode: archive

client/Dockerfile

@@ -4,7 +4,7 @@
 #   sudo podman build -t localhost/netbird:latest -f client/Dockerfile --ignorefile .dockerignore-client .
 #   sudo podman run --rm -it --cap-add={BPF,NET_ADMIN,NET_RAW} localhost/netbird:latest
 
-FROM alpine:3.23.3
+FROM alpine:3.24
 # iproute2: busybox doesn't display ip rules properly
 RUN apk add --no-cache \
     bash \
@@ -21,7 +21,7 @@ ENV \
     NB_ENTRYPOINT_SERVICE_TIMEOUT="30"
 
 ENTRYPOINT [ "/usr/local/bin/netbird-entrypoint.sh" ]
-
-ARG NETBIRD_BINARY=netbird
+ARG TARGETPLATFORM
+ARG NETBIRD_BINARY=$TARGETPLATFORM/netbird
 COPY client/netbird-entrypoint.sh /usr/local/bin/netbird-entrypoint.sh
 COPY "${NETBIRD_BINARY}"  /usr/local/bin/netbird

client/Dockerfile-rootless

@@ -4,7 +4,7 @@
 #   podman build -t localhost/netbird:latest -f client/Dockerfile --ignorefile .dockerignore-client .
 #   podman run --rm -it --cap-add={BPF,NET_ADMIN,NET_RAW} localhost/netbird:latest
 
-FROM alpine:3.22.0
+FROM alpine:3.24
 
 RUN apk add --no-cache \
       bash \
@@ -27,7 +27,7 @@ ENV \
     NB_ENTRYPOINT_SERVICE_TIMEOUT="30"
 
 ENTRYPOINT [ "/usr/local/bin/netbird-entrypoint.sh" ]
-
-ARG NETBIRD_BINARY=netbird
+ARG TARGETPLATFORM
+ARG NETBIRD_BINARY=$TARGETPLATFORM/netbird
 COPY client/netbird-entrypoint.sh /usr/local/bin/netbird-entrypoint.sh
 COPY "${NETBIRD_BINARY}"  /usr/local/bin/netbird

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/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/up.go

@@ -201,10 +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

@@ -264,32 +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():
-		// 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.
+	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/internal/connect.go

@@ -18,6 +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"
@@ -48,13 +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
-	config         *profilemanager.Config
 	statusRecorder *peer.Status
 
 	engine        *Engine
@@ -63,35 +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 {
-	return &ConnectClient{
+	c := &ConnectClient{
 		ctx:            ctx,
-		config:         config,
 		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,
@@ -110,10 +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,
@@ -131,10 +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 {
+// 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
@@ -198,18 +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
 	}
@@ -243,13 +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)
@@ -257,8 +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))
 		}
@@ -275,7 +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)
@@ -284,13 +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.Stop()
+				// 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)
@@ -344,7 +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)
@@ -388,7 +430,7 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		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)
 		}
@@ -396,12 +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()
@@ -410,14 +453,12 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 		c.engine = nil
 		c.engineMutex.Unlock()
 
-		// todo: consider to remove this condition. Is not thread safe.
-		// We should always call Stop(), but we need to verify that it is idempotent
-		if engine.wgInterface != nil {
-			log.Infof("ensuring %s is removed, Netbird engine context cancelled", engine.wgInterface.Name())
-
-			if err := engine.Stop(); err != nil {
-				log.Errorf("Failed to stop engine: %v", err)
-			}
+		// 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)
 		}
 		c.statusRecorder.ClientTeardown()
 
@@ -437,8 +478,9 @@ func (c *ConnectClient) run(mobileDependency MobileDependency, runningChan chan
 	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.Stop()
 		}
 		return err
 	}
@@ -459,6 +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
@@ -515,14 +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 {
-	engine := c.Engine()
-	if engine != nil {
-		if err := engine.Stop(); err != nil {
-			return fmt.Errorf("stop engine: %w", err)
-		}
-	}
-	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/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"
@@ -1125,6 +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() {
@@ -1151,9 +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
 		}
@@ -1235,9 +1251,9 @@ func (e *Engine) receiveManagementEvents() {
 
 		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
 		}
@@ -1714,6 +1730,13 @@ func (e *Engine) receiveSignalEvents() {
 				return e.ctx.Err()
 			}
 
+			// Self-addressed heartbeat: the signal client's receive watchdog
+			// round-trips this through the server to confirm the receive stream
+			// is delivering. Liveness is already recorded before this handler.
+			if msg.GetBody().GetType() == sProto.Body_HEARTBEAT {
+				return nil
+			}
+
 			conn, ok := e.peerStore.PeerConn(msg.Key)
 			if !ok {
 				return fmt.Errorf("wrongly addressed message %s", msg.Key)
@@ -1754,9 +1777,9 @@ func (e *Engine) receiveSignalEvents() {
 			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
 		}

client/internal/peer/status.go

@@ -1024,14 +1024,17 @@ func (d *Status) GetRelayStates() []relay.ProbeResult {
 		return d.relayStates
 	}
 
-	// extend the list of stun, turn servers with relay address
+	// extend the list of stun, turn servers with the relay server connections
 	relayStates := slices.Clone(d.relayStates)
 
-	// if the server connection is not established then we will use the general address
-	// in case of connection we will use the instance specific address
-	instanceAddr, _, err := d.relayMgr.RelayInstanceAddress()
-	if err != nil {
-		// TODO add their status
+	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 := d.relayMgr.RelayConnectError(); connErr != nil {
+			err = connErr
+		}
 		for _, r := range d.relayMgr.ServerURLs() {
 			relayStates = append(relayStates, relay.ProbeResult{
 				URI: r,
@@ -1041,10 +1044,14 @@ func (d *Status) GetRelayStates() []relay.ProbeResult {
 		return relayStates
 	}
 
-	relayState := relay.ProbeResult{
-		URI: instanceAddr,
+	for _, rs := range states {
+		relayStates = append(relayStates, relay.ProbeResult{
+			URI:       rs.URL,
+			Err:       rs.Err,
+			Transport: rs.Transport,
+		})
 	}
-	return append(relayStates, relayState)
+	return relayStates
 }
 
 func (d *Status) ForwardingRules() []firewall.ForwardRule {
@@ -1405,6 +1412,7 @@ func (fs FullStatus) ToProto() *proto.FullStatus {
 		pbRelayState := &proto.RelayState{
 			URI:       relayState.URI,
 			Available: relayState.Err == nil,
+			Transport: relayState.Transport,
 		}
 		if err := relayState.Err; err != nil {
 			pbRelayState.Error = err.Error()

client/internal/profilemanager/service.go

@@ -575,8 +575,8 @@ func (s *ServiceManager) activeProfileID() (ID, bool) {
 }
 
 // ResolveProfile turns a user-supplied handle into a Profile. Resolution
-// precedence is: exact ID match, then unique ID prefix, then unique exact
-// name. Ambiguous matches return *ErrAmbiguousHandle so callers can
+// precedence is: exact ID match, then unique exact name, then unique ID
+// prefix. Ambiguous matches return *ErrAmbiguousHandle so callers can
 // surface the candidates.
 func (s *ServiceManager) ResolveProfile(handle, username string) (*Profile, error) {
 	if handle == "" {
@@ -594,6 +594,23 @@ func (s *ServiceManager) ResolveProfile(handle, username string) (*Profile, erro
 		}
 	}
 
+	var nameMatches []Profile
+	for i := range profiles {
+		if profiles[i].Name == handle {
+			nameMatches = append(nameMatches, profiles[i])
+		}
+	}
+	if len(nameMatches) == 1 {
+		return &nameMatches[0], nil
+	}
+	if len(nameMatches) > 1 {
+		return nil, &ErrAmbiguousHandle{
+			Handle:     handle,
+			Candidates: nameMatches,
+			Kind:       AmbiguityKindName,
+		}
+	}
+
 	// ID prefix match. Skip the default profile so `select d` does not
 	// accidentally pick it via prefix.
 	var prefixMatches []Profile
@@ -616,22 +633,5 @@ func (s *ServiceManager) ResolveProfile(handle, username string) (*Profile, erro
 		}
 	}
 
-	var nameMatches []Profile
-	for i := range profiles {
-		if profiles[i].Name == handle {
-			nameMatches = append(nameMatches, profiles[i])
-		}
-	}
-	if len(nameMatches) == 1 {
-		return &nameMatches[0], nil
-	}
-	if len(nameMatches) > 1 {
-		return nil, &ErrAmbiguousHandle{
-			Handle:     handle,
-			Candidates: nameMatches,
-			Kind:       AmbiguityKindName,
-		}
-	}
-
 	return nil, ErrProfileNotFound
 }

client/internal/relay/relay.go

@@ -32,6 +32,9 @@ type ProbeResult struct {
 	URI  string
 	Err  error
 	Addr string
+	// Transport is the negotiated relay transport, empty
+	// for stun/turn probes or when not connected.
+	Transport string
 }
 
 type StunTurnProbe struct {

client/internal/routemanager/manager.go

@@ -333,6 +333,8 @@ func (m *DefaultManager) Stop(stateManager *statemanager.Manager) {
 		}
 	}
 
+	m.notifier.Close()
+
 	m.mux.Lock()
 	defer m.mux.Unlock()
 	m.clientRoutes = nil

client/internal/routemanager/notifier/notifier_android.go

@@ -16,7 +16,7 @@ import (
 type Notifier struct {
 	initialRoutes []*route.Route
 	currentRoutes []*route.Route
-	fakeIPRoutes []*route.Route
+	fakeIPRoutes  []*route.Route
 
 	listener    listener.NetworkChangeListener
 	listenerMux sync.Mutex
@@ -119,3 +119,7 @@ func (n *Notifier) GetInitialRouteRanges() []string {
 	sort.Strings(initialStrings)
 	return initialStrings
 }
+
+func (n *Notifier) Close() {
+	// unused
+}

client/internal/routemanager/notifier/notifier_ios.go

@@ -3,6 +3,7 @@
 package notifier
 
 import (
+	"container/list"
 	"net/netip"
 	"slices"
 	"sort"
@@ -14,19 +15,26 @@ import (
 )
 
 type Notifier struct {
+	mu              sync.Mutex
+	cond            *sync.Cond
 	currentPrefixes []string
-
-	listener    listener.NetworkChangeListener
-	listenerMux sync.Mutex
+	listener        listener.NetworkChangeListener
+	queue           *list.List
+	closed          bool
 }
 
 func NewNotifier() *Notifier {
-	return &Notifier{}
+	n := &Notifier{
+		queue: list.New(),
+	}
+	n.cond = sync.NewCond(&n.mu)
+	go n.deliverLoop()
+	return n
 }
 
 func (n *Notifier) SetListener(listener listener.NetworkChangeListener) {
-	n.listenerMux.Lock()
-	defer n.listenerMux.Unlock()
+	n.mu.Lock()
+	defer n.mu.Unlock()
 	n.listener = listener
 }
 
@@ -43,32 +51,52 @@ func (n *Notifier) OnNewRoutes(route.HAMap) {
 }
 
 func (n *Notifier) OnNewPrefixes(prefixes []netip.Prefix) {
-	newNets := make([]string, 0)
+	newNets := make([]string, 0, len(prefixes))
 	for _, prefix := range prefixes {
 		newNets = append(newNets, prefix.String())
 	}
 
 	sort.Strings(newNets)
 
+	n.mu.Lock()
 	if slices.Equal(n.currentPrefixes, newNets) {
+		n.mu.Unlock()
 		return
 	}
-
 	n.currentPrefixes = newNets
-	n.notify()
+	routes := strings.Join(n.currentPrefixes, ",")
+	n.queue.PushBack(routes)
+	n.cond.Signal()
+	n.mu.Unlock()
 }
-func (n *Notifier) notify() {
-	n.listenerMux.Lock()
-	defer n.listenerMux.Unlock()
-	if n.listener == nil {
-		return
-	}
 
-	go func(l listener.NetworkChangeListener) {
-		l.OnNetworkChanged(strings.Join(n.currentPrefixes, ","))
-	}(n.listener)
+func (n *Notifier) Close() {
+	n.mu.Lock()
+	n.closed = true
+	n.cond.Signal()
+	n.mu.Unlock()
 }
 
 func (n *Notifier) GetInitialRouteRanges() []string {
 	return nil
 }
+
+func (n *Notifier) deliverLoop() {
+	for {
+		n.mu.Lock()
+		for n.queue.Len() == 0 && !n.closed {
+			n.cond.Wait()
+		}
+		if n.closed && n.queue.Len() == 0 {
+			n.mu.Unlock()
+			return
+		}
+		routes := n.queue.Remove(n.queue.Front()).(string)
+		l := n.listener
+		n.mu.Unlock()
+
+		if l != nil {
+			l.OnNetworkChanged(routes)
+		}
+	}
+}

client/internal/routemanager/notifier/notifier_other.go

@@ -38,3 +38,7 @@ func (n *Notifier) OnNewPrefixes(prefixes []netip.Prefix) {
 func (n *Notifier) GetInitialRouteRanges() []string {
 	return []string{}
 }
+
+func (n *Notifier) Close() {
+	// unused
+}

client/ios/NetBirdSDK/client.go

@@ -171,13 +171,13 @@ func (c *Client) Run(fd int32, interfaceName string, envList *EnvList) error {
 	c.onHostDnsFn = func([]string) {}
 	cfg.WgIface = interfaceName
 
-	connectClient := internal.NewConnectClient(ctx, cfg, c.recorder)
+	connectClient := internal.NewConnectClient(ctx, c.recorder)
 	c.setState(cfg, connectClient)
 	// Persist the latest sync response so DebugBundle can include the network
 	// map. On iOS this is backed by disk to keep it out of the constrained
 	// process memory (see the syncstore package).
 	connectClient.SetSyncResponsePersistence(true)
-	return connectClient.RunOniOS(fd, c.networkChangeListener, c.dnsManager, c.stateFile, c.cacheDir, c.logFilePath)
+	return connectClient.RunOniOS(cfg, fd, c.networkChangeListener, c.dnsManager, c.stateFile, c.cacheDir, c.logFilePath)
 }
 
 // Stop the internal client and free the resources

client/proto/daemon.pb.go

@@ -1849,10 +1849,13 @@ func (x *ManagementState) GetError() string {
 
 // RelayState contains the latest state of the relay
 type RelayState struct {
-	state         protoimpl.MessageState `protogen:"open.v1"`
-	URI           string                 `protobuf:"bytes,1,opt,name=URI,proto3" json:"URI,omitempty"`
-	Available     bool                   `protobuf:"varint,2,opt,name=available,proto3" json:"available,omitempty"`
-	Error         string                 `protobuf:"bytes,3,opt,name=error,proto3" json:"error,omitempty"`
+	state     protoimpl.MessageState `protogen:"open.v1"`
+	URI       string                 `protobuf:"bytes,1,opt,name=URI,proto3" json:"URI,omitempty"`
+	Available bool                   `protobuf:"varint,2,opt,name=available,proto3" json:"available,omitempty"`
+	Error     string                 `protobuf:"bytes,3,opt,name=error,proto3" json:"error,omitempty"`
+	// transport is the negotiated relay transport (e.g. "ws", "quic"),
+	// empty for stun/turn probes or when not connected.
+	Transport     string `protobuf:"bytes,4,opt,name=transport,proto3" json:"transport,omitempty"`
 	unknownFields protoimpl.UnknownFields
 	sizeCache     protoimpl.SizeCache
 }
@@ -1908,6 +1911,13 @@ func (x *RelayState) GetError() string {
 	return ""
 }
 
+func (x *RelayState) GetTransport() string {
+	if x != nil {
+		return x.Transport
+	}
+	return ""
+}
+
 type NSGroupState struct {
 	state         protoimpl.MessageState `protogen:"open.v1"`
 	Servers       []string               `protobuf:"bytes,1,rep,name=servers,proto3" json:"servers,omitempty"`
@@ -6646,12 +6656,13 @@ const file_daemon_proto_rawDesc = "" +
 	"\x0fManagementState\x12\x10\n" +
 	"\x03URL\x18\x01 \x01(\tR\x03URL\x12\x1c\n" +
 	"\tconnected\x18\x02 \x01(\bR\tconnected\x12\x14\n" +
-	"\x05error\x18\x03 \x01(\tR\x05error\"R\n" +
+	"\x05error\x18\x03 \x01(\tR\x05error\"p\n" +
 	"\n" +
 	"RelayState\x12\x10\n" +
 	"\x03URI\x18\x01 \x01(\tR\x03URI\x12\x1c\n" +
 	"\tavailable\x18\x02 \x01(\bR\tavailable\x12\x14\n" +
-	"\x05error\x18\x03 \x01(\tR\x05error\"r\n" +
+	"\x05error\x18\x03 \x01(\tR\x05error\x12\x1c\n" +
+	"\ttransport\x18\x04 \x01(\tR\ttransport\"r\n" +
 	"\fNSGroupState\x12\x18\n" +
 	"\aservers\x18\x01 \x03(\tR\aservers\x12\x18\n" +
 	"\adomains\x18\x02 \x03(\tR\adomains\x12\x18\n" +

client/proto/daemon.proto

@@ -380,6 +380,9 @@ message RelayState {
   string URI = 1;
   bool available = 2;
   string error = 3;
+  // transport is the negotiated relay transport (e.g. "ws", "quic"),
+  // empty for stun/turn probes or when not connected.
+  string transport = 4;
 }
 
 message NSGroupState {

client/server/capture.go

@@ -344,9 +344,6 @@ func (s *Server) clearCaptureIfOwner(sess *capture.Session, engine *internal.Eng
 }
 
 func (s *Server) getCaptureEngineLocked() (*internal.Engine, error) {
-	if s.connectClient == nil {
-		return nil, status.Error(codes.FailedPrecondition, "client not connected")
-	}
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return nil, status.Error(codes.FailedPrecondition, "engine not initialized")

client/server/debug.go

@@ -5,7 +5,6 @@ package server
 import (
 	"bytes"
 	"context"
-	"errors"
 	"fmt"
 	"runtime/pprof"
 
@@ -28,11 +27,9 @@ func (s *Server) DebugBundle(_ context.Context, req *proto.DebugBundleRequest) (
 	}
 
 	var clientMetrics debug.MetricsExporter
-	if s.connectClient != nil {
-		if engine := s.connectClient.Engine(); engine != nil {
-			if cm := engine.GetClientMetrics(); cm != nil {
-				clientMetrics = cm
-			}
+	if engine := s.connectClient.Engine(); engine != nil {
+		if cm := engine.GetClientMetrics(); cm != nil {
+			clientMetrics = cm
 		}
 	}
 
@@ -48,13 +45,10 @@ func (s *Server) DebugBundle(_ context.Context, req *proto.DebugBundleRequest) (
 	defer s.cleanupBundleCapture()
 
 	var refreshStatus func()
-	if s.connectClient != nil {
-		engine := s.connectClient.Engine()
-		if engine != nil {
-			refreshStatus = func() {
-				log.Debug("refreshing system health status for debug bundle")
-				engine.RunHealthProbes(true)
-			}
+	if engine := s.connectClient.Engine(); engine != nil {
+		refreshStatus = func() {
+			log.Debug("refreshing system health status for debug bundle")
+			engine.RunHealthProbes(true)
 		}
 	}
 
@@ -118,9 +112,7 @@ func (s *Server) SetLogLevel(_ context.Context, req *proto.SetLogLevelRequest) (
 
 	log.SetLevel(level)
 
-	if s.connectClient != nil {
-		s.connectClient.SetLogLevel(level)
-	}
+	s.connectClient.SetLogLevel(level)
 
 	log.Infof("Log level set to %s", level.String())
 
@@ -134,20 +126,13 @@ func (s *Server) SetSyncResponsePersistence(_ context.Context, req *proto.SetSyn
 
 	enabled := req.GetEnabled()
 	s.persistSyncResponse = enabled
-	if s.connectClient != nil {
-		s.connectClient.SetSyncResponsePersistence(enabled)
-	}
+	s.connectClient.SetSyncResponsePersistence(enabled)
 
 	return &proto.SetSyncResponsePersistenceResponse{}, nil
 }
 
 func (s *Server) getLatestSyncResponse() (*mgmProto.SyncResponse, error) {
-	cClient := s.connectClient
-	if cClient == nil {
-		return nil, errors.New("connect client is not initialized")
-	}
-
-	return cClient.GetLatestSyncResponse()
+	return s.connectClient.GetLatestSyncResponse()
 }
 
 // StartCPUProfile starts CPU profiling in the daemon.

client/server/mdm.go

@@ -3,7 +3,6 @@ package server
 import (
 	"context"
 	"fmt"
-	"time"
 
 	log "github.com/sirupsen/logrus"
 	"google.golang.org/grpc/codes"
@@ -39,12 +38,11 @@ type conflictCheck struct {
 // OS-native managed-config store reports a diff vs the last observation.
 //
 // Restart sequence:
-//  1. Cancel the active engine context (terminates connectWithRetryRuns).
-//  2. Wait briefly for that goroutine to exit (giveUpChan is closed on exit).
-//  3. Re-resolve Config from disk + MDM policy (Config.apply re-runs
+//  1. Stop the in-flight run via the supervisor (blocks until fully torn down).
+//  2. Re-resolve Config from disk + MDM policy (Config.apply re-runs
 //     applyMDMPolicy with the freshly loaded Policy).
-//  4. Spawn a fresh connectWithRetryRuns with the new context and config.
-//  5. Broadcast a SystemEvent so any GUI / CLI subscriber (SubscribeEvents
+//  3. Start a fresh run with the new config.
+//  4. Broadcast a SystemEvent so any GUI / CLI subscriber (SubscribeEvents
 //     RPC) can refresh its cached config view without polling.
 //
 // The callback runs in the ticker's own goroutine. Ticker has already
@@ -52,39 +50,24 @@ type conflictCheck struct {
 func (s *Server) onMDMPolicyChange(_, _ *mdm.Policy) error {
 	log.Warn("MDM policy changed; restarting engine to apply new configuration")
 
-	// Hold s.mutex for the entire restart sequence (cancel + quiescence
-	// wait + re-spawn). Any concurrent Up/Down/Status arriving while
-	// MDM is restarting blocks on the Lock until we are done — they
-	// then observe the post-restart state coherently. This is safe
-	// because the connectWithRetryRuns goroutine no longer acquires
-	// s.mutex in its defer (intent vs. goroutine-alive concerns are
-	// fully separated; see the connectionGoroutineRunning helper).
+	// Hold s.mutex for the entire restart sequence (stop + re-start). Any
+	// concurrent Up/Down/Status arriving while MDM is restarting blocks on the
+	// Lock until we are done — they then observe the post-restart state coherently.
 	s.mutex.Lock()
 	defer s.mutex.Unlock()
 
-	if !s.clientRunning {
-		// The client is not running, so there's no engine to restart.
+	if !s.connectClient.ConnectionRunning() {
+		// No run in flight, so there's no engine to restart.
 		return nil
 	}
+
+	// Cancel daemon-side login/status activities tied to the old run; the run
+	// itself is torn down atomically by the supervisor inside Restart (see
+	// restartEngineForMDMLocked), which stops and re-starts in one operation.
 	if s.actCancel != nil {
 		s.actCancel()
 	}
 
-	// Wait for previous connectWithRetryRuns to exit so we don't end up
-	// with two goroutines fighting over the same status recorder + engine.
-	// The teardown engages a fan-out of engine goroutines (peer workers,
-	// signal handler, route manager, ...). close(clientGiveUpChan)
-	// happens in the function-scope defer of connectWithRetryRuns, on
-	// every exit path (ctx cancel, backoff exhausted, panic) — see the
-	// defer in server.go.
-	if s.clientGiveUpChan != nil {
-		select {
-		case <-s.clientGiveUpChan:
-		case <-time.After(10 * time.Second):
-			return fmt.Errorf("failed to restart the engine due to timeout")
-		}
-	}
-
 	if err := s.restartEngineForMDMLocked(); err != nil {
 		log.Errorf("MDM restart failed: %v", err)
 		return err
@@ -131,14 +114,13 @@ func (s *Server) publishConfigChangedEvent(source string) {
 }
 
 // restartEngineForMDMLocked re-resolves the active profile config
-// (re-running applyMDMPolicy via Config.apply) and re-spawns
-// connectWithRetryRuns. Mirrors the tail of Server.Start so a runtime
-// MDM change behaves identically to a fresh boot under the new policy.
+// (re-running applyMDMPolicy via Config.apply) and starts a fresh run.
+// Mirrors the tail of Server.Start so a runtime MDM change behaves
+// identically to a fresh boot under the new policy.
 //
 // MUST be called with s.mutex held — onMDMPolicyChange holds the lock
-// for the entire restart sequence (cancel + quiescence wait + re-spawn)
-// so concurrent Up/Down/Status RPCs observe a coherent post-restart
-// state.
+// for the entire restart sequence so concurrent Up/Down/Status RPCs
+// observe a coherent post-restart state.
 func (s *Server) restartEngineForMDMLocked() error {
 	activeProf, err := s.profileManager.GetActiveProfileState()
 	if err != nil {
@@ -154,13 +136,13 @@ func (s *Server) restartEngineForMDMLocked() error {
 	s.statusRecorder.UpdateRosenpass(config.RosenpassEnabled, config.RosenpassPermissive)
 	s.statusRecorder.UpdateLazyConnection(config.LazyConnectionEnabled)
 
-	ctx, cancel := context.WithCancel(s.rootCtx)
+	_, cancel := context.WithCancel(s.rootCtx)
 	s.actCancel = cancel
-	s.clientRunning = true
-	s.clientRunningChan = make(chan struct{})
-	s.clientGiveUpChan = make(chan struct{})
-	log.Info("MDM restart: spawning connectWithRetryRuns with re-resolved config")
-	go s.connectWithRetryRuns(ctx, config, s.statusRecorder, s.clientRunningChan, s.clientGiveUpChan)
+	log.Info("MDM restart: atomically restarting the run with re-resolved config")
+	// MDM restart has no incoming RPC metadata; fire and forget. Restart is a
+	// single supervisor op (atomic stop+start), so there is no observable
+	// "stopped" window between tearing down the old run and starting the new.
+	s.connectClient.Restart(config, nil)
 	s.publishConfigChangedEvent("mdm")
 	return nil
 }

client/server/network.go

@@ -34,10 +34,6 @@ func (s *Server) ListNetworks(context.Context, *proto.ListNetworksRequest) (*pro
 		return nil, gstatus.Errorf(codes.Unavailable, errNetworksDisabled)
 	}
 
-	if s.connectClient == nil {
-		return nil, fmt.Errorf("not connected")
-	}
-
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return nil, fmt.Errorf("not connected")
@@ -147,10 +143,6 @@ func (s *Server) SelectNetworks(_ context.Context, req *proto.SelectNetworksRequ
 		return nil, gstatus.Errorf(codes.Unavailable, errNetworksDisabled)
 	}
 
-	if s.connectClient == nil {
-		return nil, fmt.Errorf("not connected")
-	}
-
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return nil, fmt.Errorf("not connected")
@@ -199,10 +191,6 @@ func (s *Server) DeselectNetworks(_ context.Context, req *proto.SelectNetworksRe
 		return nil, gstatus.Errorf(codes.Unavailable, errNetworksDisabled)
 	}
 
-	if s.connectClient == nil {
-		return nil, fmt.Errorf("not connected")
-	}
-
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return nil, fmt.Errorf("not connected")

client/server/server.go

@@ -8,12 +8,10 @@ import (
 	"os"
 	"os/exec"
 	"runtime"
-	"strconv"
 	"sync"
 	"sync/atomic"
 	"time"
 
-	"github.com/cenkalti/backoff/v4"
 	log "github.com/sirupsen/logrus"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
 	"google.golang.org/grpc/codes"
@@ -39,15 +37,7 @@ import (
 )
 
 const (
-	probeThreshold          = time.Second * 5
-	retryInitialIntervalVar = "NB_CONN_RETRY_INTERVAL_TIME"
-	maxRetryIntervalVar     = "NB_CONN_MAX_RETRY_INTERVAL_TIME"
-	maxRetryTimeVar         = "NB_CONN_MAX_RETRY_TIME_TIME"
-	retryMultiplierVar      = "NB_CONN_RETRY_MULTIPLIER"
-	defaultInitialRetryTime = 30 * time.Minute
-	defaultMaxRetryInterval = 60 * time.Minute
-	defaultMaxRetryTime     = 14 * 24 * time.Hour
-	defaultRetryMultiplier  = 1.7
+	probeThreshold = time.Second * 5
 
 	// JWT token cache TTL for the client daemon (disabled by default)
 	defaultJWTCacheTTL = 0
@@ -72,15 +62,8 @@ type Server struct {
 	mutex  sync.Mutex
 	config *profilemanager.Config
 	proto.UnimplementedDaemonServiceServer
-	// clientRunning tracks "the daemon wants to be connected" — set true by
-	// Start / Up, cleared by Down / Logout. Persists across retry
-	// loops, signal disconnects, and ErrResetConnection cycles. NOT
-	// changed by connectWithRetryRuns goroutine exit — for that
-	// (goroutine-still-alive) check, see connectionGoroutineRunning() which
-	// derives from clientGiveUpChan close state. Protected by s.mutex.
-	clientRunning          bool
-	clientRunningChan chan struct{}
-	clientGiveUpChan  chan struct{} // closed when connectWithRetryRuns goroutine exits
+	// Run state (in-flight? established/done channels?) is owned entirely by the
+	// supervisor inside connectClient — the daemon keeps no per-run fields.
 
 	connectClient *internal.ConnectClient
 
@@ -136,6 +119,13 @@ func New(ctx context.Context, logFile string, configFile string, profilesDisable
 		networksDisabled:       networksDisabled,
 		jwtCache:               newJWTCache(),
 	}
+	// The ConnectClient is daemon-lifetime: build it exactly once, here. Its
+	// supervisor lives as long as the daemon; Up/Down/MDM and reconnects all
+	// drive this same instance. updateManager isn't ready yet (created in
+	// Start) and is injected there via SetUpdateManager.
+	s.connectClient = internal.NewConnectClient(ctx, s.statusRecorder)
+	s.connectClient.SetSyncResponsePersistence(s.persistSyncResponse)
+
 	agent := &serverAgent{s}
 	s.sleepHandler = sleephandler.New(agent)
 	s.startSleepDetector()
@@ -147,7 +137,7 @@ func (s *Server) Start() error {
 	s.mutex.Lock()
 	defer s.mutex.Unlock()
 
-	if s.clientRunning {
+	if s.connectClient.ConnectionRunning() {
 		return nil
 	}
 
@@ -165,6 +155,7 @@ func (s *Server) Start() error {
 		stateMgr := statemanager.New(s.profileManager.GetStatePath())
 		s.updateManager = updater.NewManager(s.statusRecorder, stateMgr)
 		s.updateManager.CheckUpdateSuccess(s.rootCtx)
+		s.connectClient.SetUpdateManager(s.updateManager)
 	}
 
 	// MDM policy reload ticker: every minute the desktop daemon re-reads
@@ -190,7 +181,9 @@ func (s *Server) Start() error {
 		return nil
 	}
 
-	ctx, cancel := context.WithCancel(s.rootCtx)
+	// actCancel cancels in-flight foreground operations (login/status); the run
+	// itself is owned by the supervisor and stopped via Stop, not this cancel.
+	_, cancel := context.WithCancel(s.rootCtx)
 	s.actCancel = cancel
 
 	// copy old default config
@@ -232,99 +225,14 @@ func (s *Server) Start() error {
 		return nil
 	}
 
-	s.clientRunning = true
-	s.clientRunningChan = make(chan struct{})
-	s.clientGiveUpChan = make(chan struct{})
-	go s.connectWithRetryRuns(ctx, config, s.statusRecorder, s.clientRunningChan, s.clientGiveUpChan)
+	// Boot autoconnect: no incoming RPC metadata. The supervisor runs the
+	// client and reconnects internally; we just fire and forget (the run owns
+	// its established/done channels).
+	s.connectClient.RunAsync(config, nil)
 	s.publishConfigChangedEvent("startup")
 	return nil
 }
 
-// connectWithRetryRuns runs the client connection with a backoff strategy where we retry the operation as additional
-// mechanism to keep the client connected even when the connection is lost.
-// we cancel retry if the client receive a stop or down command, or if disable auto connect is configured.
-//
-// The goroutine's exit is signalled to the daemon via close(giveUpChan)
-// — placed in the function-scope defer so every return path (panic,
-// DisableAutoConnect early-exit, backoff exhausted, ctx cancel) closes
-// it. Callers that need to observe "is the goroutine still alive?" use
-// Server.connectionGoroutineRunning() which non-blockingly checks the close state
-// of clientGiveUpChan. The defer does NOT touch s.mutex; the daemon's
-// "intent" (clientRunning) is maintained by the RPC handlers, not by this
-// goroutine.
-func (s *Server) connectWithRetryRuns(ctx context.Context, profileConfig *profilemanager.Config, statusRecorder *peer.Status, runningChan chan struct{}, giveUpChan chan struct{}) {
-	defer func() {
-		if giveUpChan != nil {
-			close(giveUpChan)
-		}
-	}()
-
-	if s.config.DisableAutoConnect {
-		if err := s.connect(ctx, s.config, s.statusRecorder, runningChan); err != nil {
-			log.Debugf("run client connection exited with error: %v", err)
-		}
-		log.Tracef("client connection exited")
-		return
-	}
-
-	backOff := getConnectWithBackoff(ctx)
-	go func() {
-		t := time.NewTicker(24 * time.Hour)
-		for {
-			select {
-			case <-ctx.Done():
-				t.Stop()
-				return
-			case <-t.C:
-				mgmtState := statusRecorder.GetManagementState()
-				signalState := statusRecorder.GetSignalState()
-				if mgmtState.Connected && signalState.Connected {
-					log.Tracef("resetting status")
-					backOff.Reset()
-				} else {
-					log.Tracef("not resetting status: mgmt: %v, signal: %v", mgmtState.Connected, signalState.Connected)
-				}
-			}
-		}
-	}()
-
-	runOperation := func() error {
-		err := s.connect(ctx, profileConfig, statusRecorder, runningChan)
-		if err != nil {
-			log.Debugf("run client connection exited with error: %v. Will retry in the background", err)
-			return err
-		}
-
-		log.Tracef("client connection exited gracefully, do not need to retry")
-		return nil
-	}
-
-	if err := backoff.Retry(runOperation, backOff); err != nil {
-		log.Errorf("operation failed: %v", err)
-	}
-	// giveUpChan is closed by the function-scope defer.
-}
-
-// connectionGoroutineRunning reports whether the connectWithRetryRuns goroutine is
-// still running. Returns false when no goroutine has ever been started
-// AND when the most recent one has already closed clientGiveUpChan on
-// exit (whether due to ctx cancel, DisableAutoConnect single-shot
-// completion, or backoff retry exhaustion).
-//
-// MUST be called with s.mutex held — accesses s.clientGiveUpChan which
-// is written by Start/Up under the same lock.
-func (s *Server) connectionGoroutineRunning() bool {
-	if s.clientGiveUpChan == nil {
-		return false
-	}
-	select {
-	case <-s.clientGiveUpChan:
-		return false
-	default:
-		return true
-	}
-}
-
 // loginAttempt attempts to login using the provided information. it returns a status in case something fails
 func (s *Server) loginAttempt(ctx context.Context, setupKey, jwtToken string) (internal.StatusType, error) {
 	authClient, err := auth.NewAuth(ctx, s.config.PrivateKey, s.config.ManagementURL, s.config)
@@ -375,7 +283,7 @@ func (s *Server) SetConfig(callerCtx context.Context, msg *proto.SetConfigReques
 		return nil, err
 	}
 
-	config, err := setConfigInputFromRequest(msg)
+	config, err := s.setConfigInputFromRequest(msg)
 	if err != nil {
 		return nil, err
 	}
@@ -398,17 +306,18 @@ func (s *Server) SetConfig(callerCtx context.Context, msg *proto.SetConfigReques
 // field is its own optional case. Returns the resolved ConfigInput
 // and a non-nil error only when the active profile file path cannot
 // be determined.
-func setConfigInputFromRequest(msg *proto.SetConfigRequest) (profilemanager.ConfigInput, error) {
+func (s *Server) setConfigInputFromRequest(msg *proto.SetConfigRequest) (profilemanager.ConfigInput, error) {
 	var config profilemanager.ConfigInput
 
 	resolved, err := s.resolveProfileHandle(msg.ProfileName, msg.Username)
 	if err != nil {
 		log.Errorf("failed to resolve profile %q: %v", msg.ProfileName, err)
-		return nil, err
+		return config, err
 	}
 	profPath := resolved.Path
 	if profPath == "" {
 		profPath = profilemanager.DefaultConfigPath
+	}
 	config.ConfigPath = profPath
 
 	if msg.ManagementUrl != "" {
@@ -719,13 +628,22 @@ func (s *Server) WaitSSOLogin(callerCtx context.Context, msg *proto.WaitSSOLogin
 // Up starts engine work in the daemon.
 func (s *Server) Up(callerCtx context.Context, msg *proto.UpRequest) (*proto.UpResponse, error) {
 	s.mutex.Lock()
-	// clientRunning is the daemon-intent flag (set by previous Up/Start, cleared
-	// by Down). connectionGoroutineRunning() reports whether the previous retry-loop
-	// goroutine is still trying. When intent is up AND goroutine is alive,
-	// the existing engine is on the job — just wait for it. When intent
-	// is up but the goroutine has given up (backoff exhausted) OR when
-	// intent is down, fall through to spawn a fresh retry loop.
-	if s.clientRunning && s.connectionGoroutineRunning() {
+
+	// The client (and its supervisor) is built once in New(), so a nil here
+	// never happens in production — Up is only reachable after New() has run and
+	// the gRPC server is serving. The real case this guards is the daemon
+	// SHUTTING DOWN: rootCtx is cancelled, the supervisor is no longer accepting
+	// commands, so ServiceRunning() is false even though the client exists. Bail
+	// loud instead of enqueuing a run that will never start. (nil only happens in
+	// tests that build a Server without New(); ServiceRunning is nil-safe.)
+	if !s.connectClient.ServiceRunning() {
+		s.mutex.Unlock()
+		return nil, fmt.Errorf("service is not running, start the netbird service for 'up' to take effect")
+	}
+
+	// If a connection run is already in flight, the existing engine is on the
+	// job — just wait for it. Otherwise fall through to start a fresh run.
+	if s.connectClient.ConnectionRunning() {
 		state := internal.CtxGetState(s.rootCtx)
 		status, err := state.Status()
 		if err != nil {
@@ -763,14 +681,14 @@ func (s *Server) Up(callerCtx context.Context, msg *proto.UpRequest) (*proto.UpR
 	if s.actCancel != nil {
 		s.actCancel()
 	}
-	ctx, cancel := context.WithCancel(s.rootCtx)
-	md, ok := metadata.FromIncomingContext(callerCtx)
-	if ok {
-		ctx = metadata.NewOutgoingContext(ctx, md)
-	}
-
+	// actCancel cancels in-flight foreground ops (login/status); the run is
+	// owned by the supervisor and stopped via Stop, not this cancel.
+	_, cancel := context.WithCancel(s.rootCtx)
 	s.actCancel = cancel
 
+	// Forward the caller's gRPC metadata (e.g. UI user-agent) into the run.
+	md, _ := metadata.FromIncomingContext(callerCtx)
+
 	if s.config == nil {
 		s.mutex.Unlock()
 		return nil, fmt.Errorf("config is not defined, please call login command first")
@@ -811,35 +729,26 @@ func (s *Server) Up(callerCtx context.Context, msg *proto.UpRequest) (*proto.UpR
 	s.statusRecorder.UpdateManagementAddress(s.config.ManagementURL.String())
 	s.statusRecorder.UpdateRosenpass(s.config.RosenpassEnabled, s.config.RosenpassPermissive)
 
-	s.clientRunning = true
-	s.clientRunningChan = make(chan struct{})
-	s.clientGiveUpChan = make(chan struct{})
-
-	go s.connectWithRetryRuns(ctx, s.config, s.statusRecorder, s.clientRunningChan, s.clientGiveUpChan)
+	s.connectClient.RunAsync(s.config, md)
 	s.publishConfigChangedEvent("up_rpc")
 
 	s.mutex.Unlock()
 	return s.waitForUp(callerCtx)
 }
 
-// todo: handle potential race conditions
+// waitForUp blocks until the in-flight run becomes established (success) or ends
+// before that (failure). The wait is owned by the supervisor (via the client) —
+// the daemon holds no per-run state here.
 func (s *Server) waitForUp(callerCtx context.Context) (*proto.UpResponse, error) {
 	timeoutCtx, cancel := context.WithTimeout(callerCtx, 50*time.Second)
 	defer cancel()
 
-	select {
-	case <-s.clientGiveUpChan:
-		return nil, fmt.Errorf("client gave up to connect")
-	case <-s.clientRunningChan:
-		s.isSessionActive.Store(true)
-		return &proto.UpResponse{}, nil
-	case <-callerCtx.Done():
-		log.Debug("context done, stopping the wait for engine to become ready")
-		return nil, callerCtx.Err()
-	case <-timeoutCtx.Done():
-		log.Debug("up is timed out, stopping the wait for engine to become ready")
-		return nil, timeoutCtx.Err()
+	if err := s.connectClient.WaitEstablishedOrDone(timeoutCtx); err != nil {
+		log.Debugf("waiting for the connection to be established failed: %v", err)
+		return nil, fmt.Errorf("connection not established: %w", err)
 	}
+	s.isSessionActive.Store(true)
+	return &proto.UpResponse{}, nil
 }
 
 // resolveProfileHandle resolves a wire-level profile handle (display
@@ -934,11 +843,11 @@ func (s *Server) SwitchProfile(callerCtx context.Context, msg *proto.SwitchProfi
 // Down engine work in the daemon.
 func (s *Server) Down(ctx context.Context, _ *proto.DownRequest) (*proto.DownResponse, error) {
 	s.mutex.Lock()
+	defer s.mutex.Unlock()
 
-	giveUpChan := s.clientGiveUpChan
-
+	// cleanupConnection stops the run through the supervisor, which blocks until
+	// the run has fully unwound — no separate goroutine-quiescence wait needed.
 	if err := s.cleanupConnection(); err != nil {
-		s.mutex.Unlock()
 		// todo review to update the status in case any type of error
 		log.Errorf("failed to shut down properly: %v", err)
 		return nil, err
@@ -947,20 +856,6 @@ func (s *Server) Down(ctx context.Context, _ *proto.DownRequest) (*proto.DownRes
 	state := internal.CtxGetState(s.rootCtx)
 	state.Set(internal.StatusIdle)
 
-	s.mutex.Unlock()
-
-	// Wait for the connectWithRetryRuns goroutine to finish with a short timeout.
-	// This prevents the goroutine from setting ErrResetConnection after Down() returns.
-	// The giveUpChan is closed at the end of connectWithRetryRuns.
-	if giveUpChan != nil {
-		select {
-		case <-giveUpChan:
-			log.Debugf("client goroutine finished successfully")
-		case <-time.After(5 * time.Second):
-			log.Warnf("timeout waiting for client goroutine to finish, proceeding anyway")
-		}
-	}
-
 	return &proto.DownResponse{}, nil
 }
 
@@ -971,34 +866,19 @@ func (s *Server) cleanupConnection() error {
 		return ErrServiceNotUp
 	}
 
-	// Daemon intent flips to "down" — all callers (Down RPC,
-	// Logout RPC handlers) tear down the connection because the user
-	// explicitly asked for it. MDM restart does NOT go through this
-	// path, so its clientRunning stays true.
-	s.clientRunning = false
-
-	// Capture the engine reference before cancelling the context.
-	// After actCancel(), the connectWithRetryRuns goroutine wakes up
-	// and sets connectClient.engine = nil, causing connectClient.Stop()
-	// to skip the engine shutdown entirely.
-	var engine *internal.Engine
-	if s.connectClient != nil {
-		engine = s.connectClient.Engine()
+	// Tear the client down through the lifecycle supervisor BEFORE cancelling
+	// the retry context. Stop serializes on the supervisor queue and blocks
+	// until the in-flight run has fully unwound (a clean, synchronous teardown).
+	// It must run before actCancel: cancelling the context first would make
+	// Stop observe a dead context and return early without waiting.
+	if err := s.connectClient.Stop(); err != nil {
+		return err
 	}
 
+	// Stop the retry goroutine so it does not start a fresh run. The client
+	// itself is daemon-lifetime and intentionally kept (a later Up reuses it).
 	s.actCancel()
 
-	if s.connectClient == nil {
-		return nil
-	}
-
-	if engine != nil {
-		if err := engine.Stop(); err != nil {
-			return err
-		}
-	}
-
-	s.connectClient = nil
 	s.isSessionActive.Store(false)
 
 	log.Infof("service is down")
@@ -1133,7 +1013,7 @@ func (s *Server) validateProfileOperation(id profilemanager.ID, allowActiveProfi
 
 func (s *Server) logoutFromProfile(ctx context.Context, profile *profilemanager.Profile) error {
 	activeProf, err := s.profileManager.GetActiveProfileState()
-	if err == nil && activeProf.ID == profile.ID && s.connectClient != nil {
+	if err == nil && activeProf.ID == profile.ID && s.connectClient.ConnectionRunning() {
 		return s.sendLogoutRequest(ctx)
 	}
 
@@ -1179,48 +1059,13 @@ func (s *Server) Status(
 	ctx context.Context,
 	msg *proto.StatusRequest,
 ) (*proto.StatusResponse, error) {
-	s.mutex.Lock()
-	// Only wait if the retry-loop goroutine is alive and making
-	// progress. clientRunning=true with connectionGoroutineRunning=false means the
-	// backoff has given up — there is nothing to wait for; let the
-	// caller observe the failed status directly.
-	alive := s.connectionGoroutineRunning()
-	s.mutex.Unlock()
-
-	if msg.WaitForReady != nil && *msg.WaitForReady && alive {
-		state := internal.CtxGetState(s.rootCtx)
-		status, err := state.Status()
-		if err != nil {
-			return nil, err
-		}
-
-		if status != internal.StatusIdle && status != internal.StatusConnected && status != internal.StatusConnecting {
-			s.actCancel()
-		}
-
-		ticker := time.NewTicker(1 * time.Second)
-		defer ticker.Stop()
-	loop:
-		for {
-			select {
-			case <-s.clientGiveUpChan:
-				ticker.Stop()
-				break loop
-			case <-s.clientRunningChan:
-				ticker.Stop()
-				break loop
-			case <-ticker.C:
-				status, err := state.Status()
-				if err != nil {
-					continue
-				}
-				if status != internal.StatusIdle && status != internal.StatusConnected && status != internal.StatusConnecting {
-					s.actCancel()
-				}
-				continue
-			case <-ctx.Done():
-				return nil, ctx.Err()
-			}
+	// A run that hits a terminal auth failure now exits on its own (engine marks
+	// NeedsLogin), so we no longer poll-and-cancel: we wait for the in-flight run
+	// to become established or to end. With no run in flight this returns
+	// immediately (errNoRunInFlight); either way we then report the status below.
+	if msg.WaitForReady != nil && *msg.WaitForReady {
+		if err := s.connectClient.WaitEstablishedOrDone(ctx); err != nil && ctx.Err() != nil {
+			return nil, ctx.Err()
 		}
 	}
 
@@ -1258,10 +1103,6 @@ func (s *Server) getSSHServerState() *proto.SSHServerState {
 	connectClient := s.connectClient
 	s.mutex.Unlock()
 
-	if connectClient == nil {
-		return nil
-	}
-
 	engine := connectClient.Engine()
 	if engine == nil {
 		return nil
@@ -1299,10 +1140,6 @@ func (s *Server) GetPeerSSHHostKey(
 	statusRecorder := s.statusRecorder
 	s.mutex.Unlock()
 
-	if connectClient == nil {
-		return nil, errors.New("client not initialized")
-	}
-
 	engine := connectClient.Engine()
 	if engine == nil {
 		return nil, errors.New("engine not started")
@@ -1469,17 +1306,13 @@ func (s *Server) WaitJWTToken(
 // ExposeService exposes a local port via the NetBird reverse proxy.
 func (s *Server) ExposeService(req *proto.ExposeServiceRequest, srv proto.DaemonService_ExposeServiceServer) error {
 	s.mutex.Lock()
-	if !s.clientRunning {
+	if !s.connectClient.ConnectionRunning() {
 		s.mutex.Unlock()
 		return gstatus.Errorf(codes.FailedPrecondition, "client is not running, run 'netbird up' first")
 	}
 	connectClient := s.connectClient
 	s.mutex.Unlock()
 
-	if connectClient == nil {
-		return gstatus.Errorf(codes.FailedPrecondition, "client not initialized")
-	}
-
 	engine := connectClient.Engine()
 	if engine == nil {
 		return gstatus.Errorf(codes.FailedPrecondition, "engine not initialized")
@@ -1533,10 +1366,6 @@ func isUnixRunningDesktop() bool {
 }
 
 func (s *Server) runProbes(waitForProbeResult bool) {
-	if s.connectClient == nil {
-		return
-	}
-
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return
@@ -1815,22 +1644,6 @@ func (s *Server) GetFeatures(ctx context.Context, msg *proto.GetFeaturesRequest)
 	return features, nil
 }
 
-func (s *Server) connect(ctx context.Context, config *profilemanager.Config, statusRecorder *peer.Status, runningChan chan struct{}) error {
-	log.Tracef("running client connection")
-	client := internal.NewConnectClient(ctx, config, statusRecorder)
-	client.SetUpdateManager(s.updateManager)
-	client.SetSyncResponsePersistence(s.persistSyncResponse)
-
-	s.mutex.Lock()
-	s.connectClient = client
-	s.mutex.Unlock()
-
-	if err := client.Run(runningChan, s.logFile); err != nil {
-		return err
-	}
-	return nil
-}
-
 // MDM authority: when the platform-native MDM source sets a kill switch
 // key (regardless of true/false value), that value wins. The CLI flag
 // supplied at service install time is the fallback used only when the
@@ -1892,45 +1705,6 @@ func (s *Server) onSessionExpire() {
 	}
 }
 
-// getConnectWithBackoff returns a backoff with exponential backoff strategy for connection retries
-func getConnectWithBackoff(ctx context.Context) backoff.BackOff {
-	initialInterval := parseEnvDuration(retryInitialIntervalVar, defaultInitialRetryTime)
-	maxInterval := parseEnvDuration(maxRetryIntervalVar, defaultMaxRetryInterval)
-	maxElapsedTime := parseEnvDuration(maxRetryTimeVar, defaultMaxRetryTime)
-	multiplier := defaultRetryMultiplier
-
-	if envValue := os.Getenv(retryMultiplierVar); envValue != "" {
-		// parse the multiplier from the environment variable string value to float64
-		value, err := strconv.ParseFloat(envValue, 64)
-		if err != nil {
-			log.Warnf("unable to parse environment variable %s: %s. using default: %f", retryMultiplierVar, envValue, multiplier)
-		} else {
-			multiplier = value
-		}
-	}
-
-	return backoff.WithContext(&backoff.ExponentialBackOff{
-		InitialInterval:     initialInterval,
-		RandomizationFactor: 1,
-		Multiplier:          multiplier,
-		MaxInterval:         maxInterval,
-		MaxElapsedTime:      maxElapsedTime, // 14 days
-		Stop:                backoff.Stop,
-		Clock:               backoff.SystemClock,
-	}, ctx)
-}
-
-// parseEnvDuration parses the environment variable and returns the duration
-func parseEnvDuration(envVar string, defaultDuration time.Duration) time.Duration {
-	if envValue := os.Getenv(envVar); envValue != "" {
-		if duration, err := time.ParseDuration(envValue); err == nil {
-			return duration
-		}
-		log.Warnf("unable to parse environment variable %s: %s. using default: %s", envVar, envValue, defaultDuration)
-	}
-	return defaultDuration
-}
-
 // sendTerminalNotification sends a terminal notification message
 // to inform the user that the NetBird connection session has expired.
 func sendTerminalNotification() error {

client/server/server_connect_test.go

@@ -15,14 +15,19 @@ import (
 )
 
 func newTestServer() *Server {
-	return &Server{
-		rootCtx:        context.Background(),
+	ctx := context.Background()
+	s := &Server{
+		rootCtx:        ctx,
 		statusRecorder: peer.NewRecorder(""),
 	}
+	// Honor the production invariant: the daemon-lifetime client always exists
+	// (built in New). Server methods rely on s.connectClient being non-nil.
+	s.connectClient = internal.NewConnectClient(ctx, s.statusRecorder)
+	return s
 }
 
 func newDummyConnectClient(ctx context.Context) *internal.ConnectClient {
-	return internal.NewConnectClient(ctx, nil, nil)
+	return internal.NewConnectClient(ctx, nil)
 }
 
 // TestConnectSetsClientWithMutex validates that connect() sets s.connectClient
@@ -87,41 +92,36 @@ func TestConcurrentConnectClientAccess(t *testing.T) {
 	assert.Equal(t, 50, nilCount+setCount, "all goroutines should complete without panic")
 }
 
-// TestCleanupConnection_ClearsConnectClient validates that cleanupConnection
-// properly nils out connectClient.
-func TestCleanupConnection_ClearsConnectClient(t *testing.T) {
+// TestCleanupConnection_KeepsClientStopsRunning validates that cleanupConnection
+// clears the daemon "up" intent but KEEPS the daemon-lifetime ConnectClient
+// (it is reused across Up/Down; only the run is stopped).
+func TestCleanupConnection_KeepsClientStopsRunning(t *testing.T) {
 	s := newTestServer()
 	_, cancel := context.WithCancel(context.Background())
 	s.actCancel = cancel
 
-	s.connectClient = newDummyConnectClient(context.Background())
-	s.clientRunning = true
-
 	err := s.cleanupConnection()
 	require.NoError(t, err)
 
-	assert.Nil(t, s.connectClient, "connectClient should be nil after cleanup")
-	assert.False(t, s.clientRunning, "clientRunning should be cleared after cleanup (intent = down)")
+	assert.NotNil(t, s.connectClient, "connectClient is daemon-lifetime and must persist after cleanup")
+	assert.False(t, s.connectClient.ConnectionRunning(), "no run should be in flight after cleanup")
 }
 
-// TestCleanState_NilConnectClient validates that CleanState doesn't panic
-// when connectClient is nil.
-func TestCleanState_NilConnectClient(t *testing.T) {
+// TestCleanState_NotConnected validates that CleanState doesn't panic when no
+// connection run is in flight.
+func TestCleanState_NotConnected(t *testing.T) {
 	s := newTestServer()
-	s.connectClient = nil
-	s.profileManager = nil // will cause error if it tries to proceed past the nil check
+	s.profileManager = nil // will cause error if it tries to proceed
 
-	// Should not panic — the nil check should prevent calling Status() on nil
 	assert.NotPanics(t, func() {
 		_, _ = s.CleanState(context.Background(), &proto.CleanStateRequest{All: true})
 	})
 }
 
-// TestDeleteState_NilConnectClient validates that DeleteState doesn't panic
-// when connectClient is nil.
-func TestDeleteState_NilConnectClient(t *testing.T) {
+// TestDeleteState_NotConnected validates that DeleteState doesn't panic when no
+// connection run is in flight.
+func TestDeleteState_NotConnected(t *testing.T) {
 	s := newTestServer()
-	s.connectClient = nil
 	s.profileManager = nil
 
 	assert.NotPanics(t, func() {
@@ -129,60 +129,6 @@ func TestDeleteState_NilConnectClient(t *testing.T) {
 	})
 }
 
-// TestDownThenUp_StaleRunningChan documents the known state issue where
-// clientRunningChan from a previous connection is already closed, causing
-// waitForUp() to return immediately on reconnect.
-func TestDownThenUp_StaleRunningChan(t *testing.T) {
-	s := newTestServer()
-
-	// Simulate state after a successful connection
-	s.clientRunning = true
-	s.clientRunningChan = make(chan struct{})
-	close(s.clientRunningChan) // closed when engine started
-	s.clientGiveUpChan = make(chan struct{})
-	s.connectClient = newDummyConnectClient(context.Background())
-
-	_, cancel := context.WithCancel(context.Background())
-	s.actCancel = cancel
-
-	// Simulate Down(): cleanupConnection sets connectClient = nil and
-	// flips clientRunning to false (intent = down). The connectionGoroutineRunning state
-	// remains independent of intent — derived from clientGiveUpChan.
-	s.mutex.Lock()
-	err := s.cleanupConnection()
-	s.mutex.Unlock()
-	require.NoError(t, err)
-
-	// After cleanup: connectClient is nil, clientRunning is false (intent
-	// cleared by cleanupConnection), connectionGoroutineRunning may still be true
-	// (goroutine teardown is independent of the intent flag).
-	s.mutex.Lock()
-	assert.Nil(t, s.connectClient, "connectClient should be nil after cleanup")
-	assert.False(t, s.clientRunning, "clientRunning should be cleared by cleanupConnection (intent = down)")
-	s.mutex.Unlock()
-
-	// waitForUp() returns immediately due to stale closed clientRunningChan
-	ctx, ctxCancel := context.WithTimeout(context.Background(), 2*time.Second)
-	defer ctxCancel()
-
-	waitDone := make(chan error, 1)
-	go func() {
-		_, err := s.waitForUp(ctx)
-		waitDone <- err
-	}()
-
-	select {
-	case err := <-waitDone:
-		assert.NoError(t, err, "waitForUp returns success on stale channel")
-		// But connectClient is still nil — this is the stale state issue
-		s.mutex.Lock()
-		assert.Nil(t, s.connectClient, "connectClient is nil despite waitForUp success")
-		s.mutex.Unlock()
-	case <-time.After(1 * time.Second):
-		t.Fatal("waitForUp should have returned immediately due to stale closed channel")
-	}
-}
-
 // TestConnectClient_EngineNilOnFreshClient validates that a newly created
 // ConnectClient has nil Engine (before Run is called).
 func TestConnectClient_EngineNilOnFreshClient(t *testing.T) {

client/server/server_test.go

@@ -31,7 +31,6 @@ import (
 	"google.golang.org/grpc/keepalive"
 
 	"github.com/netbirdio/netbird/client/internal"
-	"github.com/netbirdio/netbird/client/internal/peer"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	daemonProto "github.com/netbirdio/netbird/client/proto"
 	"github.com/netbirdio/netbird/management/server"
@@ -61,65 +60,6 @@ var (
 	}
 )
 
-// TestConnectWithRetryRuns checks that the connectWithRetry function runs and runs the retries according to the times specified via environment variables
-// we will use a management server started via to simulate the server and capture the number of retries
-func TestConnectWithRetryRuns(t *testing.T) {
-	// start the signal server
-	_, signalAddr, err := startSignal(t)
-	if err != nil {
-		t.Fatalf("failed to start signal server: %v", err)
-	}
-
-	counter := 0
-	// start the management server
-	_, mgmtAddr, err := startManagement(t, signalAddr, &counter)
-	if err != nil {
-		t.Fatalf("failed to start management server: %v", err)
-	}
-
-	ctx := internal.CtxInitState(context.Background())
-
-	ctx, cancel := context.WithDeadline(ctx, time.Now().Add(30*time.Second))
-	defer cancel()
-	// create new server
-	ic := profilemanager.ConfigInput{
-		ManagementURL: "http://" + mgmtAddr,
-		ConfigPath:    t.TempDir() + "/test-profile.json",
-	}
-
-	config, err := profilemanager.UpdateOrCreateConfig(ic)
-	if err != nil {
-		t.Fatalf("failed to create config: %v", err)
-	}
-
-	currUser, err := user.Current()
-	require.NoError(t, err)
-
-	pm := profilemanager.ServiceManager{}
-	err = pm.SetActiveProfileState(&profilemanager.ActiveProfileState{
-		ID:       "test-profile",
-		Username: currUser.Username,
-	})
-	if err != nil {
-		t.Fatalf("failed to set active profile state: %v", err)
-	}
-
-	s := New(ctx, "debug", "", false, false, false, false)
-
-	s.config = config
-
-	s.statusRecorder = peer.NewRecorder(config.ManagementURL.String())
-	t.Setenv(retryInitialIntervalVar, "1s")
-	t.Setenv(maxRetryIntervalVar, "2s")
-	t.Setenv(maxRetryTimeVar, "5s")
-	t.Setenv(retryMultiplierVar, "1")
-
-	s.connectWithRetryRuns(ctx, config, s.statusRecorder, nil, nil)
-	if counter < 3 {
-		t.Fatalf("expected counter > 2, got %d", counter)
-	}
-}
-
 func TestServer_Up(t *testing.T) {
 	tempDir := t.TempDir()
 	origDefaultProfileDir := profilemanager.DefaultConfigPathDir

client/server/setconfig_mdm_test.go

@@ -62,7 +62,7 @@ func setupServerWithProfile(t *testing.T) (s *Server, ctx context.Context, profN
 
 	pm := profilemanager.ServiceManager{}
 	require.NoError(t, pm.SetActiveProfileState(&profilemanager.ActiveProfileState{
-		Name:     profName,
+		ID:       profilemanager.ID(profName),
 		Username: currUser.Username,
 	}))
 
@@ -107,9 +107,9 @@ func TestSetConfig_MDMReject_SingleField(t *testing.T) {
 
 func TestSetConfig_MDMReject_MultipleFields(t *testing.T) {
 	withMDMPolicy(t, mdm.NewPolicy(map[string]any{
-		mdm.KeyManagementURL:     "https://mdm.example.com:443",
-		mdm.KeyBlockInbound:      true,
-		mdm.KeyRosenpassEnabled:  true,
+		mdm.KeyManagementURL:    "https://mdm.example.com:443",
+		mdm.KeyBlockInbound:     true,
+		mdm.KeyRosenpassEnabled: true,
 	}))
 
 	s, ctx, profName, username, _ := setupServerWithProfile(t)

client/server/state.go

@@ -9,7 +9,6 @@ import (
 	"google.golang.org/grpc/status"
 
 	nberrors "github.com/netbirdio/netbird/client/errors"
-	"github.com/netbirdio/netbird/client/internal"
 	"github.com/netbirdio/netbird/client/internal/routemanager/systemops"
 	"github.com/netbirdio/netbird/client/internal/statemanager"
 	"github.com/netbirdio/netbird/client/proto"
@@ -38,7 +37,7 @@ func (s *Server) ListStates(_ context.Context, _ *proto.ListStatesRequest) (*pro
 
 // CleanState handles cleaning of states (performing cleanup operations)
 func (s *Server) CleanState(ctx context.Context, req *proto.CleanStateRequest) (*proto.CleanStateResponse, error) {
-	if s.connectClient != nil && (s.connectClient.Status() == internal.StatusConnected || s.connectClient.Status() == internal.StatusConnecting) {
+	if s.connectClient.ConnectionRunning() {
 		return nil, status.Errorf(codes.FailedPrecondition, "cannot clean state while connecting or connected, run 'netbird down' first.")
 	}
 
@@ -81,7 +80,7 @@ func (s *Server) CleanState(ctx context.Context, req *proto.CleanStateRequest) (
 
 // DeleteState handles deletion of states without cleanup
 func (s *Server) DeleteState(ctx context.Context, req *proto.DeleteStateRequest) (*proto.DeleteStateResponse, error) {
-	if s.connectClient != nil && (s.connectClient.Status() == internal.StatusConnected || s.connectClient.Status() == internal.StatusConnecting) {
+	if s.connectClient.ConnectionRunning() {
 		return nil, status.Errorf(codes.FailedPrecondition, "cannot clean state while connecting or connected, run 'netbird down' first.")
 	}
 

client/server/trace.go

@@ -62,10 +62,6 @@ func (s *Server) TracePacket(_ context.Context, req *proto.TracePacketRequest) (
 }
 
 func (s *Server) getPacketTracer() (packetTracer, *internal.Engine, error) {
-	if s.connectClient == nil {
-		return nil, nil, fmt.Errorf("connect client not initialized")
-	}
-
 	engine := s.connectClient.Engine()
 	if engine == nil {
 		return nil, nil, fmt.Errorf("engine not initialized")

client/status/status.go

@@ -98,6 +98,7 @@ type RelayStateOutputDetail struct {
 	URI       string `json:"uri" yaml:"uri"`
 	Available bool   `json:"available" yaml:"available"`
 	Error     string `json:"error" yaml:"error"`
+	Transport string `json:"transport,omitempty" yaml:"transport,omitempty"`
 }
 
 type RelayStateOutput struct {
@@ -219,7 +220,8 @@ func mapRelays(relays []*proto.RelayState) RelayStateOutput {
 			RelayStateOutputDetail{
 				URI:       relay.URI,
 				Available: available,
-				Error:     relay.GetError(),
+				Error:     relayErrorString(relay.GetError()),
+				Transport: relay.GetTransport(),
 			},
 		)
 
@@ -235,6 +237,12 @@ func mapRelays(relays []*proto.RelayState) RelayStateOutput {
 	}
 }
 
+// relayErrorString flattens a newline-joined aggregated relay error onto a
+// single line for status output.
+func relayErrorString(s string) string {
+	return strings.ReplaceAll(s, "\n", "; ")
+}
+
 func mapNSGroups(servers []*proto.NSGroupState) []NsServerGroupStateOutput {
 	mappedNSGroups := make([]NsServerGroupStateOutput, 0, len(servers))
 	for _, pbNsGroupServer := range servers {
@@ -441,6 +449,8 @@ func (o *OutputOverview) GeneralSummary(showURL bool, showRelays bool, showNameS
 					available = "Unavailable"
 					reason = fmt.Sprintf(", reason: %s", relay.Error)
 				}
+			} else if relay.Transport != "" {
+				available = fmt.Sprintf("%s via %s", available, relay.Transport)
 			}
 
 			relaysString += fmt.Sprintf("\n  [%s] is %s%s", relay.URI, available, reason)

client/status/status_test.go

@@ -647,3 +647,13 @@ func TestTimeAgo(t *testing.T) {
 		})
 	}
 }
+
+func TestMapRelaysTransport(t *testing.T) {
+	out := mapRelays([]*proto.RelayState{
+		{URI: "rels://relay.example:443", Available: true, Transport: "quic"},
+		{URI: "rels://relay2.example:443", Available: true, Transport: "ws"},
+	})
+	require.Len(t, out.Details, 2)
+	assert.Equal(t, "quic", out.Details[0].Transport)
+	assert.Equal(t, "ws", out.Details[1].Transport)
+}

combined/Dockerfile

@@ -2,4 +2,5 @@ FROM ubuntu:24.04
 RUN apt update && apt install -y ca-certificates && rm -fr /var/cache/apt
 ENTRYPOINT [ "/go/bin/netbird-server" ]
 CMD ["--config", "/etc/netbird/config.yaml"]
-COPY netbird-server /go/bin/netbird-server
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-server /go/bin/netbird-server

docker/build-env/README.md

@@ -0,0 +1,56 @@
+# Build environments
+
+Dockerfiles that pin the same toolchain CI uses, so a developer can
+reproduce a CI build locally without installing platform SDKs on their
+workstation. The version pins in each `Dockerfile` must stay in lockstep
+with `.github/workflows/`.
+
+## `android/`
+
+Mirrors `.github/workflows/mobile-build-validation.yml` (`android_build`
+job). Carries Go 1.25.5, Adopt JDK 11, Android cmdline-tools 8512546,
+NDK 23.1.7779620 and gomobile pinned at the CI commit. Use it to
+produce `netbird.aar` from `./client/android`:
+
+```bash
+docker build -t netbird/build-android docker/build-env/android
+docker run --rm -v "$PWD:/src" -w /src netbird/build-android \
+    gomobile bind \
+    -o netbird.aar \
+    -javapkg=io.netbird.gomobile \
+    -ldflags="-checklinkname=0 \
+              -X golang.zx2c4.com/wireguard/ipc.socketDirectory=/data/data/io.netbird.client/cache/wireguard \
+              -X github.com/netbirdio/netbird/version.version=local" \
+    ./client/android
+```
+
+To build the full Android APK, bind-mount the `android-client` repo as
+well and run its own `./gradlew assembleDebug` from inside the
+container (the gradle wrapper ships with `android-client`).
+
+## `windows-cross/`
+
+Cross-compiles Windows binaries from Linux using `mingw-w64`. Lets you
+verify that `GOOS=windows go build ./...` compiles cleanly without
+needing a Windows VM. Cannot run Windows tests — the `golang-test-windows`
+CI job executes on a native `windows-latest` runner with wintun.dll
+and PsExec, neither of which lives under Linux containers.
+
+```bash
+docker build -t netbird/build-windows docker/build-env/windows-cross
+docker run --rm -v "$PWD:/src" -w /src netbird/build-windows \
+    bash -c 'GOOS=windows GOARCH=amd64 go build ./...'
+```
+
+## What is NOT here
+
+- **iOS / macOS**: cannot legally run macOS in Docker (Apple EULA),
+  and Xcode is not redistributable. The `ios_build` CI job uses a
+  `macos-latest` GitHub runner; locally you need a real Mac.
+
+- **Native Windows tests**: see note above. The Linux+mingw image
+  builds, it does not execute Windows-host code paths
+  (registry, wintun, services, PsExec workflows).
+
+When CI version pins change, update the corresponding `ARG` lines in
+the Dockerfiles and the README's table of versions.

docker/build-env/android/Dockerfile

@@ -0,0 +1,86 @@
+# Android build environment.
+#
+# Mirrors the toolchain pinned by .github/workflows/mobile-build-validation.yml
+# so a `gomobile bind` against ./client/android in this image produces the
+# same netbird.aar that CI builds.
+#
+# Tooling versions (must stay in sync with the CI workflow):
+#   - Ubuntu 22.04 (matches the ubuntu-latest GitHub runner)
+#   - Go 1.25.5 (matches go.mod)
+#   - Adopt JDK 11 (matches actions/setup-java@v3 java-version: 11, distribution: adopt)
+#   - Android SDK cmdline-tools 8512546
+#   - Android NDK 23.1.7779620
+#   - gomobile commit v0.0.0-20251113184115-a159579294ab
+#
+# Usage (from the netbird repo root):
+#
+#   docker build -t netbird/build-android docker/build-env/android
+#
+#   # bind the netbird checkout in and run the same gomobile command CI runs
+#   docker run --rm -v "$PWD:/src" -w /src netbird/build-android \
+#       gomobile bind \
+#       -o netbird.aar \
+#       -javapkg=io.netbird.gomobile \
+#       -ldflags="-checklinkname=0 \
+#                 -X golang.zx2c4.com/wireguard/ipc.socketDirectory=/data/data/io.netbird.client/cache/wireguard \
+#                 -X github.com/netbirdio/netbird/version.version=local" \
+#       ./client/android
+#
+# To build the full APK, mount the android-client repo too and run
+# `./gradlew assembleDebug` from /android-client (this image carries
+# gradle's prerequisites JDK + Android SDK but not the gradle wrapper —
+# that ships with android-client).
+
+FROM ubuntu:22.04
+
+ARG DEBIAN_FRONTEND=noninteractive
+
+# Versions — bump in lockstep with .github/workflows/mobile-build-validation.yml.
+ARG GO_VERSION=1.25.5
+ARG ANDROID_CMDLINE_TOOLS_VERSION=8512546
+ARG ANDROID_NDK_VERSION=23.1.7779620
+ARG GOMOBILE_VERSION=v0.0.0-20251113184115-a159579294ab
+
+ENV ANDROID_HOME=/opt/android-sdk
+ENV ANDROID_NDK_HOME=${ANDROID_HOME}/ndk/${ANDROID_NDK_VERSION}
+ENV JAVA_HOME=/usr/lib/jvm/java-11-openjdk-amd64
+ENV GOPATH=/go
+ENV GOTOOLCHAIN=local
+ENV CGO_ENABLED=0
+ENV PATH=${GOPATH}/bin:/usr/local/go/bin:${ANDROID_HOME}/cmdline-tools/latest/bin:${ANDROID_HOME}/platform-tools:${JAVA_HOME}/bin:${PATH}
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        ca-certificates \
+        curl \
+        unzip \
+        git \
+        openjdk-11-jdk-headless \
+        build-essential \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install Go (matches go.mod). actions/setup-go fetches the same tarball.
+RUN curl -fsSL "https://go.dev/dl/go${GO_VERSION}.linux-amd64.tar.gz" \
+        | tar -C /usr/local -xz \
+    && go version
+
+# Install Android SDK command-line tools, accept licenses, install NDK.
+RUN mkdir -p "${ANDROID_HOME}/cmdline-tools" \
+    && curl -fsSL -o /tmp/cmdline.zip \
+        "https://dl.google.com/android/repository/commandlinetools-linux-${ANDROID_CMDLINE_TOOLS_VERSION}_latest.zip" \
+    && unzip -q /tmp/cmdline.zip -d "${ANDROID_HOME}/cmdline-tools" \
+    && mv "${ANDROID_HOME}/cmdline-tools/cmdline-tools" "${ANDROID_HOME}/cmdline-tools/latest" \
+    && rm /tmp/cmdline.zip \
+    && yes | sdkmanager --licenses > /dev/null \
+    && sdkmanager --install "ndk;${ANDROID_NDK_VERSION}" "platform-tools" > /dev/null
+
+# Install gomobile at the same commit CI pins. Don't run `gomobile init` here:
+# `init` resolves the NDK at runtime, do it on the first bind in the mounted
+# workspace so the cache lands on the host volume.
+RUN GOBIN=/usr/local/bin go install "golang.org/x/mobile/cmd/gomobile@${GOMOBILE_VERSION}" \
+    && gomobile version
+
+WORKDIR /src
+
+# Default entrypoint is a plain shell so the image is composable: callers pass
+# the full gomobile / gradle command they want to run.
+CMD ["/bin/bash"]

docker/build-env/windows-cross/Dockerfile

@@ -0,0 +1,63 @@
+# Windows-cross build environment.
+#
+# Cross-compiles Windows .exe targets from a Linux container using
+# mingw-w64. Mirrors the toolchain set used by
+# .github/workflows/golang-test-windows.yml insofar as that is possible
+# without a Windows kernel.
+#
+# IMPORTANT — what this image CAN do:
+#   - `GOOS=windows go build ./...` to validate that Windows builds compile
+#   - CGO Windows cross-compile via x86_64-w64-mingw32-gcc when CGO_ENABLED=1
+#     (matches CI's choco-installed mingw-w64)
+#
+# IMPORTANT — what this image CANNOT do:
+#   - Run Windows binaries (no Windows kernel under Docker on Linux).
+#   - Replicate the CI's `go test` runs which execute on a real
+#     windows-latest runner (wintun.dll, PsExec, registry, etc.).
+#     Use the CI for that or a native Windows VM.
+#
+# Usage (from the netbird repo root):
+#
+#   docker build -t netbird/build-windows docker/build-env/windows-cross
+#
+#   # Cross-compile a static client (.exe) from Linux:
+#   docker run --rm -v "$PWD:/src" -w /src netbird/build-windows \
+#       bash -c 'CGO_ENABLED=1 GOOS=windows GOARCH=amd64 \
+#           CC=x86_64-w64-mingw32-gcc CXX=x86_64-w64-mingw32-g++ \
+#           go build -o netbird.exe ./client'
+#
+#   # Just validate that everything *compiles* on Windows (no CGO):
+#   docker run --rm -v "$PWD:/src" -w /src netbird/build-windows \
+#       bash -c 'GOOS=windows GOARCH=amd64 go build ./...'
+#
+# Tooling versions (keep in sync with go.mod and any future explicit pin
+# documented in golang-test-windows.yml):
+#   - Ubuntu 22.04
+#   - Go 1.25.5 (matches go.mod)
+#   - mingw-w64 (Ubuntu package — pin further if drift becomes a problem)
+
+FROM ubuntu:22.04
+
+ARG DEBIAN_FRONTEND=noninteractive
+ARG GO_VERSION=1.25.5
+
+ENV GOPATH=/go
+ENV GOTOOLCHAIN=local
+ENV PATH=${GOPATH}/bin:/usr/local/go/bin:${PATH}
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        ca-certificates \
+        curl \
+        git \
+        build-essential \
+        mingw-w64 \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install Go (matches go.mod).
+RUN curl -fsSL "https://go.dev/dl/go${GO_VERSION}.linux-amd64.tar.gz" \
+        | tar -C /usr/local -xz \
+    && go version
+
+WORKDIR /src
+
+CMD ["/bin/bash"]

management/Dockerfile

@@ -2,4 +2,5 @@ FROM ubuntu:24.04
 RUN apt update && apt install -y ca-certificates && rm -fr /var/cache/apt
 ENTRYPOINT [ "/go/bin/netbird-mgmt","management"]
 CMD ["--log-file", "console"]
-COPY netbird-mgmt /go/bin/netbird-mgmt
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-mgmt /go/bin/netbird-mgmt

management/Dockerfile.debug

@@ -1,5 +0,0 @@
-FROM ubuntu:24.04
-RUN apt update && apt install -y ca-certificates && rm -fr /var/cache/apt
-ENTRYPOINT [ "/go/bin/netbird-mgmt","management","--log-level","debug"]
-CMD ["--log-file", "console"]
-COPY netbird-mgmt /go/bin/netbird-mgmt

management/internals/controllers/network_map/controller/controller.go

@@ -585,66 +585,66 @@ func (b *bufferAffectedUpdate) setTimer(d time.Duration, f func()) {
 	b.next.Reset(d)
 }
 
-func (c *Controller) GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, peer *nbpeer.Peer) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, int64, error) {
+func (c *Controller) GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, peerID string) (*types.NetworkMap, []*posture.Checks, int64, error) {
 	if isRequiresApproval {
 		network, err := c.repo.GetAccountNetwork(ctx, accountID)
 		if err != nil {
-			return nil, nil, nil, 0, err
+			return nil, nil, 0, err
 		}
 
 		emptyMap := &types.NetworkMap{
 			Network: network.Copy(),
 		}
-		return peer, emptyMap, nil, 0, nil
+		return emptyMap, nil, 0, nil
 	}
 
 	account, err := c.requestBuffer.GetAccountWithBackpressure(ctx, accountID)
 	if err != nil {
-		return nil, nil, nil, 0, err
+		return nil, nil, 0, err
 	}
 
 	account.InjectProxyPolicies(ctx)
 
 	approvedPeersMap, err := c.integratedPeerValidator.GetValidatedPeers(ctx, account.Id, maps.Values(account.Groups), maps.Values(account.Peers), account.Settings.Extra)
 	if err != nil {
-		return nil, nil, nil, 0, err
+		return nil, nil, 0, err
 	}
 
 	startPosture := time.Now()
-	postureChecks, err := c.getPeerPostureChecks(account, peer.ID)
+	postureChecks, err := c.getPeerPostureChecks(account, peerID)
 	if err != nil {
-		return nil, nil, nil, 0, err
+		return nil, nil, 0, err
 	}
 	log.WithContext(ctx).Debugf("getPeerPostureChecks took %s", time.Since(startPosture))
 
 	accountZones, err := c.repo.GetAccountZones(ctx, account.Id)
 	if err != nil {
 		log.WithContext(ctx).Errorf("failed to get account zones: %v", err)
-		return nil, nil, nil, 0, err
+		return nil, nil, 0, err
 	}
 
 	dnsDomain := c.GetDNSDomain(account.Settings)
 	peersCustomZone := account.GetPeersCustomZone(ctx, dnsDomain)
 
-	proxyNetworkMaps, err := c.proxyController.GetProxyNetworkMaps(ctx, account.Id, peer.ID, account.Peers)
+	proxyNetworkMaps, err := c.proxyController.GetProxyNetworkMaps(ctx, account.Id, peerID, account.Peers)
 	if err != nil {
 		log.WithContext(ctx).Errorf("failed to get proxy network maps: %v", err)
-		return nil, nil, nil, 0, err
+		return nil, nil, 0, err
 	}
 
 	resourcePolicies := account.GetResourcePoliciesMap()
 	routers := account.GetResourceRoutersMap()
 	groupIDToUserIDs := account.GetActiveGroupUsers()
-	networkMap := account.GetPeerNetworkMapFromComponents(ctx, peer.ID, peersCustomZone, accountZones, approvedPeersMap, resourcePolicies, routers, c.accountManagerMetrics, groupIDToUserIDs)
+	networkMap := account.GetPeerNetworkMapFromComponents(ctx, peerID, peersCustomZone, accountZones, approvedPeersMap, resourcePolicies, routers, c.accountManagerMetrics, groupIDToUserIDs)
 
-	proxyNetworkMap, ok := proxyNetworkMaps[peer.ID]
+	proxyNetworkMap, ok := proxyNetworkMaps[peerID]
 	if ok {
 		networkMap.Merge(proxyNetworkMap)
 	}
 
 	dnsFwdPort := computeForwarderPort(maps.Values(account.Peers), network_map.DnsForwarderPortMinVersion)
 
-	return peer, networkMap, postureChecks, dnsFwdPort, nil
+	return networkMap, postureChecks, dnsFwdPort, nil
 }
 
 // GetDNSDomain returns the configured dnsDomain

management/internals/controllers/network_map/interface.go

@@ -23,7 +23,7 @@ type Controller interface {
 	BufferUpdateAffectedPeers(ctx context.Context, accountID string, peerIDs []string, reason types.UpdateReason) error
 	UpdateAccountPeer(ctx context.Context, accountId string, peerId string) error
 	BufferUpdateAccountPeers(ctx context.Context, accountID string, reason types.UpdateReason) error
-	GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, p *nbpeer.Peer) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, int64, error)
+	GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, peerID string) (*types.NetworkMap, []*posture.Checks, int64, error)
 	GetDNSDomain(settings *types.Settings) string
 	StartWarmup(context.Context)
 	GetNetworkMap(ctx context.Context, peerID string) (*types.NetworkMap, error)

management/internals/controllers/network_map/interface_mock.go

@@ -127,21 +127,20 @@ func (mr *MockControllerMockRecorder) GetNetworkMap(ctx, peerID any) *gomock.Cal
 }
 
 // GetValidatedPeerWithMap mocks base method.
-func (m *MockController) GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, p *peer.Peer) (*peer.Peer, *types.NetworkMap, []*posture.Checks, int64, error) {
+func (m *MockController) GetValidatedPeerWithMap(ctx context.Context, isRequiresApproval bool, accountID string, peerID string) (*types.NetworkMap, []*posture.Checks, int64, error) {
 	m.ctrl.T.Helper()
-	ret := m.ctrl.Call(m, "GetValidatedPeerWithMap", ctx, isRequiresApproval, accountID, p)
-	ret0, _ := ret[0].(*peer.Peer)
-	ret1, _ := ret[1].(*types.NetworkMap)
-	ret2, _ := ret[2].([]*posture.Checks)
-	ret3, _ := ret[3].(int64)
-	ret4, _ := ret[4].(error)
-	return ret0, ret1, ret2, ret3, ret4
+	ret := m.ctrl.Call(m, "GetValidatedPeerWithMap", ctx, isRequiresApproval, accountID, peerID)
+	ret0, _ := ret[0].(*types.NetworkMap)
+	ret1, _ := ret[1].([]*posture.Checks)
+	ret2, _ := ret[2].(int64)
+	ret3, _ := ret[3].(error)
+	return ret0, ret1, ret2, ret3
 }
 
 // GetValidatedPeerWithMap indicates an expected call of GetValidatedPeerWithMap.
-func (mr *MockControllerMockRecorder) GetValidatedPeerWithMap(ctx, isRequiresApproval, accountID, p any) *gomock.Call {
+func (mr *MockControllerMockRecorder) GetValidatedPeerWithMap(ctx, isRequiresApproval, accountID, peerID any) *gomock.Call {
 	mr.mock.ctrl.T.Helper()
-	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetValidatedPeerWithMap", reflect.TypeOf((*MockController)(nil).GetValidatedPeerWithMap), ctx, isRequiresApproval, accountID, p)
+	return mr.mock.ctrl.RecordCallWithMethodType(mr.mock, "GetValidatedPeerWithMap", reflect.TypeOf((*MockController)(nil).GetValidatedPeerWithMap), ctx, isRequiresApproval, accountID, peerID)
 }
 
 // OnPeerConnected mocks base method.

management/internals/modules/peers/manager.go

@@ -242,7 +242,7 @@ func (m *managerImpl) CreateProxyPeer(ctx context.Context, accountID string, pee
 		},
 	}
 
-	_, _, _, err = m.accountManager.AddPeer(ctx, accountID, "", "", peer, true)
+	_, _, _, _, err = m.accountManager.AddPeer(ctx, accountID, "", "", peer, true)
 	if err != nil {
 		return fmt.Errorf("failed to create proxy peer: %w", err)
 	}

management/internals/shared/grpc/server.go

@@ -778,7 +778,7 @@ func (s *Server) Login(ctx context.Context, req *proto.EncryptedMessage) (*proto
 		sshKey = loginReq.GetPeerKeys().GetSshPubKey()
 	}
 
-	peer, netMap, postureChecks, err := s.accountManager.LoginPeer(ctx, types.PeerLogin{
+	peer, network, postureChecks, enableSSH, err := s.accountManager.LoginPeer(ctx, types.PeerLogin{
 		WireGuardPubKey: peerKey.String(),
 		SSHKey:          string(sshKey),
 		Meta:            peerMeta,
@@ -792,7 +792,7 @@ func (s *Server) Login(ctx context.Context, req *proto.EncryptedMessage) (*proto
 		return nil, mapError(ctx, err)
 	}
 
-	loginResp, err := s.prepareLoginResponse(ctx, peer, netMap, postureChecks)
+	loginResp, err := s.prepareLoginResponse(ctx, peer, network, postureChecks, enableSSH)
 	if err != nil {
 		log.WithContext(ctx).Warnf("failed preparing login response for peer %s: %s", peerKey, err)
 		return nil, status.Errorf(codes.Internal, "failed logging in peer")
@@ -895,7 +895,7 @@ func (s *Server) ExtendAuthSession(ctx context.Context, req *proto.EncryptedMess
 	}, nil
 }
 
-func (s *Server) prepareLoginResponse(ctx context.Context, peer *nbpeer.Peer, netMap *types.NetworkMap, postureChecks []*posture.Checks) (*proto.LoginResponse, error) {
+func (s *Server) prepareLoginResponse(ctx context.Context, peer *nbpeer.Peer, network *types.Network, postureChecks []*posture.Checks, enableSSH bool) (*proto.LoginResponse, error) {
 	var relayToken *Token
 	var err error
 	if s.config.Relay != nil && len(s.config.Relay.Addresses) > 0 {
@@ -914,7 +914,7 @@ func (s *Server) prepareLoginResponse(ctx context.Context, peer *nbpeer.Peer, ne
 	// if peer has reached this point then it has logged in
 	loginResp := &proto.LoginResponse{
 		NetbirdConfig: toNetbirdConfig(s.config, nil, relayToken, nil),
-		PeerConfig:    toPeerConfig(peer, netMap.Network, s.networkMapController.GetDNSDomain(settings), settings, s.config.HttpConfig, s.config.DeviceAuthorizationFlow, netMap.EnableSSH),
+		PeerConfig:    toPeerConfig(peer, network, s.networkMapController.GetDNSDomain(settings), settings, s.config.HttpConfig, s.config.DeviceAuthorizationFlow, enableSSH),
 		Checks:        toProtocolChecks(ctx, postureChecks),
 	}
 

management/server/account/manager.go

@@ -70,7 +70,7 @@ type Manager interface {
 	UpdatePeerIPv6(ctx context.Context, accountID, userID, peerID string, newIPv6 netip.Addr) error
 	GetNetworkMap(ctx context.Context, peerID string) (*types.NetworkMap, error)
 	GetPeerNetwork(ctx context.Context, peerID string) (*types.Network, error)
-	AddPeer(ctx context.Context, accountID, setupKey, userID string, p *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error)
+	AddPeer(ctx context.Context, accountID, setupKey, userID string, p *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error)
 	CreatePAT(ctx context.Context, accountID string, initiatorUserID string, targetUserID string, tokenName string, expiresIn int) (*types.PersonalAccessTokenGenerated, error)
 	DeletePAT(ctx context.Context, accountID string, initiatorUserID string, targetUserID string, tokenID string) error
 	GetPAT(ctx context.Context, accountID string, initiatorUserID string, targetUserID string, tokenID string) (*types.PersonalAccessToken, error)
@@ -109,7 +109,7 @@ type Manager interface {
 	GetPeer(ctx context.Context, accountID, peerID, userID string) (*nbpeer.Peer, error)
 	UpdateAccountSettings(ctx context.Context, accountID, userID string, newSettings *types.Settings) (*types.Settings, error)
 	UpdateAccountOnboarding(ctx context.Context, accountID, userID string, newOnboarding *types.AccountOnboarding) (*types.AccountOnboarding, error)
-	LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error)                       // used by peer gRPC API
+	LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error)                    // used by peer gRPC API
 	ExtendPeerSession(ctx context.Context, peerPubKey, userID string) (time.Time, error)                                                    // used by peer gRPC API for ExtendAuthSession
 	SyncPeer(ctx context.Context, sync types.PeerSync, accountID string) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, int64, error) // used by peer gRPC API
 	GetExternalCacheManager() ExternalCacheManager

management/server/account/manager_mock.go

@@ -80,14 +80,15 @@ func (mr *MockManagerMockRecorder) AccountExists(ctx, accountID interface{}) *go
 }
 
 // AddPeer mocks base method.
-func (m *MockManager) AddPeer(ctx context.Context, accountID, setupKey, userID string, p *peer.Peer, temporary bool) (*peer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (m *MockManager) AddPeer(ctx context.Context, accountID, setupKey, userID string, p *peer.Peer, temporary bool) (*peer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	m.ctrl.T.Helper()
 	ret := m.ctrl.Call(m, "AddPeer", ctx, accountID, setupKey, userID, p, temporary)
 	ret0, _ := ret[0].(*peer.Peer)
-	ret1, _ := ret[1].(*types.NetworkMap)
+	ret1, _ := ret[1].(*types.Network)
 	ret2, _ := ret[2].([]*posture.Checks)
-	ret3, _ := ret[3].(error)
-	return ret0, ret1, ret2, ret3
+	ret3, _ := ret[3].(bool)
+	ret4, _ := ret[4].(error)
+	return ret0, ret1, ret2, ret3, ret4
 }
 
 // AddPeer indicates an expected call of AddPeer.
@@ -1289,14 +1290,15 @@ func (mr *MockManagerMockRecorder) ListUsers(ctx, accountID interface{}) *gomock
 }
 
 // LoginPeer mocks base method.
-func (m *MockManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*peer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (m *MockManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*peer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	m.ctrl.T.Helper()
 	ret := m.ctrl.Call(m, "LoginPeer", ctx, login)
 	ret0, _ := ret[0].(*peer.Peer)
-	ret1, _ := ret[1].(*types.NetworkMap)
+	ret1, _ := ret[1].(*types.Network)
 	ret2, _ := ret[2].([]*posture.Checks)
-	ret3, _ := ret[3].(error)
-	return ret0, ret1, ret2, ret3
+	ret3, _ := ret[3].(bool)
+	ret4, _ := ret[4].(error)
+	return ret0, ret1, ret2, ret3, ret4
 }
 
 // LoginPeer indicates an expected call of LoginPeer.

management/server/account_test.go

@@ -84,7 +84,7 @@ func verifyCanAddPeerToAccount(t *testing.T, manager nbAccount.Manager, account
 		setupKey = key.Key
 	}
 
-	_, _, _, err := manager.AddPeer(context.Background(), "", setupKey, userID, peer, false)
+	_, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey, userID, peer, false)
 	if err != nil {
 		t.Error("expected to add new peer successfully after creating new account, but failed", err)
 	}
@@ -1092,7 +1092,7 @@ func TestAccountManager_AddPeer(t *testing.T) {
 	}
 	expectedPeerKey := key.PublicKey().String()
 
-	peer, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  expectedPeerKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
 	}, false)
@@ -1156,7 +1156,7 @@ func TestAccountManager_AddPeerWithUserID(t *testing.T) {
 	expectedPeerKey := key.PublicKey().String()
 	expectedUserID := userID
 
-	peer, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  expectedPeerKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
 	}, false)
@@ -1504,7 +1504,7 @@ func TestAccountManager_DeletePeer(t *testing.T) {
 
 	peerKey := key.PublicKey().String()
 
-	peer, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: peerKey},
 	}, false)
@@ -1826,7 +1826,7 @@ func TestDefaultAccountManager_UpdatePeer_PeerLoginExpiration(t *testing.T) {
 
 	key, err := wgtypes.GenerateKey()
 	require.NoError(t, err, "unable to generate WireGuard key")
-	peer, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:                    key.PublicKey().String(),
 		Meta:                   nbpeer.PeerSystemMeta{Hostname: "test-peer"},
 		LoginExpirationEnabled: true,
@@ -1882,7 +1882,7 @@ func TestDefaultAccountManager_MarkPeerConnected_PeerLoginExpiration(t *testing.
 
 	key, err := wgtypes.GenerateKey()
 	require.NoError(t, err, "unable to generate WireGuard key")
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:                    key.PublicKey().String(),
 		Meta:                   nbpeer.PeerSystemMeta{Hostname: "test-peer"},
 		LoginExpirationEnabled: true,
@@ -1927,7 +1927,7 @@ func TestDefaultAccountManager_OnPeerDisconnected_LastSeenCheck(t *testing.T) {
 	require.NoError(t, err, "unable to generate WireGuard key")
 	peerPubKey := key.PublicKey().String()
 
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  peerPubKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer"},
 	}, false)
@@ -2017,7 +2017,7 @@ func TestDefaultAccountManager_MarkPeerConnected_ConcurrentRace(t *testing.T) {
 	require.NoError(t, err, "unable to generate WireGuard key")
 	peerPubKey := key.PublicKey().String()
 
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  peerPubKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: "race-peer"},
 	}, false)
@@ -2080,7 +2080,7 @@ func TestDefaultAccountManager_UpdateAccountSettings_PeerLoginExpiration(t *test
 
 	key, err := wgtypes.GenerateKey()
 	require.NoError(t, err, "unable to generate WireGuard key")
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:                    key.PublicKey().String(),
 		Meta:                   nbpeer.PeerSystemMeta{Hostname: "test-peer"},
 		LoginExpirationEnabled: true,
@@ -3276,7 +3276,7 @@ func setupNetworkMapTest(t *testing.T) (*DefaultAccountManager, *update_channel.
 		}
 		expectedPeerKey := key.PublicKey().String()
 
-		peer, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+		peer, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 			Key:  expectedPeerKey,
 			Meta: nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
 			Status: &nbpeer.PeerStatus{
@@ -3444,7 +3444,7 @@ func BenchmarkLoginPeer_ExistingPeer(b *testing.B) {
 			b.ResetTimer()
 			start := time.Now()
 			for i := 0; i < b.N; i++ {
-				_, _, _, err := manager.LoginPeer(context.Background(), types.PeerLogin{
+				_, _, _, _, err := manager.LoginPeer(context.Background(), types.PeerLogin{
 					WireGuardPubKey: account.Peers["peer-1"].Key,
 					SSHKey:          "someKey",
 					Meta:            nbpeer.PeerSystemMeta{Hostname: strconv.Itoa(i)},
@@ -3513,7 +3513,7 @@ func BenchmarkLoginPeer_NewPeer(b *testing.B) {
 			b.ResetTimer()
 			start := time.Now()
 			for i := 0; i < b.N; i++ {
-				_, _, _, err := manager.LoginPeer(context.Background(), types.PeerLogin{
+				_, _, _, _, err := manager.LoginPeer(context.Background(), types.PeerLogin{
 					WireGuardPubKey: "some-new-key" + strconv.Itoa(i),
 					SSHKey:          "someKey",
 					Meta:            nbpeer.PeerSystemMeta{Hostname: strconv.Itoa(i)},
@@ -3908,13 +3908,13 @@ func TestDefaultAccountManager_UpdatePeerIP(t *testing.T) {
 	key2, err := wgtypes.GenerateKey()
 	require.NoError(t, err, "unable to generate WireGuard key")
 
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-1"},
 	}, false)
 	require.NoError(t, err, "unable to add peer1")
 
-	peer2, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer2, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)

management/server/affected_peers_test.go

@@ -1663,7 +1663,7 @@ func addPeerToAccount(t *testing.T, manager *DefaultAccountManager, _, setupKeyK
 	key, err := wgtypes.GeneratePrivateKey()
 	require.NoError(t, err)
 
-	peer, _, _, err := manager.AddPeer(context.Background(), "", setupKeyKey, "", &nbpeer.Peer{
+	peer, _, _, _, err := manager.AddPeer(context.Background(), "", setupKeyKey, "", &nbpeer.Peer{
 		Key:  key.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: key.PublicKey().String()},
 	}, false)

management/server/dns_test.go

@@ -298,11 +298,11 @@ func initTestDNSAccount(t *testing.T, am *DefaultAccountManager) (*types.Account
 		return nil, err
 	}
 
-	savedPeer1, _, _, err := am.AddPeer(context.Background(), "", "", dnsAdminUserID, peer1, false)
+	savedPeer1, _, _, _, err := am.AddPeer(context.Background(), "", "", dnsAdminUserID, peer1, false)
 	if err != nil {
 		return nil, err
 	}
-	_, _, _, err = am.AddPeer(context.Background(), "", "", dnsAdminUserID, peer2, false)
+	_, _, _, _, err = am.AddPeer(context.Background(), "", "", dnsAdminUserID, peer2, false)
 	if err != nil {
 		return nil, err
 	}

management/server/group_ipv6_test.go

@@ -55,7 +55,7 @@ func TestGroupIPv6Assignment(t *testing.T) {
 	key, err := wgtypes.GeneratePrivateKey()
 	require.NoError(t, err)
 
-	peer, _, _, err := am.AddPeer(ctx, "", setupKey.Key, "", &nbpeer.Peer{
+	peer, _, _, _, err := am.AddPeer(ctx, "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  key.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "ipv6-test-host"},
 	}, false)

management/server/http/handlers/peers/peers_handler.go

@@ -479,7 +479,7 @@ func (h *Handler) CreateTemporaryAccess(w http.ResponseWriter, r *http.Request)
 		return
 	}
 
-	peer, _, _, err := h.accountManager.AddPeer(r.Context(), userAuth.AccountId, "", userAuth.UserId, newPeer, true)
+	peer, _, _, _, err := h.accountManager.AddPeer(r.Context(), userAuth.AccountId, "", userAuth.UserId, newPeer, true)
 	if err != nil {
 		util.WriteError(r.Context(), err, w)
 		return

management/server/management_proto_test.go

@@ -728,7 +728,7 @@ func Test_LoginPerformance(t *testing.T) {
 						}
 
 						login := func() error {
-							_, _, _, err = am.LoginPeer(context.Background(), peerLogin)
+							_, _, _, _, err = am.LoginPeer(context.Background(), peerLogin)
 							if err != nil {
 								t.Logf("failed to login peer: %v", err)
 								return err
@@ -746,7 +746,7 @@ func Test_LoginPerformance(t *testing.T) {
 
 						go func(peerLogin types.PeerLogin, counterStart *int32) {
 							defer wgPeer.Done()
-							_, _, _, err = am.LoginPeer(context.Background(), peerLogin)
+							_, _, _, _, err = am.LoginPeer(context.Background(), peerLogin)
 							if err != nil {
 								t.Logf("failed to login peer: %v", err)
 								return

management/server/mock_server/account_mock.go

@@ -45,7 +45,7 @@ type MockAccountManager struct {
 	DeletePeerFunc                        func(ctx context.Context, accountID, peerKey, userID string) error
 	GetNetworkMapFunc                     func(ctx context.Context, peerKey string) (*types.NetworkMap, error)
 	GetPeerNetworkFunc                    func(ctx context.Context, peerKey string) (*types.Network, error)
-	AddPeerFunc                           func(ctx context.Context, accountID string, setupKey string, userId string, peer *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error)
+	AddPeerFunc                           func(ctx context.Context, accountID string, setupKey string, userId string, peer *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error)
 	GetGroupFunc                          func(ctx context.Context, accountID, groupID, userID string) (*types.Group, error)
 	GetAllGroupsFunc                      func(ctx context.Context, accountID, userID string) ([]*types.Group, error)
 	GetGroupByNameFunc                    func(ctx context.Context, groupName, accountID, userID string) (*types.Group, error)
@@ -98,7 +98,7 @@ type MockAccountManager struct {
 	SaveDNSSettingsFunc                   func(ctx context.Context, accountID, userID string, dnsSettingsToSave *types.DNSSettings) error
 	GetPeerFunc                           func(ctx context.Context, accountID, peerID, userID string) (*nbpeer.Peer, error)
 	UpdateAccountSettingsFunc             func(ctx context.Context, accountID, userID string, newSettings *types.Settings) (*types.Settings, error)
-	LoginPeerFunc                         func(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error)
+	LoginPeerFunc                         func(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error)
 	ExtendPeerSessionFunc                 func(ctx context.Context, peerPubKey, userID string) (time.Time, error)
 	SyncPeerFunc                          func(ctx context.Context, sync types.PeerSync, accountID string) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, int64, error)
 	InviteUserFunc                        func(ctx context.Context, accountID string, initiatorUserID string, targetUserEmail string) error
@@ -424,11 +424,11 @@ func (am *MockAccountManager) AddPeer(
 	userId string,
 	peer *nbpeer.Peer,
 	temporary bool,
-) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	if am.AddPeerFunc != nil {
 		return am.AddPeerFunc(ctx, accountID, setupKey, userId, peer, temporary)
 	}
-	return nil, nil, nil, status.Errorf(codes.Unimplemented, "method AddPeer is not implemented")
+	return nil, nil, nil, false, status.Errorf(codes.Unimplemented, "method AddPeer is not implemented")
 }
 
 // GetGroupByName mock implementation of GetGroupByName from server.AccountManager interface
@@ -862,11 +862,11 @@ func (am *MockAccountManager) UpdateAccountSettings(ctx context.Context, account
 }
 
 // LoginPeer mocks LoginPeer of the AccountManager interface
-func (am *MockAccountManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (am *MockAccountManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	if am.LoginPeerFunc != nil {
 		return am.LoginPeerFunc(ctx, login)
 	}
-	return nil, nil, nil, status.Errorf(codes.Unimplemented, "method LoginPeer is not implemented")
+	return nil, nil, nil, false, status.Errorf(codes.Unimplemented, "method LoginPeer is not implemented")
 }
 
 // ExtendPeerSession mocks ExtendPeerSession of the AccountManager interface

management/server/nameserver_test.go

@@ -896,11 +896,11 @@ func initTestNSAccount(t *testing.T, am *DefaultAccountManager) (*types.Account,
 		return nil, err
 	}
 
-	_, _, _, err = am.AddPeer(context.Background(), "", "", userID, peer1, false)
+	_, _, _, _, err = am.AddPeer(context.Background(), "", "", userID, peer1, false)
 	if err != nil {
 		return nil, err
 	}
-	_, _, _, err = am.AddPeer(context.Background(), "", "", userID, peer2, false)
+	_, _, _, _, err = am.AddPeer(context.Background(), "", "", userID, peer2, false)
 	if err != nil {
 		return nil, err
 	}

management/server/peer.go

@@ -718,10 +718,10 @@ func (am *DefaultAccountManager) handleSetupKeyAddedPeer(ctx context.Context, en
 // to it. We also add the User ID to the peer metadata to identify registrant. If no userID provided, then fail with status.PermissionDenied
 // Each new Peer will be assigned a new next net.IP from the Account.Network and Account.Network.LastIP will be updated (IP's are not reused).
 // The peer property is just a placeholder for the Peer properties to pass further
-func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKey, userID string, peer *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKey, userID string, peer *nbpeer.Peer, temporary bool) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	if setupKey == "" && userID == "" && !peer.ProxyMeta.Embedded {
 		// no auth method provided => reject access
-		return nil, nil, nil, status.Errorf(status.Unauthenticated, "no peer auth method provided, please use a setup key or interactive SSO login")
+		return nil, nil, nil, false, status.Errorf(status.Unauthenticated, "no peer auth method provided, please use a setup key or interactive SSO login")
 	}
 
 	upperKey := strings.ToUpper(setupKey)
@@ -737,7 +737,7 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 	// The connecting peer should be able to recover with a retry.
 	_, err := am.Store.GetPeerByPeerPubKey(ctx, store.LockingStrengthNone, peer.Key)
 	if err == nil {
-		return nil, nil, nil, status.Errorf(status.PreconditionFailed, "peer has been already registered")
+		return nil, nil, nil, false, status.Errorf(status.PreconditionFailed, "peer has been already registered")
 	}
 
 	opEvent := &activity.Event{
@@ -748,7 +748,7 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 
 	peerAddConfig, err := am.processPeerAddAuth(ctx, accountID, userID, encodedHashedKey, peer, temporary, addedByUser, addedBySetupKey, opEvent)
 	if err != nil {
-		return nil, nil, nil, err
+		return nil, nil, nil, false, err
 	}
 	accountID = peerAddConfig.AccountID
 	ephemeral := peerAddConfig.Ephemeral
@@ -763,7 +763,7 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 	}
 
 	if err := domain.ValidateDomainsList(peer.ExtraDNSLabels); err != nil {
-		return nil, nil, nil, status.Errorf(status.InvalidArgument, "invalid extra DNS labels: %v", err)
+		return nil, nil, nil, false, status.Errorf(status.InvalidArgument, "invalid extra DNS labels: %v", err)
 	}
 
 	registrationTime := time.Now().UTC()
@@ -789,7 +789,7 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 	}
 	settings, err := am.Store.GetAccountSettings(ctx, store.LockingStrengthNone, accountID)
 	if err != nil {
-		return nil, nil, nil, fmt.Errorf("failed to get account settings: %w", err)
+		return nil, nil, nil, false, fmt.Errorf("failed to get account settings: %w", err)
 	}
 
 	if am.geo != nil && newPeer.Location.ConnectionIP != nil {
@@ -807,30 +807,30 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 
 	network, err := am.Store.GetAccountNetwork(ctx, store.LockingStrengthNone, accountID)
 	if err != nil {
-		return nil, nil, nil, fmt.Errorf("failed getting network: %w", err)
+		return nil, nil, nil, false, fmt.Errorf("failed getting network: %w", err)
 	}
 
 	maxAttempts := 10
 	for attempt := 1; attempt <= maxAttempts; attempt++ {
 		netPrefix, err := netip.ParsePrefix(network.Net.String())
 		if err != nil {
-			return nil, nil, nil, fmt.Errorf("parse network prefix: %w", err)
+			return nil, nil, nil, false, fmt.Errorf("parse network prefix: %w", err)
 		}
 		freeIP, err := types.AllocateRandomPeerIP(netPrefix)
 		if err != nil {
-			return nil, nil, nil, fmt.Errorf("failed to get free IP: %w", err)
+			return nil, nil, nil, false, fmt.Errorf("failed to get free IP: %w", err)
 		}
 
 		var freeLabel string
 		if ephemeral || attempt > 1 {
 			freeLabel, err = getPeerIPDNSLabel(freeIP, peer.Meta.Hostname)
 			if err != nil {
-				return nil, nil, nil, fmt.Errorf("failed to get free DNS label: %w", err)
+				return nil, nil, nil, false, fmt.Errorf("failed to get free DNS label: %w", err)
 			}
 		} else {
 			freeLabel, err = nbdns.GetParsedDomainLabel(peer.Meta.Hostname)
 			if err != nil {
-				return nil, nil, nil, fmt.Errorf("failed to get free DNS label: %w", err)
+				return nil, nil, nil, false, fmt.Errorf("failed to get free DNS label: %w", err)
 			}
 		}
 		newPeer.DNSLabel = freeLabel
@@ -852,11 +852,11 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 			if allocate {
 				v6Prefix, err := netip.ParsePrefix(network.NetV6.String())
 				if err != nil {
-					return nil, nil, nil, fmt.Errorf("parse IPv6 prefix: %w", err)
+					return nil, nil, nil, false, fmt.Errorf("parse IPv6 prefix: %w", err)
 				}
 				freeIPv6, err := types.AllocateRandomPeerIPv6(v6Prefix)
 				if err != nil {
-					return nil, nil, nil, fmt.Errorf("allocate peer IPv6: %w", err)
+					return nil, nil, nil, false, fmt.Errorf("allocate peer IPv6: %w", err)
 				}
 				newPeer.IPv6 = freeIPv6
 			}
@@ -929,10 +929,10 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 			continue
 		}
 
-		return nil, nil, nil, fmt.Errorf("failed to add peer to database: %w", err)
+		return nil, nil, nil, false, fmt.Errorf("failed to add peer to database: %w", err)
 	}
 	if newPeer == nil {
-		return nil, nil, nil, fmt.Errorf("new peer is nil")
+		return nil, nil, nil, false, fmt.Errorf("new peer is nil")
 	}
 
 	opEvent.TargetID = newPeer.ID
@@ -940,7 +940,8 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 	if !addedByUser {
 		opEvent.Meta["setup_key_name"] = peerAddConfig.SetupKeyName
 	}
-	if newPeer.Status != nil && newPeer.Status.RequiresApproval {
+	requiresApproval := newPeer.Status != nil && newPeer.Status.RequiresApproval
+	if requiresApproval {
 		opEvent.Meta["pending_approval"] = true
 	}
 
@@ -948,18 +949,18 @@ func (am *DefaultAccountManager) AddPeer(ctx context.Context, accountID, setupKe
 		am.StoreEvent(ctx, opEvent.InitiatorID, opEvent.TargetID, opEvent.AccountID, opEvent.Activity, opEvent.Meta)
 	}
 
-	p, nmap, pc, _, err := am.networkMapController.GetValidatedPeerWithMap(ctx, false, accountID, newPeer)
+	network, postureChecks, enableSSH, err := getPeerLoginInfo(ctx, am.Store, accountID, newPeer, !requiresApproval)
 	if err != nil {
-		return p, nmap, pc, err
+		return nil, nil, nil, false, err
 	}
 
 	changedPeerIDs := []string{newPeer.ID}
-	affectedPeerIDs := affectedPeerIDsFromNetworkMap(nmap, newPeer.ID)
+	affectedPeerIDs := am.resolveAffectedPeersForPeerChanges(ctx, am.Store, accountID, changedPeerIDs)
 	if err := am.networkMapController.OnPeersAdded(ctx, accountID, changedPeerIDs, affectedPeerIDs); err != nil {
 		log.WithContext(ctx).Errorf("failed to update network map cache for peer %s: %v", newPeer.ID, err)
 	}
 
-	return p, nmap, pc, nil
+	return newPeer, network, postureChecks, enableSSH, nil
 }
 
 func getPeerIPDNSLabel(ip netip.Addr, peerHostName string) (string, error) {
@@ -1041,7 +1042,7 @@ func (am *DefaultAccountManager) SyncPeer(ctx context.Context, sync types.PeerSy
 		return nil, nil, nil, 0, err
 	}
 
-	resPeer, nmap, resPostureChecks, dnsFwdPort, err := am.networkMapController.GetValidatedPeerWithMap(ctx, peerNotValid, accountID, peer)
+	nmap, resPostureChecks, dnsFwdPort, err := am.networkMapController.GetValidatedPeerWithMap(ctx, peerNotValid, accountID, peer.ID)
 	if err != nil {
 		return nil, nil, nil, 0, err
 	}
@@ -1054,7 +1055,7 @@ func (am *DefaultAccountManager) SyncPeer(ctx context.Context, sync types.PeerSy
 		}
 	}
 
-	return resPeer, nmap, resPostureChecks, dnsFwdPort, nil
+	return peer, nmap, resPostureChecks, dnsFwdPort, nil
 }
 
 // syncPeerAffectedPeers resolves the peers affected by a SyncPeer change. The
@@ -1085,7 +1086,7 @@ func (am *DefaultAccountManager) markConnectedAffectedPeers(ctx context.Context,
 	return affectedPeerIDsFromNetworkMap(nmap, peerID)
 }
 
-func (am *DefaultAccountManager) handlePeerLoginNotFound(ctx context.Context, login types.PeerLogin, err error) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (am *DefaultAccountManager) handlePeerLoginNotFound(ctx context.Context, login types.PeerLogin, err error) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	if errStatus, ok := status.FromError(err); ok && errStatus.Type() == status.NotFound {
 		// we couldn't find this peer by its public key which can mean that peer hasn't been registered yet.
 		// Try registering it.
@@ -1101,12 +1102,12 @@ func (am *DefaultAccountManager) handlePeerLoginNotFound(ctx context.Context, lo
 	}
 
 	log.WithContext(ctx).Errorf("failed while logging in peer %s: %v", login.WireGuardPubKey, err)
-	return nil, nil, nil, status.Errorf(status.Internal, "failed while logging in peer")
+	return nil, nil, nil, false, status.Errorf(status.Internal, "failed while logging in peer")
 }
 
 // LoginPeer logs in or registers a peer.
 // If peer doesn't exist the function checks whether a setup key or a user is present and registers a new peer if so.
-func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.NetworkMap, []*posture.Checks, error) {
+func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.PeerLogin) (*nbpeer.Peer, *types.Network, []*posture.Checks, bool, error) {
 	accountID, err := am.Store.GetAccountIDByPeerPubKey(ctx, login.WireGuardPubKey)
 	if err != nil {
 		return am.handlePeerLoginNotFound(ctx, login, err)
@@ -1118,20 +1119,17 @@ func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.Peer
 	if login.UserID == "" {
 		err = am.checkIFPeerNeedsLoginWithoutLock(ctx, accountID, login)
 		if err != nil {
-			return nil, nil, nil, err
+			return nil, nil, nil, false, err
 		}
 	}
 
 	var peer *nbpeer.Peer
-	var updateRemotePeers bool
-	var isPeerUpdated bool
-	var ipv6CapabilityChanged bool
-	var postureChecks []*posture.Checks
+	var shouldStorePeer bool
 	var peerGroupIDs []string
 
 	settings, err := am.Store.GetAccountSettings(ctx, store.LockingStrengthNone, accountID)
 	if err != nil {
-		return nil, nil, nil, err
+		return nil, nil, nil, false, err
 	}
 
 	err = am.Store.ExecuteInTransaction(ctx, func(transaction store.Store) error {
@@ -1140,9 +1138,6 @@ func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.Peer
 			return err
 		}
 
-		// this flag prevents unnecessary calls to the persistent store.
-		shouldStorePeer := false
-
 		if login.UserID != "" {
 			if peer.UserID != login.UserID {
 				log.Warnf("user mismatch when logging in peer %s: peer user %s, login user %s ", peer.ID, peer.UserID, login.UserID)
@@ -1156,7 +1151,6 @@ func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.Peer
 
 			if changed {
 				shouldStorePeer = true
-				updateRemotePeers = true
 			}
 		}
 
@@ -1165,23 +1159,9 @@ func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.Peer
 			return err
 		}
 
-		oldHasIPv6Cap := peer.HasCapability(nbpeer.PeerCapabilityIPv6Overlay)
-		isPeerUpdated, _ = peer.UpdateMetaIfNew(login.Meta)
-		ipv6CapabilityChanged = oldHasIPv6Cap != peer.HasCapability(nbpeer.PeerCapabilityIPv6Overlay)
-		if isPeerUpdated {
-			am.metrics.AccountManagerMetrics().CountPeerMetUpdate()
-			shouldStorePeer = true
-
-			postureChecks, err = getPeerPostureChecks(ctx, transaction, accountID, peer.ID)
-			if err != nil {
-				return err
-			}
-		}
-
 		if peer.SSHKey != login.SSHKey {
 			peer.SSHKey = login.SSHKey
 			shouldStorePeer = true
-			updateRemotePeers = true
 		}
 
 		if !peer.AllowExtraDNSLabels && len(login.ExtraDNSLabels) > 0 {
@@ -1197,28 +1177,28 @@ func (am *DefaultAccountManager) LoginPeer(ctx context.Context, login types.Peer
 		return nil
 	})
 	if err != nil {
-		return nil, nil, nil, err
+		return nil, nil, nil, false, err
 	}
 
 	isRequiresApproval, isStatusChanged, err := am.integratedPeerValidator.IsNotValidPeer(ctx, accountID, peer, peerGroupIDs, settings.Extra)
 	if err != nil {
-		return nil, nil, nil, err
+		return nil, nil, nil, false, err
 	}
 
-	p, nmap, pc, _, err := am.networkMapController.GetValidatedPeerWithMap(ctx, isRequiresApproval, accountID, peer)
+	network, postureChecks, enableSSH, err := getPeerLoginInfo(ctx, am.Store, accountID, peer, !isRequiresApproval)
 	if err != nil {
-		return nil, nil, nil, err
+		return nil, nil, nil, false, err
 	}
 
-	if updateRemotePeers || isStatusChanged || ipv6CapabilityChanged || (isPeerUpdated && len(postureChecks) > 0) {
+	if isStatusChanged || shouldStorePeer {
 		changedPeerIDs := []string{peer.ID}
-		affectedPeerIDs := am.syncPeerAffectedPeers(ctx, accountID, peer.ID, nmap, isRequiresApproval, isPeerUpdated, len(postureChecks) > 0)
+		affectedPeerIDs := am.resolveAffectedPeersForPeerChanges(ctx, am.Store, accountID, changedPeerIDs)
 		if err = am.networkMapController.OnPeersUpdated(ctx, accountID, changedPeerIDs, affectedPeerIDs); err != nil {
-			return nil, nil, nil, fmt.Errorf("notify network map controller of peer update: %w", err)
+			return nil, nil, nil, false, fmt.Errorf("notify network map controller of peer update: %w", err)
 		}
 	}
 
-	return p, nmap, pc, nil
+	return peer, network, postureChecks, enableSSH, nil
 }
 
 // ExtendPeerSession refreshes the peer's SSO session deadline by updating
@@ -1294,6 +1274,50 @@ func (am *DefaultAccountManager) ExtendPeerSession(ctx context.Context, peerPubK
 	return refreshed.SessionExpiresAt(settings.PeerLoginExpirationEnabled, settings.PeerLoginExpiration), nil
 }
 
+// getPeerLoginInfo computes the login/register response data (network, posture
+// checks, SSH) from the store without building the peer's full network map.
+func getPeerLoginInfo(ctx context.Context, transaction store.Store, accountID string, peer *nbpeer.Peer, isValid bool) (*types.Network, []*posture.Checks, bool, error) {
+	network, err := transaction.GetAccountNetwork(ctx, store.LockingStrengthNone, accountID)
+	if err != nil {
+		return nil, nil, false, fmt.Errorf("get account network: %w", err)
+	}
+
+	if !isValid {
+		return network, nil, false, nil
+	}
+
+	postureChecks, err := getPeerPostureChecks(ctx, transaction, accountID, peer.ID)
+	if err != nil {
+		return nil, nil, false, err
+	}
+
+	enableSSH, err := isPeerSSHEnabled(ctx, transaction, accountID, peer)
+	if err != nil {
+		return nil, nil, false, err
+	}
+
+	return network, postureChecks, enableSSH, nil
+}
+
+func isPeerSSHEnabled(ctx context.Context, transaction store.Store, accountID string, peer *nbpeer.Peer) (bool, error) {
+	policies, err := transaction.GetAccountPolicies(ctx, store.LockingStrengthNone, accountID)
+	if err != nil {
+		return false, err
+	}
+
+	peerGroups, err := transaction.GetPeerGroups(ctx, store.LockingStrengthNone, accountID, peer.ID)
+	if err != nil {
+		return false, err
+	}
+
+	peerGroupIDs := make(map[string]struct{}, len(peerGroups))
+	for _, g := range peerGroups {
+		peerGroupIDs[g.ID] = struct{}{}
+	}
+
+	return types.PeerSSHEnabledFromPolicies(policies, peer.ID, peerGroupIDs, peer.SSHEnabled), nil
+}
+
 // getPeerPostureChecks returns the posture checks for the peer.
 func getPeerPostureChecks(ctx context.Context, transaction store.Store, accountID, peerID string) ([]*posture.Checks, error) {
 	policies, err := transaction.GetAccountPolicies(ctx, store.LockingStrengthNone, accountID)

management/server/peer_test.go

@@ -205,7 +205,7 @@ func testGetNetworkMapGeneral(t *testing.T) {
 		return
 	}
 
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-1"},
 	}, false)
@@ -219,7 +219,7 @@ func testGetNetworkMapGeneral(t *testing.T) {
 		t.Fatal(err)
 		return
 	}
-	_, _, _, err = manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)
@@ -278,7 +278,7 @@ func TestAccountManager_GetNetworkMapWithPolicy(t *testing.T) {
 		return
 	}
 
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-1"},
 	}, false)
@@ -292,7 +292,7 @@ func TestAccountManager_GetNetworkMapWithPolicy(t *testing.T) {
 		t.Fatal(err)
 		return
 	}
-	peer2, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer2, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)
@@ -454,7 +454,7 @@ func TestAccountManager_GetPeerNetwork(t *testing.T) {
 		return
 	}
 
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-1"},
 	}, false)
@@ -468,7 +468,7 @@ func TestAccountManager_GetPeerNetwork(t *testing.T) {
 		t.Fatal(err)
 		return
 	}
-	_, _, _, err = manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)
@@ -526,7 +526,7 @@ func TestDefaultAccountManager_GetPeer(t *testing.T) {
 		return
 	}
 
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", "", someUser, &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", "", someUser, &nbpeer.Peer{
 		Key:  peerKey1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)
@@ -542,7 +542,7 @@ func TestDefaultAccountManager_GetPeer(t *testing.T) {
 	}
 
 	// the second peer added with a setup key
-	peer2, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
+	peer2, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", &nbpeer.Peer{
 		Key:  peerKey2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 	}, false)
@@ -698,7 +698,7 @@ func TestDefaultAccountManager_GetPeers(t *testing.T) {
 				return
 			}
 
-			_, _, _, err = manager.AddPeer(context.Background(), "", "", someUser, &nbpeer.Peer{
+			_, _, _, _, err = manager.AddPeer(context.Background(), "", "", someUser, &nbpeer.Peer{
 				Key:  peerKey1.PublicKey().String(),
 				Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-1"},
 			}, false)
@@ -707,7 +707,7 @@ func TestDefaultAccountManager_GetPeers(t *testing.T) {
 				return
 			}
 
-			_, _, _, err = manager.AddPeer(context.Background(), "", "", adminUser, &nbpeer.Peer{
+			_, _, _, _, err = manager.AddPeer(context.Background(), "", "", adminUser, &nbpeer.Peer{
 				Key:  peerKey2.PublicKey().String(),
 				Meta: nbpeer.PeerSystemMeta{Hostname: "test-peer-2"},
 			}, false)
@@ -1332,7 +1332,7 @@ func Test_RegisterPeerByUser(t *testing.T) {
 		},
 	}
 
-	addedPeer, _, _, err := am.AddPeer(context.Background(), "", "", existingUserID, newPeer, false)
+	addedPeer, _, _, _, err := am.AddPeer(context.Background(), "", "", existingUserID, newPeer, false)
 	require.NoError(t, err)
 	assert.Equal(t, newPeer.ExtraDNSLabels, addedPeer.ExtraDNSLabels)
 
@@ -1465,7 +1465,7 @@ func Test_RegisterPeerBySetupKey(t *testing.T) {
 				ExtraDNSLabels: newPeerTemplate.ExtraDNSLabels,
 			}
 
-			addedPeer, _, _, err := am.AddPeer(context.Background(), "", tc.existingSetupKeyID, "", currentPeer, false)
+			addedPeer, _, _, _, err := am.AddPeer(context.Background(), "", tc.existingSetupKeyID, "", currentPeer, false)
 
 			if tc.expectAddPeerError {
 				require.Error(t, err, "Expected an error when adding peer with setup key: %s", tc.existingSetupKeyID)
@@ -1577,7 +1577,7 @@ func Test_RegisterPeerRollbackOnFailure(t *testing.T) {
 		SSHEnabled: false,
 	}
 
-	_, _, _, err = am.AddPeer(context.Background(), "", faultyKey, "", newPeer, false)
+	_, _, _, _, err = am.AddPeer(context.Background(), "", faultyKey, "", newPeer, false)
 	require.Error(t, err)
 
 	_, err = s.GetPeerByPeerPubKey(context.Background(), store.LockingStrengthNone, newPeer.Key)
@@ -1723,7 +1723,7 @@ func Test_LoginPeer(t *testing.T) {
 			if sk.AllowExtraDNSLabels {
 				currentPeer.ExtraDNSLabels = newPeerTemplate.ExtraDNSLabels
 			}
-			_, _, _, err = am.AddPeer(context.Background(), "", tc.setupKey, "", currentPeer, false)
+			_, _, _, _, err = am.AddPeer(context.Background(), "", tc.setupKey, "", currentPeer, false)
 			require.NoError(t, err, "Expected no error when adding peer with setup key: %s", tc.setupKey)
 
 			loginInput := types.PeerLogin{
@@ -1739,12 +1739,12 @@ func Test_LoginPeer(t *testing.T) {
 				loginInput.ExtraDNSLabels = tc.extraDNSLabels
 			}
 
-			loggedinPeer, networkMap, postureChecks, loginErr := am.LoginPeer(context.Background(), loginInput)
+			loggedinPeer, network, postureChecks, _, loginErr := am.LoginPeer(context.Background(), loginInput)
 			if tc.expectLoginError {
 				require.Error(t, loginErr, "Expected an error during LoginPeer with setup key: %s", tc.setupKey)
 				assert.Contains(t, loginErr.Error(), tc.expectedErrorMsgSubstring, "Error message mismatch")
 				assert.Nil(t, loggedinPeer, "LoggedinPeer should be nil on error")
-				assert.Nil(t, networkMap, "NetworkMap should be nil on error")
+				assert.Nil(t, network, "Network should be nil on error")
 				assert.Nil(t, postureChecks, "PostureChecks should be empty or nil on error")
 				return
 			}
@@ -1757,7 +1757,7 @@ func Test_LoginPeer(t *testing.T) {
 			} else {
 				assert.Equal(t, currentPeer.ExtraDNSLabels, loggedinPeer.ExtraDNSLabels, "ExtraDNSLabels mismatch on loggedinPeer")
 			}
-			assert.NotNil(t, networkMap, "networkMap should not be nil on success")
+			assert.NotNil(t, network, "network should not be nil on success")
 
 			assert.Equal(t, existingAccountID, loggedinPeer.AccountID, "AccountID mismatch for logged peer")
 
@@ -1863,7 +1863,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 		require.NoError(t, err)
 
 		expectedPeerKey := key.PublicKey().String()
-		peer4, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser1", &nbpeer.Peer{
+		peer4, _, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser1", &nbpeer.Peer{
 			Key:  expectedPeerKey,
 			Meta: nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
 		}, false)
@@ -1986,7 +1986,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 		require.NoError(t, err)
 
 		expectedPeerKey := key.PublicKey().String()
-		peer4, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser1", &nbpeer.Peer{
+		peer4, _, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser1", &nbpeer.Peer{
 			Key:                    expectedPeerKey,
 			LoginExpirationEnabled: true,
 			Meta:                   nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
@@ -2053,7 +2053,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 		require.NoError(t, err)
 
 		expectedPeerKey := key.PublicKey().String()
-		peer5, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser2", &nbpeer.Peer{
+		peer5, _, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser2", &nbpeer.Peer{
 			Key:                    expectedPeerKey,
 			LoginExpirationEnabled: true,
 			Meta:                   nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
@@ -2108,7 +2108,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 		require.NoError(t, err)
 
 		expectedPeerKey := key.PublicKey().String()
-		peer6, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser3", &nbpeer.Peer{
+		peer6, _, _, _, err = manager.AddPeer(context.Background(), "", "", "regularUser3", &nbpeer.Peer{
 			Key:                    expectedPeerKey,
 			LoginExpirationEnabled: true,
 			Meta:                   nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
@@ -2286,7 +2286,7 @@ func Test_AddPeer(t *testing.T) {
 
 			<-start
 
-			_, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", newPeer, false)
+			_, _, _, _, err := manager.AddPeer(context.Background(), "", setupKey.Key, "", newPeer, false)
 			if err != nil {
 				errs <- fmt.Errorf("AddPeer failed for peer %d: %w", i, err)
 				return
@@ -2366,7 +2366,7 @@ func TestAddPeer_UserPendingApprovalBlocked(t *testing.T) {
 		},
 	}
 
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", pendingUser.Id, peer, false)
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", pendingUser.Id, peer, false)
 	require.Error(t, err)
 	assert.Contains(t, err.Error(), "user pending approval cannot add peers")
 }
@@ -2401,7 +2401,7 @@ func TestAddPeer_ApprovedUserCanAddPeers(t *testing.T) {
 		},
 	}
 
-	_, _, _, err = manager.AddPeer(context.Background(), "", "", regularUser.Id, peer, false)
+	_, _, _, _, err = manager.AddPeer(context.Background(), "", "", regularUser.Id, peer, false)
 	require.NoError(t, err, "Regular user should be able to add peers")
 }
 
@@ -2444,7 +2444,7 @@ func TestLoginPeer_UserPendingApprovalBlocked(t *testing.T) {
 			WtVersion: "0.28.0",
 		},
 	}
-	existingPeer, _, _, err := manager.AddPeer(context.Background(), "", "", pendingUser.Id, newPeer, false)
+	existingPeer, _, _, _, err := manager.AddPeer(context.Background(), "", "", pendingUser.Id, newPeer, false)
 	require.NoError(t, err)
 
 	// Now set the user back to pending approval after peer was created
@@ -2463,7 +2463,7 @@ func TestLoginPeer_UserPendingApprovalBlocked(t *testing.T) {
 		},
 	}
 
-	_, _, _, err = manager.LoginPeer(context.Background(), login)
+	_, _, _, _, err = manager.LoginPeer(context.Background(), login)
 	require.Error(t, err)
 	e, ok := status.FromError(err)
 	require.True(t, ok, "error is not a gRPC status error")
@@ -2500,7 +2500,7 @@ func TestLoginPeer_ApprovedUserCanLogin(t *testing.T) {
 			WtVersion: "0.28.0",
 		},
 	}
-	existingPeer, _, _, err := manager.AddPeer(context.Background(), "", "", regularUser.Id, newPeer, false)
+	existingPeer, _, _, _, err := manager.AddPeer(context.Background(), "", "", regularUser.Id, newPeer, false)
 	require.NoError(t, err)
 
 	// Try to login with regular user
@@ -2513,7 +2513,7 @@ func TestLoginPeer_ApprovedUserCanLogin(t *testing.T) {
 		},
 	}
 
-	_, _, _, err = manager.LoginPeer(context.Background(), login)
+	_, _, _, _, err = manager.LoginPeer(context.Background(), login)
 	require.NoError(t, err, "Regular user should be able to login peers")
 }
 
@@ -2837,7 +2837,7 @@ func TestUpdatePeer_DnsLabelCollisionWithFQDN(t *testing.T) {
 	// Add first peer with hostname that produces DNS label "netbird1"
 	key1, err := wgtypes.GenerateKey()
 	require.NoError(t, err)
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "netbird1.netbird.cloud"},
 	}, false)
@@ -2847,7 +2847,7 @@ func TestUpdatePeer_DnsLabelCollisionWithFQDN(t *testing.T) {
 	// Add second peer with a different hostname
 	key2, err := wgtypes.GenerateKey()
 	require.NoError(t, err)
-	peer2, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer2, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "ip-10-29-5-130"},
 	}, false)
@@ -2871,7 +2871,7 @@ func TestUpdatePeer_DnsLabelUniqueName(t *testing.T) {
 
 	key1, err := wgtypes.GenerateKey()
 	require.NoError(t, err)
-	peer1, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer1, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key1.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "web-server"},
 	}, false)
@@ -2881,7 +2881,7 @@ func TestUpdatePeer_DnsLabelUniqueName(t *testing.T) {
 	// Add second peer and rename it to a unique FQDN whose first label doesn't collide
 	key2, err := wgtypes.GenerateKey()
 	require.NoError(t, err)
-	peer2, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
+	peer2, _, _, _, err := manager.AddPeer(context.Background(), "", "", userID, &nbpeer.Peer{
 		Key:  key2.PublicKey().String(),
 		Meta: nbpeer.PeerSystemMeta{Hostname: "old-name"},
 	}, false)

management/server/store/sql_store_test.go

@@ -6,7 +6,6 @@ import (
 	b64 "encoding/base64"
 	"encoding/binary"
 	"fmt"
-	"math/rand"
 	"net"
 	"net/netip"
 	"os"
@@ -92,7 +91,7 @@ func runLargeTest(t *testing.T, store Store) {
 	account.SetupKeys[setupKey.Key] = setupKey
 	const numPerAccount = 6000
 	for n := 0; n < numPerAccount; n++ {
-		netIP := randomIPv4()
+		netIP := sequentialIPv4(n)
 		peerID := fmt.Sprintf("%s-peer-%d", account.Id, n)
 		addr, _ := netip.AddrFromSlice(netIP)
 
@@ -216,12 +215,12 @@ func runLargeTest(t *testing.T, store Store) {
 	}
 }
 
-func randomIPv4() net.IP {
-	rand.New(rand.NewSource(time.Now().UnixNano()))
+// sequentialIPv4 returns a unique IPv4 address for the given index, avoiding
+// the random collisions that would otherwise violate the unique (account_id, ip)
+// index when generating a large number of peers.
+func sequentialIPv4(n int) net.IP {
 	b := make([]byte, 4)
-	for i := range b {
-		b[i] = byte(rand.Intn(256))
-	}
+	binary.BigEndian.PutUint32(b, 0x0A000000+uint32(n))
 	return net.IP(b)
 }
 

management/server/types/account.go

@@ -1156,6 +1156,47 @@ func policyRuleImpliesLegacySSH(rule *PolicyRule) bool {
 	return rule.Protocol == PolicyRuleProtocolALL || (rule.Protocol == PolicyRuleProtocolTCP && (portsIncludesSSH(rule.Ports) || portRangeIncludesSSH(rule.PortRanges)))
 }
 
+// PeerSSHEnabledFromPolicies is the network-map-free equivalent of the sshEnabled
+// determination in GetPeerConnectionResources / CalculateNetworkMapFromComponents.
+func PeerSSHEnabledFromPolicies(policies []*Policy, peerID string, peerGroupIDs map[string]struct{}, peerSSHEnabled bool) bool {
+	for _, policy := range policies {
+		if !policy.Enabled {
+			continue
+		}
+
+		for _, rule := range policy.Rules {
+			if !rule.Enabled {
+				continue
+			}
+
+			isSSHRule := rule.Protocol == PolicyRuleProtocolNetbirdSSH ||
+				(policyRuleImpliesLegacySSH(rule) && peerSSHEnabled)
+			if !isSSHRule {
+				continue
+			}
+
+			if ruleHasDestination(rule, peerID, peerGroupIDs) {
+				return true
+			}
+		}
+	}
+
+	return false
+}
+
+func ruleHasDestination(rule *PolicyRule, peerID string, peerGroupIDs map[string]struct{}) bool {
+	if rule.DestinationResource.Type == ResourceTypePeer && rule.DestinationResource.ID != "" {
+		return rule.DestinationResource.ID == peerID
+	}
+
+	for _, groupID := range rule.Destinations {
+		if _, ok := peerGroupIDs[groupID]; ok {
+			return true
+		}
+	}
+	return false
+}
+
 func portRangeIncludesSSH(portRanges []RulePortRange) bool {
 	for _, pr := range portRanges {
 		if (pr.Start <= defaultSSHPortNumber && pr.End >= defaultSSHPortNumber) || (pr.Start <= nativeSSHPortNumber && pr.End >= nativeSSHPortNumber) {

management/server/types/networkmap_components_correctness_test.go

@@ -1233,3 +1233,97 @@ func TestComponents_DisabledRuleInEnabledPolicy(t *testing.T) {
 	assert.True(t, has3000, "enabled rule should generate firewall rule for port 3000")
 	assert.False(t, has3001, "disabled rule should NOT generate firewall rule for port 3001")
 }
+
+func peerGroupIDSet(account *types.Account, peerID string) map[string]struct{} {
+	return account.GetPeerGroups(peerID)
+}
+
+func assertSSHEquivalence(t *testing.T, account *types.Account, peerID string, validatedPeers map[string]struct{}) {
+	t.Helper()
+	nm := componentsNetworkMap(account, peerID, validatedPeers)
+	require.NotNil(t, nm)
+
+	got := types.PeerSSHEnabledFromPolicies(account.Policies, peerID, peerGroupIDSet(account, peerID), account.Peers[peerID].SSHEnabled)
+	assert.Equalf(t, nm.EnableSSH, got, "PeerSSHEnabledFromPolicies mismatch for %s", peerID)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_NetbirdSSHProtocol(t *testing.T) {
+	account, validatedPeers := scalableTestAccount(20, 2)
+	account.Groups["ssh-users"] = &types.Group{ID: "ssh-users", Name: "SSH Users", Peers: []string{}}
+	account.Policies = append(account.Policies, &types.Policy{
+		ID: "policy-ssh", Name: "SSH Access", Enabled: true, AccountID: "test-account",
+		Rules: []*types.PolicyRule{{
+			ID: "rule-ssh", Name: "Allow SSH", Enabled: true,
+			Action: types.PolicyTrafficActionAccept, Protocol: types.PolicyRuleProtocolNetbirdSSH,
+			Bidirectional: false,
+			Sources:       []string{"group-0"}, Destinations: []string{"group-1"},
+			AuthorizedGroups: map[string][]string{"ssh-users": {"root"}},
+		}},
+	})
+
+	assertSSHEquivalence(t, account, "peer-10", validatedPeers)
+	assertSSHEquivalence(t, account, "peer-0", validatedPeers)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_NoSSHPolicy(t *testing.T) {
+	account, validatedPeers := scalableTestAccount(20, 2)
+	assertSSHEquivalence(t, account, "peer-0", validatedPeers)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_LegacyImpliedSSH(t *testing.T) {
+	account, validatedPeers := scalableTestAccount(20, 2)
+	account.Peers["peer-10"].SSHEnabled = true
+	assertSSHEquivalence(t, account, "peer-10", validatedPeers)
+	assertSSHEquivalence(t, account, "peer-11", validatedPeers)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_PeerAsDestinationResource(t *testing.T) {
+	account, validatedPeers := scalableTestAccountWithoutDefaultPolicy(20, 2)
+	account.Policies = append(account.Policies, &types.Policy{
+		ID: "policy-ssh-res", Name: "SSH to peer", Enabled: true, AccountID: "test-account",
+		Rules: []*types.PolicyRule{{
+			ID: "rule-ssh-res", Name: "SSH to peer-5", Enabled: true,
+			Action: types.PolicyTrafficActionAccept, Protocol: types.PolicyRuleProtocolNetbirdSSH,
+			Sources:             []string{"group-0"},
+			DestinationResource: types.Resource{ID: "peer-5", Type: types.ResourceTypePeer},
+		}},
+	})
+
+	assertSSHEquivalence(t, account, "peer-5", validatedPeers)
+	assertSSHEquivalence(t, account, "peer-6", validatedPeers)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_DisabledSSHPolicy(t *testing.T) {
+	account, validatedPeers := scalableTestAccountWithoutDefaultPolicy(20, 2)
+	account.Policies = append(account.Policies, &types.Policy{
+		ID: "policy-ssh-off", Name: "SSH disabled", Enabled: false, AccountID: "test-account",
+		Rules: []*types.PolicyRule{{
+			ID: "rule-ssh-off", Name: "Allow SSH", Enabled: true,
+			Action: types.PolicyTrafficActionAccept, Protocol: types.PolicyRuleProtocolNetbirdSSH,
+			Sources: []string{"group-0"}, Destinations: []string{"group-1"},
+		}},
+	})
+	assertSSHEquivalence(t, account, "peer-10", validatedPeers)
+}
+
+func TestPeerSSHEnabledFromPolicies_MatchesMap_Sweep(t *testing.T) {
+	account, validatedPeers := scalableTestAccount(60, 6)
+	account.Policies = append(account.Policies, &types.Policy{
+		ID: "policy-ssh-sweep", Name: "SSH sweep", Enabled: true, AccountID: "test-account",
+		Rules: []*types.PolicyRule{{
+			ID: "rule-ssh-sweep", Name: "Allow SSH", Enabled: true,
+			Action: types.PolicyTrafficActionAccept, Protocol: types.PolicyRuleProtocolNetbirdSSH,
+			Sources: []string{"group-0"}, Destinations: []string{"group-2"},
+		}},
+	})
+	for peerID := range account.Peers {
+		account.Peers[peerID].SSHEnabled = len(peerID)%2 == 0
+	}
+
+	for peerID := range account.Peers {
+		if _, ok := validatedPeers[peerID]; !ok {
+			continue
+		}
+		assertSSHEquivalence(t, account, peerID, validatedPeers)
+	}
+}

management/server/user_test.go

@@ -1565,7 +1565,7 @@ func TestUserAccountPeersUpdate(t *testing.T) {
 	require.NoError(t, err)
 
 	expectedPeerKey := key.PublicKey().String()
-	peer4, _, _, err := manager.AddPeer(context.Background(), "", "", "regularUser2", &nbpeer.Peer{
+	peer4, _, _, _, err := manager.AddPeer(context.Background(), "", "", "regularUser2", &nbpeer.Peer{
 		Key:  expectedPeerKey,
 		Meta: nbpeer.PeerSystemMeta{Hostname: expectedPeerKey},
 	}, false)

proxy/Dockerfile

@@ -7,7 +7,8 @@ RUN echo "netbird:x:1000:1000:netbird:/var/lib/netbird:/sbin/nologin" > /tmp/pas
     mkdir -p /tmp/certs
 
 FROM gcr.io/distroless/base:debug
-COPY netbird-proxy /go/bin/netbird-proxy
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-proxy /go/bin/netbird-proxy
 COPY --from=builder /tmp/passwd /etc/passwd
 COPY --from=builder /tmp/group /etc/group
 COPY --from=builder --chown=1000:1000 /tmp/var/lib/netbird /var/lib/netbird

relay/Dockerfile

@@ -1,4 +1,5 @@
 FROM gcr.io/distroless/base:debug
 ENTRYPOINT [ "/go/bin/netbird-relay" ]
 ENV NB_LOG_FILE=console
-COPY netbird-relay /go/bin/netbird-relay
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-relay /go/bin/netbird-relay

relay/metrics/realy.go

@@ -6,6 +6,7 @@ import (
 	"time"
 
 	log "github.com/sirupsen/logrus"
+	"go.opentelemetry.io/otel/attribute"
 	"go.opentelemetry.io/otel/metric"
 )
 
@@ -119,8 +120,8 @@ func NewMetrics(ctx context.Context, meter metric.Meter) (*Metrics, error) {
 }
 
 // PeerConnected increments the number of connected peers and increments number of idle connections
-func (m *Metrics) PeerConnected(id string) {
-	m.peers.Add(m.ctx, 1)
+func (m *Metrics) PeerConnected(id, transport string) {
+	m.peers.Add(m.ctx, 1, metric.WithAttributes(attribute.String("transport", transport)))
 	m.mutexActivity.Lock()
 	defer m.mutexActivity.Unlock()
 
@@ -138,8 +139,8 @@ func (m *Metrics) RecordPeerStoreTime(duration time.Duration) {
 }
 
 // PeerDisconnected decrements the number of connected peers and decrements number of idle or active connections
-func (m *Metrics) PeerDisconnected(id string) {
-	m.peers.Add(m.ctx, -1)
+func (m *Metrics) PeerDisconnected(id, transport string) {
+	m.peers.Add(m.ctx, -1, metric.WithAttributes(attribute.String("transport", transport)))
 	m.mutexActivity.Lock()
 	defer m.mutexActivity.Unlock()
 

relay/server/listener/conn.go

@@ -11,4 +11,6 @@ type Conn interface {
 	Write(ctx context.Context, b []byte) (n int, err error)
 	RemoteAddr() net.Addr
 	Close() error
+	// Protocol returns the transport name.
+	Protocol() string
 }

relay/server/listener/quic/conn.go

@@ -42,6 +42,11 @@ func (c *Conn) RemoteAddr() net.Addr {
 	return c.session.RemoteAddr()
 }
 
+// Protocol returns the transport name for this connection.
+func (c *Conn) Protocol() string {
+	return "quic"
+}
+
 func (c *Conn) Close() error {
 	c.closedMu.Lock()
 	if c.closed {

relay/server/listener/ws/conn.go

@@ -64,6 +64,11 @@ func (c *Conn) RemoteAddr() net.Addr {
 	return c.rAddr
 }
 
+// Protocol returns the transport name for this connection.
+func (c *Conn) Protocol() string {
+	return "ws"
+}
+
 func (c *Conn) Close() error {
 	c.closedMu.Lock()
 	c.closed = true

relay/server/relay.go

@@ -154,15 +154,16 @@ func (r *Relay) Accept(conn listener.Conn) {
 	}
 	r.notifier.PeerCameOnline(peer.ID())
 
+	transport := conn.Protocol()
 	r.metrics.RecordPeerStoreTime(time.Since(storeTime))
-	r.metrics.PeerConnected(peer.String())
+	r.metrics.PeerConnected(peer.String(), transport)
 	go func() {
 		peer.Work()
 		if deleted := r.store.DeletePeer(peer); deleted {
 			r.notifier.PeerWentOffline(peer.ID())
 		}
 		peer.log.Debugf("relay connection closed")
-		r.metrics.PeerDisconnected(peer.String())
+		r.metrics.PeerDisconnected(peer.String(), transport)
 	}()
 
 	if err := h.handshakeResponse(hsCtx); err != nil {

shared/relay/client/client.go

@@ -145,6 +145,11 @@ func (cc *connContainer) close() {
 	}
 }
 
+// transportConn is implemented by relay connections that know their transport.
+type transportConn interface {
+	Protocol() string
+}
+
 // Client is a client for the relay server. It is responsible for establishing a connection to the relay server and
 // managing connections to other peers. All exported functions are safe to call concurrently. After close the connection,
 // the client can be reused by calling Connect again. When the client is closed, all connections are closed too.
@@ -182,6 +187,18 @@ type Client struct {
 	// datagramFallbackTriggered guards a single fallback per connection so a
 	// burst of oversized datagrams triggers one reconnect, not many.
 	datagramFallbackTriggered atomic.Bool
+
+	// transport is the negotiated relay transport of the
+	// current connection, guarded by mu.
+	transport string
+}
+
+// Transport returns the negotiated relay transport of the current connection,
+// or an empty string when not connected.
+func (c *Client) Transport() string {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	return c.transport
 }
 
 // SetTransportFallback wires the shared datagram-transport fallback tracker.
@@ -402,6 +419,9 @@ func (c *Client) connect(ctx context.Context) (*RelayAddr, error) {
 	}
 	c.relayConn = conn
 	c.datagramFallbackTriggered.Store(false)
+	if tc, ok := conn.(transportConn); ok {
+		c.transport = tc.Protocol()
+	}
 
 	instanceURL, err := c.handShake(ctx)
 	if err != nil {
@@ -792,6 +812,7 @@ func (c *Client) close(gracefullyExit bool) error {
 		return nil
 	}
 	c.serviceIsRunning = false
+	c.transport = ""
 
 	c.muInstanceURL.Lock()
 	c.instanceURL = nil

shared/relay/client/dialer/quic/conn.go

@@ -57,6 +57,11 @@ func (c *Conn) Write(b []byte) (int, error) {
 	return len(b), nil
 }
 
+// Protocol returns the transport name for this connection.
+func (c *Conn) Protocol() string {
+	return Network
+}
+
 func (c *Conn) RemoteAddr() net.Addr {
 	return c.session.RemoteAddr()
 }

shared/relay/client/dialer/quic/quic.go

@@ -59,14 +59,12 @@ func (d Dialer) Dial(ctx context.Context, address, serverName string) (net.Conn,
 
 	udpConn, err := nbnet.ListenUDP("udp", &net.UDPAddr{Port: 0})
 	if err != nil {
-		log.Errorf("failed to listen on UDP: %s", err)
-		return nil, err
+		return nil, fmt.Errorf("listen udp: %w", err)
 	}
 
 	udpAddr, err := net.ResolveUDPAddr("udp", quicURL)
 	if err != nil {
-		log.Errorf("failed to resolve UDP address: %s", err)
-		return nil, err
+		return nil, fmt.Errorf("resolve %s: %w", quicURL, err)
 	}
 
 	session, err := quic.Dial(ctx, udpConn, udpAddr, tlsClientConfig, quicConfig)
@@ -74,7 +72,7 @@ func (d Dialer) Dial(ctx context.Context, address, serverName string) (net.Conn,
 		if errors.Is(err, context.Canceled) {
 			return nil, err
 		}
-		log.Errorf("failed to dial to Relay server via QUIC '%s': %s", quicURL, err)
+		log.Debugf("failed to dial to Relay server via QUIC '%s': %s", quicURL, err)
 		return nil, err
 	}
 

shared/relay/client/dialer/race_dialer.go

@@ -3,6 +3,7 @@ package dialer
 import (
 	"context"
 	"errors"
+	"fmt"
 	"net"
 	"time"
 
@@ -71,6 +72,7 @@ func (r *RaceDial) Dial(ctx context.Context) (net.Conn, error) {
 
 	connChan := make(chan dialResult, len(r.dialerFns))
 	winnerConn := make(chan net.Conn, 1)
+	errChan := make(chan error, 1)
 	abortCtx, abort := context.WithCancel(ctx)
 	defer abort()
 
@@ -78,11 +80,11 @@ func (r *RaceDial) Dial(ctx context.Context) (net.Conn, error) {
 		go r.dial(dfn, abortCtx, connChan)
 	}
 
-	go r.processResults(connChan, winnerConn, abort)
+	go r.processResults(connChan, winnerConn, errChan, abort)
 
 	conn, ok := <-winnerConn
 	if !ok {
-		return nil, errors.New("failed to dial to Relay server on any protocol")
+		return nil, <-errChan
 	}
 	return conn, nil
 }
@@ -90,6 +92,7 @@ func (r *RaceDial) Dial(ctx context.Context) (net.Conn, error) {
 // dialSequential tries each dialer in order, returning the first connection and
 // falling back to the next on failure.
 func (r *RaceDial) dialSequential(ctx context.Context) (net.Conn, error) {
+	var errs []error
 	for _, dfn := range r.dialerFns {
 		if err := ctx.Err(); err != nil {
 			return nil, err
@@ -103,12 +106,13 @@ func (r *RaceDial) dialSequential(ctx context.Context) (net.Conn, error) {
 				return nil, err
 			}
 			r.log.Errorf("failed to dial via %s: %s", dfn.Protocol(), err)
+			errs = append(errs, fmt.Errorf("%s: %w", dfn.Protocol(), err))
 			continue
 		}
 		r.log.Infof("successfully dialed via: %s", dfn.Protocol())
 		return conn, nil
 	}
-	return nil, errors.New("failed to dial to Relay server on any protocol")
+	return nil, dialErr(errs)
 }
 
 func (r *RaceDial) dial(dfn DialeFn, abortCtx context.Context, connChan chan dialResult) {
@@ -120,8 +124,9 @@ func (r *RaceDial) dial(dfn DialeFn, abortCtx context.Context, connChan chan dia
 	connChan <- dialResult{Conn: conn, Protocol: dfn.Protocol(), Err: err}
 }
 
-func (r *RaceDial) processResults(connChan chan dialResult, winnerConn chan net.Conn, abort context.CancelFunc) {
+func (r *RaceDial) processResults(connChan chan dialResult, winnerConn chan net.Conn, errChan chan error, abort context.CancelFunc) {
 	var hasWinner bool
+	errsByProtocol := make(map[string]error)
 	for i := 0; i < len(r.dialerFns); i++ {
 		dr := <-connChan
 		if dr.Err != nil {
@@ -129,6 +134,7 @@ func (r *RaceDial) processResults(connChan chan dialResult, winnerConn chan net.
 				r.log.Infof("connection attempt aborted via: %s", dr.Protocol)
 			} else {
 				r.log.Errorf("failed to dial via %s: %s", dr.Protocol, dr.Err)
+				errsByProtocol[dr.Protocol] = fmt.Errorf("%s: %w", dr.Protocol, dr.Err)
 			}
 			continue
 		}
@@ -146,5 +152,29 @@ func (r *RaceDial) processResults(connChan chan dialResult, winnerConn chan net.
 		hasWinner = true
 		winnerConn <- dr.Conn
 	}
+	if !hasWinner {
+		errChan <- dialErr(r.orderedErrs(errsByProtocol))
+	}
 	close(winnerConn)
 }
+
+// orderedErrs returns the per-protocol errors in dialer order, so the combined
+// error is stable regardless of which attempt failed first.
+func (r *RaceDial) orderedErrs(byProtocol map[string]error) []error {
+	errs := make([]error, 0, len(byProtocol))
+	for _, dfn := range r.dialerFns {
+		if err, ok := byProtocol[dfn.Protocol()]; ok {
+			errs = append(errs, err)
+		}
+	}
+	return errs
+}
+
+// dialErr combines per-dialer failures, preserving the underlying reasons
+// (e.g. "connection refused") rather than a generic message.
+func dialErr(errs []error) error {
+	if len(errs) == 0 {
+		return errors.New("no relay transport available")
+	}
+	return errors.Join(errs...)
+}

shared/relay/client/dialer/ws/conn.go

@@ -33,6 +33,11 @@ func NewConn(wsConn *websocket.Conn, serverAddress string, underlying net.Conn)
 	}
 }
 
+// Protocol returns the transport name for this connection.
+func (c *Conn) Protocol() string {
+	return Network
+}
+
 func (c *Conn) Read(b []byte) (n int, err error) {
 	t, ioReader, err := c.Conn.Reader(c.ctx)
 	if err != nil {

shared/relay/client/dialer/ws/ws.go

@@ -22,7 +22,7 @@ type Dialer struct {
 }
 
 func (d Dialer) Protocol() string {
-	return "WS"
+	return Network
 }
 
 func (d Dialer) Dial(ctx context.Context, address, serverName string) (net.Conn, error) {
@@ -39,7 +39,12 @@ func (d Dialer) Dial(ctx context.Context, address, serverName string) (net.Conn,
 		if errors.Is(err, context.Canceled) {
 			return nil, err
 		}
-		log.Errorf("failed to dial to Relay server '%s': %s", wsURL, err)
+		// websocket.Dial wraps the cause in verbose layers; surface the
+		// underlying network error when present.
+		var opErr *net.OpError
+		if errors.As(err, &opErr) {
+			return nil, opErr
+		}
 		return nil, err
 	}
 	if resp.Body != nil {

shared/relay/client/dialers_generic_test.go

@@ -41,14 +41,14 @@ func TestGetDialers(t *testing.T) {
 		preferWS bool
 		want     []string
 	}{
-		{name: "auto races quic and ws", mode: "auto", mtu: iface.DefaultMTU, want: []string{"quic", "WS"}},
-		{name: "ws pinned", mode: "ws", mtu: iface.DefaultMTU, want: []string{"WS"}},
+		{name: "auto races quic and ws", mode: "auto", mtu: iface.DefaultMTU, want: []string{"quic", "ws"}},
+		{name: "ws pinned", mode: "ws", mtu: iface.DefaultMTU, want: []string{"ws"}},
 		{name: "quic pinned", mode: "quic", mtu: iface.DefaultMTU, want: []string{"quic"}},
-		{name: "prefer-quic orders quic first", mode: "prefer-quic", mtu: iface.DefaultMTU, want: []string{"quic", "WS"}},
-		{name: "prefer-ws orders ws first", mode: "prefer-ws", mtu: iface.DefaultMTU, want: []string{"WS", "quic"}},
-		{name: "mtu above default forces ws", mode: "auto", mtu: iface.DefaultMTU + 100, want: []string{"WS"}},
-		{name: "sticky fallback forces ws in auto", mode: "auto", mtu: iface.DefaultMTU, preferWS: true, want: []string{"WS"}},
-		{name: "sticky fallback forces ws in prefer-quic", mode: "prefer-quic", mtu: iface.DefaultMTU, preferWS: true, want: []string{"WS"}},
+		{name: "prefer-quic orders quic first", mode: "prefer-quic", mtu: iface.DefaultMTU, want: []string{"quic", "ws"}},
+		{name: "prefer-ws orders ws first", mode: "prefer-ws", mtu: iface.DefaultMTU, want: []string{"ws", "quic"}},
+		{name: "mtu above default forces ws", mode: "auto", mtu: iface.DefaultMTU + 100, want: []string{"ws"}},
+		{name: "sticky fallback forces ws in auto", mode: "auto", mtu: iface.DefaultMTU, preferWS: true, want: []string{"ws"}},
+		{name: "sticky fallback forces ws in prefer-quic", mode: "prefer-quic", mtu: iface.DefaultMTU, preferWS: true, want: []string{"ws"}},
 		{name: "quic pin overrides sticky fallback", mode: "quic", mtu: iface.DefaultMTU, preferWS: true, want: []string{"quic"}},
 	}
 
@@ -91,11 +91,11 @@ func TestStickyFallbackAfterDatagramTooLarge(t *testing.T) {
 	}
 
 	// First dial races both transports.
-	assert.Equal(t, []string{"quic", "WS"}, protocols(c.getDialers(transportModeFromEnv())))
+	assert.Equal(t, []string{"quic", "ws"}, protocols(c.getDialers(transportModeFromEnv())))
 
 	// An oversized datagram records the fallback for this server.
 	c.onDatagramTooLarge(&closeTrackingConn{}, netErr.ErrDatagramTooLarge)
 
 	// The reconnect now sticks to WebSocket.
-	assert.Equal(t, []string{"WS"}, protocols(c.getDialers(transportModeFromEnv())))
+	assert.Equal(t, []string{"ws"}, protocols(c.getDialers(transportModeFromEnv())))
 }

shared/relay/client/guard.go

@@ -2,6 +2,7 @@ package client
 
 import (
 	"context"
+	"sync/atomic"
 	"time"
 
 	"github.com/cenkalti/backoff/v4"
@@ -20,6 +21,10 @@ type Guard struct {
 	// maxBackoffInterval caps the exponential backoff between reconnect
 	// attempts.
 	maxBackoffInterval time.Duration
+
+	// lastErr is the error from the most recent failed reconnect attempt,
+	// surfaced as the home relay status while disconnected.
+	lastErr atomic.Pointer[error]
 }
 
 // NewGuard creates a new guard for the relay client. A non-positive
@@ -37,6 +42,15 @@ func NewGuard(sp *ServerPicker, maxBackoffInterval time.Duration) *Guard {
 	return g
 }
 
+// LastError returns the error from the most recent failed reconnect attempt, or
+// nil if reconnection last succeeded.
+func (g *Guard) LastError() error {
+	if p := g.lastErr.Load(); p != nil {
+		return *p
+	}
+	return nil
+}
+
 // StartReconnectTrys is called when the relay client is disconnected from the relay server.
 // It attempts to reconnect to the relay server. The function first tries a quick reconnect
 // to the same server that was used before, if the server URL is still valid. If the quick
@@ -63,6 +77,7 @@ func (g *Guard) StartReconnectTrys(ctx context.Context, relayClient *Client) {
 		case <-ticker.C:
 			if err := g.retry(ctx); err != nil {
 				log.Errorf("failed to pick new Relay server: %s", err)
+				g.setLastError(err)
 				continue
 			}
 			return
@@ -72,6 +87,10 @@ func (g *Guard) StartReconnectTrys(ctx context.Context, relayClient *Client) {
 	}
 }
 
+func (g *Guard) setLastError(err error) {
+	g.lastErr.Store(&err)
+}
+
 func (g *Guard) tryToQuickReconnect(parentCtx context.Context, rc *Client) bool {
 	if rc == nil {
 		return false
@@ -89,6 +108,7 @@ func (g *Guard) tryToQuickReconnect(parentCtx context.Context, rc *Client) bool
 
 	if err := rc.Connect(parentCtx); err != nil {
 		log.Errorf("failed to reconnect to relay server: %s", err)
+		g.setLastError(err)
 		return false
 	}
 	return true
@@ -100,6 +120,7 @@ func (g *Guard) retry(ctx context.Context) error {
 	if err != nil {
 		return err
 	}
+	g.setLastError(nil)
 
 	// prevent to work with a deprecated Relay client instance
 	g.drainRelayClientChan()
@@ -125,6 +146,7 @@ func (g *Guard) isServerURLStillValid(rc *Client) bool {
 }
 
 func (g *Guard) notifyReconnected() {
+	g.setLastError(nil)
 	select {
 	case g.OnReconnected <- struct{}{}:
 	default:

shared/relay/client/manager.go

@@ -43,6 +43,17 @@ type OnServerCloseListener func()
 // ManagerOption configures a Manager at construction time.
 type ManagerOption func(*Manager)
 
+// RelayConnState is the connection state of a single relay server.
+type RelayConnState struct {
+	// URL is the server's instance address when connected, otherwise the
+	// configured server URL.
+	URL string
+	// Transport is the negotiated transport, empty if not connected.
+	Transport string
+	// Err is set when the relay is not connected.
+	Err error
+}
+
 // WithMaxBackoffInterval caps the exponential backoff between reconnect
 // attempts to the home relay. A non-positive value keeps the default.
 func WithMaxBackoffInterval(d time.Duration) ManagerOption {
@@ -130,6 +141,9 @@ func (m *Manager) Serve() error {
 
 	client, err := m.serverPicker.PickServer(m.ctx)
 	if err != nil {
+		// record the initial failure so status shows the real reason before
+		// the guard's first retry tick
+		m.reconnectGuard.setLastError(err)
 		go m.reconnectGuard.StartReconnectTrys(m.ctx, nil)
 	} else {
 		m.storeClient(client)
@@ -242,6 +256,56 @@ func (m *Manager) ServerURLs() []string {
 	return m.serverPicker.ServerURLs.Load().([]string)
 }
 
+// RelayConnectError returns the error from the most recent failed home relay
+// reconnect attempt, or nil if the relay last connected successfully.
+func (m *Manager) RelayConnectError() error {
+	return m.reconnectGuard.LastError()
+}
+
+// RelayStates returns the connection state of the home relay and every foreign
+// relay the manager currently tracks.
+func (m *Manager) RelayStates() []RelayConnState {
+	var states []RelayConnState
+
+	m.relayClientMu.RLock()
+	home := m.relayClient
+	m.relayClientMu.RUnlock()
+	if home != nil {
+		st := relayConnState(home)
+		// The home relay reconnects through the guard, so the real failure
+		// reason lives there rather than on the (stale) client.
+		if st.Err != nil {
+			if gErr := m.reconnectGuard.LastError(); gErr != nil {
+				st.Err = gErr
+			}
+		}
+		states = append(states, st)
+	}
+
+	// Snapshot the tracks, then query each outside the map lock: a track can be
+	// held by an in-progress Connect, and blocking on it must not stall other
+	// relay operations.
+	m.relayClientsMutex.RLock()
+	tracks := make([]*RelayTrack, 0, len(m.relayClients))
+	for _, rt := range m.relayClients {
+		tracks = append(tracks, rt)
+	}
+	m.relayClientsMutex.RUnlock()
+
+	// Only connected foreign relays carry state; a failed connect is evicted
+	// immediately (openConnVia), so there is no error state to surface.
+	for _, rt := range tracks {
+		rt.RLock()
+		rc := rt.relayClient
+		rt.RUnlock()
+		if rc != nil {
+			states = append(states, relayConnState(rc))
+		}
+	}
+
+	return states
+}
+
 // HasRelayAddress returns true if the manager is serving. With this method can check if the peer can communicate with
 // Relay service.
 func (m *Manager) HasRelayAddress() bool {
@@ -460,3 +524,11 @@ func (m *Manager) notifyOnDisconnectListeners(serverAddress string) {
 	}
 	delete(m.onDisconnectedListeners, serverAddress)
 }
+
+func relayConnState(c *Client) RelayConnState {
+	addr, err := c.ServerInstanceURL()
+	if err != nil {
+		return RelayConnState{URL: c.connectionURL, Err: err}
+	}
+	return RelayConnState{URL: addr, Transport: c.Transport()}
+}

shared/relay/client/picker.go

@@ -40,6 +40,7 @@ func (sp *ServerPicker) PickServer(parentCtx context.Context) (*Client, error) {
 
 	connResultChan := make(chan connResult, totalServers)
 	successChan := make(chan connResult, 1)
+	errChan := make(chan error, 1)
 	concurrentLimiter := make(chan struct{}, maxConcurrentServers)
 
 	log.Debugf("pick server from list: %v", sp.ServerURLs.Load().([]string))
@@ -54,17 +55,17 @@ func (sp *ServerPicker) PickServer(parentCtx context.Context) (*Client, error) {
 		}(url)
 	}
 
-	go sp.processConnResults(connResultChan, successChan)
+	go sp.processConnResults(connResultChan, successChan, errChan)
 
 	select {
 	case cr, ok := <-successChan:
 		if !ok {
-			return nil, errors.New("failed to connect to any relay server: all attempts failed")
+			return nil, <-errChan
 		}
 		log.Infof("chosen home Relay server: %s", cr.Url)
 		return cr.RelayClient, nil
 	case <-ctx.Done():
-		return nil, fmt.Errorf("failed to connect to any relay server: %w", ctx.Err())
+		return nil, fmt.Errorf("connect to relay server: %w", ctx.Err())
 	}
 }
 
@@ -80,12 +81,14 @@ func (sp *ServerPicker) startConnection(ctx context.Context, resultChan chan con
 	}
 }
 
-func (sp *ServerPicker) processConnResults(resultChan chan connResult, successChan chan connResult) {
+func (sp *ServerPicker) processConnResults(resultChan chan connResult, successChan chan connResult, errChan chan error) {
 	var hasSuccess bool
+	var errs []error
 	for numOfResults := 0; numOfResults < cap(resultChan); numOfResults++ {
 		cr := <-resultChan
 		if cr.Err != nil {
 			log.Tracef("failed to connect to Relay server: %s: %v", cr.Url, cr.Err)
+			errs = append(errs, cr.Err)
 			continue
 		}
 		log.Infof("connected to Relay server: %s", cr.Url)
@@ -101,5 +104,16 @@ func (sp *ServerPicker) processConnResults(resultChan chan connResult, successCh
 		hasSuccess = true
 		successChan <- cr
 	}
+	if !hasSuccess {
+		errChan <- pickErr(errs)
+	}
 	close(successChan)
 }
+
+// pickErr combines per-server connection failures into a single error.
+func pickErr(errs []error) error {
+	if len(errs) == 0 {
+		return errors.New("no relay server available")
+	}
+	return errors.Join(errs...)
+}

shared/signal/client/grpc.go

@@ -2,9 +2,11 @@ package client
 
 import (
 	"context"
+	"errors"
 	"fmt"
 	"io"
 	"sync"
+	"sync/atomic"
 	"time"
 
 	"github.com/cenkalti/backoff/v4"
@@ -23,7 +25,23 @@ import (
 	"github.com/netbirdio/netbird/util/wsproxy"
 )
 
-const healthCheckTimeout = 5 * time.Second
+const (
+	// receiveInactivityThreshold is how long the receive stream may be silent
+	// before the watchdog actively probes it. The gRPC transport can stay
+	// healthy (keepalive satisfied) while the server stops delivering messages,
+	// which the transport layer cannot detect.
+	receiveInactivityThreshold = 30 * time.Second
+	// receiveProbeTimeout is how long the watchdog waits for its self-addressed
+	// probe to round-trip back on the stream before declaring the receive
+	// direction dead.
+	receiveProbeTimeout = 10 * time.Second
+	// receiveWatchdogInterval is how often the watchdog evaluates the stream.
+	receiveWatchdogInterval = 10 * time.Second
+)
+
+// errReceiveStreamStalled is reported when the receive stream is transport-alive
+// but no longer delivering messages, so the stream is torn down to reconnect.
+var errReceiveStreamStalled = errors.New("signal receive stream stalled")
 
 // ConnStateNotifier is a wrapper interface of the status recorder
 type ConnStateNotifier interface {
@@ -52,6 +70,14 @@ type GrpcClient struct {
 	decryptionWorker       *Worker
 	decryptionWorkerCancel context.CancelFunc
 	decryptionWg           sync.WaitGroup
+
+	// lastReceived holds the Unix-nano timestamp of the last message read from
+	// the receive stream, used by the receive watchdog.
+	lastReceived atomic.Int64
+	// receiveStalled is set by the receive watchdog when the stream is
+	// transport-alive but no longer delivering messages. It is the source of
+	// truth IsHealthy reads, and is cleared once any frame is received again.
+	receiveStalled atomic.Bool
 }
 
 // NewClient creates a new Signal client
@@ -148,9 +174,9 @@ func (c *GrpcClient) Receive(ctx context.Context, msgHandler func(msg *proto.Mes
 
 		// connect to Signal stream identifying ourselves with a public WireGuard key
 		// todo once the key rotation logic has been implemented, consider changing to some other identifier (received from management)
-		ctx, cancelStream := context.WithCancel(ctx)
+		streamCtx, cancelStream := context.WithCancel(ctx)
 		defer cancelStream()
-		stream, err := c.connect(ctx, c.key.PublicKey().String())
+		stream, err := c.connect(streamCtx, c.key.PublicKey().String())
 		if err != nil {
 			log.Warnf("disconnected from the Signal Exchange due to an error: %v", err)
 			return err
@@ -164,9 +190,16 @@ func (c *GrpcClient) Receive(ctx context.Context, msgHandler func(msg *proto.Mes
 		// Start worker pool if not already started
 		c.startEncryptionWorker(msgHandler)
 
+		// Guard the receive direction: the transport can stay healthy while the
+		// server stops delivering messages. The watchdog reconnects via cancelStream.
+		c.markReceived()
+		go c.watchReceiveStream(streamCtx, cancelStream)
+
 		// start receiving messages from the Signal stream (from other peers through signal)
 		err = c.receive(stream)
 		if err != nil {
+			// Check the parent context, not streamCtx: a watchdog-triggered
+			// cancelStream must reconnect, only a parent cancel is shutdown.
 			if ctx.Err() != nil {
 				log.Debugf("signal connection context has been canceled, this usually indicates shutdown")
 				return nil
@@ -252,7 +285,10 @@ func (c *GrpcClient) Ready() bool {
 	return c.signalConn.GetState() == connectivity.Ready || c.signalConn.GetState() == connectivity.Idle
 }
 
-// IsHealthy probes the gRPC connection and returns false on errors
+// IsHealthy reports whether the Signal connection is usable, based on the
+// transport state plus the receive watchdog's verdict, and updates the status
+// recorder accordingly. It does not actively probe: the watchdog
+// (watchReceiveStream) owns probing the receive path and reconnecting.
 func (c *GrpcClient) IsHealthy() bool {
 	switch c.signalConn.GetState() {
 	case connectivity.TransientFailure:
@@ -265,16 +301,8 @@ func (c *GrpcClient) IsHealthy() bool {
 	case connectivity.Ready:
 	}
 
-	ctx, cancel := context.WithTimeout(c.ctx, healthCheckTimeout)
-	defer cancel()
-	_, err := c.realClient.Send(ctx, &proto.EncryptedMessage{
-		Key:       c.key.PublicKey().String(),
-		RemoteKey: "dummy",
-		Body:      nil,
-	})
-	if err != nil {
-		c.notifyDisconnected(err)
-		log.Warnf("health check returned: %s", err)
+	if c.receiveStalled.Load() {
+		c.notifyDisconnected(errReceiveStreamStalled)
 		return false
 	}
 	c.notifyConnected()
@@ -398,6 +426,68 @@ func (c *GrpcClient) Send(msg *proto.Message) error {
 	return err
 }
 
+// markReceived records that a frame was just read from the receive stream and
+// clears the stalled flag.
+func (c *GrpcClient) markReceived() {
+	c.lastReceived.Store(time.Now().UnixNano())
+	c.receiveStalled.Store(false)
+}
+
+// idleSinceReceive returns how long the receive stream has been silent.
+func (c *GrpcClient) idleSinceReceive() time.Duration {
+	return time.Since(time.Unix(0, c.lastReceived.Load()))
+}
+
+// watchReceiveStream guards against a receive stream that is transport-alive but
+// no longer delivering messages. While the stream is idle past
+// receiveInactivityThreshold it sends a self-addressed probe that the Signal
+// server routes back to this client. If the probe does not round-trip within
+// receiveProbeTimeout the receive direction is considered dead and cancelStream
+// is called so the retry loop reconnects.
+func (c *GrpcClient) watchReceiveStream(ctx context.Context, cancelStream context.CancelFunc) {
+	ticker := time.NewTicker(receiveWatchdogInterval)
+	defer ticker.Stop()
+
+	var probeSentAt time.Time
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case <-ticker.C:
+			if c.idleSinceReceive() < receiveInactivityThreshold {
+				probeSentAt = time.Time{}
+				continue
+			}
+
+			if !probeSentAt.IsZero() && time.Since(probeSentAt) >= receiveProbeTimeout {
+				log.Warnf("signal receive stream stalled: no messages for %s and probe did not return, reconnecting", c.idleSinceReceive().Round(time.Second))
+				c.receiveStalled.Store(true)
+				c.notifyDisconnected(errReceiveStreamStalled)
+				cancelStream()
+				return
+			}
+
+			if probeSentAt.IsZero() {
+				if err := c.sendReceiveProbe(); err != nil {
+					log.Debugf("failed to send signal receive probe: %v", err)
+				}
+				probeSentAt = time.Now()
+			}
+		}
+	}
+}
+
+// sendReceiveProbe sends a self-addressed heartbeat. The Signal server routes it
+// back to this client, exercising the exact receive path the watchdog guards.
+func (c *GrpcClient) sendReceiveProbe() error {
+	self := c.key.PublicKey().String()
+	return c.Send(&proto.Message{
+		Key:       self,
+		RemoteKey: self,
+		Body:      &proto.Body{Type: proto.Body_HEARTBEAT},
+	})
+}
+
 // receive receives messages from other peers coming through the Signal Exchange
 // and distributes them to worker threads for processing
 func (c *GrpcClient) receive(stream proto.SignalExchange_ConnectStreamClient) error {
@@ -419,6 +509,9 @@ func (c *GrpcClient) receive(stream proto.SignalExchange_ConnectStreamClient) er
 			return err
 		}
 
+		// Any frame from the server proves the receive direction is alive.
+		c.markReceived()
+
 		if msg == nil {
 			continue
 		}

shared/signal/client/watchdog_test.go

@@ -0,0 +1,84 @@
+package client
+
+import (
+	"context"
+	"net"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/require"
+	"go.opentelemetry.io/otel"
+	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
+	"google.golang.org/grpc"
+
+	sigProto "github.com/netbirdio/netbird/shared/signal/proto"
+	"github.com/netbirdio/netbird/signal/server"
+)
+
+func startTestSignalServer(t *testing.T) string {
+	t.Helper()
+
+	lis, err := net.Listen("tcp", "127.0.0.1:0")
+	require.NoError(t, err)
+
+	s := grpc.NewServer()
+	srv, err := server.NewServer(context.Background(), otel.Meter(""))
+	require.NoError(t, err)
+	sigProto.RegisterSignalExchangeServer(s, srv)
+
+	go func() {
+		_ = s.Serve(lis)
+	}()
+	t.Cleanup(s.Stop)
+
+	return lis.Addr().String()
+}
+
+// TestReceiveProbeRoundTrips verifies that the watchdog's self-addressed heartbeat
+// is routed back to the same client through the signal server. This round-trip is
+// what lets the watchdog confirm the receive direction is still delivering.
+func TestReceiveProbeRoundTrips(t *testing.T) {
+	addr := startTestSignalServer(t)
+
+	key, err := wgtypes.GenerateKey()
+	require.NoError(t, err)
+
+	ctx, cancel := context.WithCancel(context.Background())
+	t.Cleanup(cancel)
+
+	client, err := NewClient(ctx, addr, key, false)
+	require.NoError(t, err)
+	t.Cleanup(func() { _ = client.Close() })
+
+	received := make(chan struct{}, 1)
+	go func() {
+		_ = client.Receive(ctx, func(msg *sigProto.Message) error {
+			if msg.GetBody().GetType() == sigProto.Body_HEARTBEAT && msg.GetKey() == key.PublicKey().String() {
+				select {
+				case received <- struct{}{}:
+				default:
+				}
+			}
+			return nil
+		})
+	}()
+
+	streamReady := make(chan struct{})
+	go func() {
+		client.WaitStreamConnected()
+		close(streamReady)
+	}()
+	select {
+	case <-streamReady:
+	case <-time.After(5 * time.Second):
+		t.Fatal("signal stream did not connect within timeout")
+	}
+
+	require.NoError(t, client.sendReceiveProbe())
+
+	select {
+	case <-received:
+	case <-time.After(3 * time.Second):
+		t.Fatal("self-addressed heartbeat did not round-trip back through the signal server")
+	}
+}

shared/signal/proto/signalexchange.pb.go

@@ -30,6 +30,7 @@ const (
 	Body_CANDIDATE Body_Type = 2
 	Body_MODE      Body_Type = 4
 	Body_GO_IDLE   Body_Type = 5
+	Body_HEARTBEAT Body_Type = 6
 )
 
 // Enum value maps for Body_Type.
@@ -40,6 +41,7 @@ var (
 		2: "CANDIDATE",
 		4: "MODE",
 		5: "GO_IDLE",
+		6: "HEARTBEAT",
 	}
 	Body_Type_value = map[string]int32{
 		"OFFER":     0,
@@ -47,6 +49,7 @@ var (
 		"CANDIDATE": 2,
 		"MODE":      4,
 		"GO_IDLE":   5,
+		"HEARTBEAT": 6,
 	}
 )
 
@@ -463,7 +466,7 @@ var file_signalexchange_proto_rawDesc = []byte{
 	0x52, 0x09, 0x72, 0x65, 0x6d, 0x6f, 0x74, 0x65, 0x4b, 0x65, 0x79, 0x12, 0x28, 0x0a, 0x04, 0x62,
 	0x6f, 0x64, 0x79, 0x18, 0x04, 0x20, 0x01, 0x28, 0x0b, 0x32, 0x14, 0x2e, 0x73, 0x69, 0x67, 0x6e,
 	0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x42, 0x6f, 0x64, 0x79, 0x52,
-	0x04, 0x62, 0x6f, 0x64, 0x79, 0x22, 0xc3, 0x04, 0x0a, 0x04, 0x42, 0x6f, 0x64, 0x79, 0x12, 0x2d,
+	0x04, 0x62, 0x6f, 0x64, 0x79, 0x22, 0xd2, 0x04, 0x0a, 0x04, 0x42, 0x6f, 0x64, 0x79, 0x12, 0x2d,
 	0x0a, 0x04, 0x74, 0x79, 0x70, 0x65, 0x18, 0x01, 0x20, 0x01, 0x28, 0x0e, 0x32, 0x19, 0x2e, 0x73,
 	0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x42, 0x6f,
 	0x64, 0x79, 0x2e, 0x54, 0x79, 0x70, 0x65, 0x52, 0x04, 0x74, 0x79, 0x70, 0x65, 0x12, 0x18, 0x0a,
@@ -491,38 +494,39 @@ var file_signalexchange_proto_rawDesc = []byte{
 	0x52, 0x09, 0x73, 0x65, 0x73, 0x73, 0x69, 0x6f, 0x6e, 0x49, 0x64, 0x88, 0x01, 0x01, 0x12, 0x29,
 	0x0a, 0x0d, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x49, 0x50, 0x18,
 	0x0b, 0x20, 0x01, 0x28, 0x0c, 0x48, 0x02, 0x52, 0x0d, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65,
-	0x72, 0x76, 0x65, 0x72, 0x49, 0x50, 0x88, 0x01, 0x01, 0x22, 0x43, 0x0a, 0x04, 0x54, 0x79, 0x70,
+	0x72, 0x76, 0x65, 0x72, 0x49, 0x50, 0x88, 0x01, 0x01, 0x22, 0x52, 0x0a, 0x04, 0x54, 0x79, 0x70,
 	0x65, 0x12, 0x09, 0x0a, 0x05, 0x4f, 0x46, 0x46, 0x45, 0x52, 0x10, 0x00, 0x12, 0x0a, 0x0a, 0x06,
 	0x41, 0x4e, 0x53, 0x57, 0x45, 0x52, 0x10, 0x01, 0x12, 0x0d, 0x0a, 0x09, 0x43, 0x41, 0x4e, 0x44,
 	0x49, 0x44, 0x41, 0x54, 0x45, 0x10, 0x02, 0x12, 0x08, 0x0a, 0x04, 0x4d, 0x4f, 0x44, 0x45, 0x10,
-	0x04, 0x12, 0x0b, 0x0a, 0x07, 0x47, 0x4f, 0x5f, 0x49, 0x44, 0x4c, 0x45, 0x10, 0x05, 0x42, 0x15,
-	0x0a, 0x13, 0x5f, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64,
-	0x64, 0x72, 0x65, 0x73, 0x73, 0x42, 0x0c, 0x0a, 0x0a, 0x5f, 0x73, 0x65, 0x73, 0x73, 0x69, 0x6f,
-	0x6e, 0x49, 0x64, 0x42, 0x10, 0x0a, 0x0e, 0x5f, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65, 0x72,
-	0x76, 0x65, 0x72, 0x49, 0x50, 0x4a, 0x04, 0x08, 0x09, 0x10, 0x0a, 0x22, 0x2e, 0x0a, 0x04, 0x4d,
-	0x6f, 0x64, 0x65, 0x12, 0x1b, 0x0a, 0x06, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x18, 0x01, 0x20,
-	0x01, 0x28, 0x08, 0x48, 0x00, 0x52, 0x06, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x88, 0x01, 0x01,
-	0x42, 0x09, 0x0a, 0x07, 0x5f, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x22, 0x6d, 0x0a, 0x0f, 0x52,
-	0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x12, 0x28,
-	0x0a, 0x0f, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73, 0x50, 0x75, 0x62, 0x4b, 0x65,
-	0x79, 0x18, 0x01, 0x20, 0x01, 0x28, 0x0c, 0x52, 0x0f, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61,
-	0x73, 0x73, 0x50, 0x75, 0x62, 0x4b, 0x65, 0x79, 0x12, 0x30, 0x0a, 0x13, 0x72, 0x6f, 0x73, 0x65,
-	0x6e, 0x70, 0x61, 0x73, 0x73, 0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64, 0x64, 0x72, 0x18,
-	0x02, 0x20, 0x01, 0x28, 0x09, 0x52, 0x13, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73,
-	0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64, 0x64, 0x72, 0x32, 0xb9, 0x01, 0x0a, 0x0e, 0x53,
-	0x69, 0x67, 0x6e, 0x61, 0x6c, 0x45, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x12, 0x4c, 0x0a,
-	0x04, 0x53, 0x65, 0x6e, 0x64, 0x12, 0x20, 0x2e, 0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78,
-	0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x65, 0x64,
-	0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x1a, 0x20, 0x2e, 0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c,
-	0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74,
-	0x65, 0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x22, 0x00, 0x12, 0x59, 0x0a, 0x0d, 0x43,
-	0x6f, 0x6e, 0x6e, 0x65, 0x63, 0x74, 0x53, 0x74, 0x72, 0x65, 0x61, 0x6d, 0x12, 0x20, 0x2e, 0x73,
-	0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e,
-	0x63, 0x72, 0x79, 0x70, 0x74, 0x65, 0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x1a, 0x20,
-	0x2e, 0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e,
-	0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x65, 0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65,
-	0x22, 0x00, 0x28, 0x01, 0x30, 0x01, 0x42, 0x08, 0x5a, 0x06, 0x2f, 0x70, 0x72, 0x6f, 0x74, 0x6f,
-	0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+	0x04, 0x12, 0x0b, 0x0a, 0x07, 0x47, 0x4f, 0x5f, 0x49, 0x44, 0x4c, 0x45, 0x10, 0x05, 0x12, 0x0d,
+	0x0a, 0x09, 0x48, 0x45, 0x41, 0x52, 0x54, 0x42, 0x45, 0x41, 0x54, 0x10, 0x06, 0x42, 0x15, 0x0a,
+	0x13, 0x5f, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64, 0x64,
+	0x72, 0x65, 0x73, 0x73, 0x42, 0x0c, 0x0a, 0x0a, 0x5f, 0x73, 0x65, 0x73, 0x73, 0x69, 0x6f, 0x6e,
+	0x49, 0x64, 0x42, 0x10, 0x0a, 0x0e, 0x5f, 0x72, 0x65, 0x6c, 0x61, 0x79, 0x53, 0x65, 0x72, 0x76,
+	0x65, 0x72, 0x49, 0x50, 0x4a, 0x04, 0x08, 0x09, 0x10, 0x0a, 0x22, 0x2e, 0x0a, 0x04, 0x4d, 0x6f,
+	0x64, 0x65, 0x12, 0x1b, 0x0a, 0x06, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x18, 0x01, 0x20, 0x01,
+	0x28, 0x08, 0x48, 0x00, 0x52, 0x06, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x88, 0x01, 0x01, 0x42,
+	0x09, 0x0a, 0x07, 0x5f, 0x64, 0x69, 0x72, 0x65, 0x63, 0x74, 0x22, 0x6d, 0x0a, 0x0f, 0x52, 0x6f,
+	0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73, 0x43, 0x6f, 0x6e, 0x66, 0x69, 0x67, 0x12, 0x28, 0x0a,
+	0x0f, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73, 0x50, 0x75, 0x62, 0x4b, 0x65, 0x79,
+	0x18, 0x01, 0x20, 0x01, 0x28, 0x0c, 0x52, 0x0f, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73,
+	0x73, 0x50, 0x75, 0x62, 0x4b, 0x65, 0x79, 0x12, 0x30, 0x0a, 0x13, 0x72, 0x6f, 0x73, 0x65, 0x6e,
+	0x70, 0x61, 0x73, 0x73, 0x53, 0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64, 0x64, 0x72, 0x18, 0x02,
+	0x20, 0x01, 0x28, 0x09, 0x52, 0x13, 0x72, 0x6f, 0x73, 0x65, 0x6e, 0x70, 0x61, 0x73, 0x73, 0x53,
+	0x65, 0x72, 0x76, 0x65, 0x72, 0x41, 0x64, 0x64, 0x72, 0x32, 0xb9, 0x01, 0x0a, 0x0e, 0x53, 0x69,
+	0x67, 0x6e, 0x61, 0x6c, 0x45, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x12, 0x4c, 0x0a, 0x04,
+	0x53, 0x65, 0x6e, 0x64, 0x12, 0x20, 0x2e, 0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63,
+	0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x65, 0x64, 0x4d,
+	0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x1a, 0x20, 0x2e, 0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65,
+	0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x65,
+	0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x22, 0x00, 0x12, 0x59, 0x0a, 0x0d, 0x43, 0x6f,
+	0x6e, 0x6e, 0x65, 0x63, 0x74, 0x53, 0x74, 0x72, 0x65, 0x61, 0x6d, 0x12, 0x20, 0x2e, 0x73, 0x69,
+	0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45, 0x6e, 0x63,
+	0x72, 0x79, 0x70, 0x74, 0x65, 0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x1a, 0x20, 0x2e,
+	0x73, 0x69, 0x67, 0x6e, 0x61, 0x6c, 0x65, 0x78, 0x63, 0x68, 0x61, 0x6e, 0x67, 0x65, 0x2e, 0x45,
+	0x6e, 0x63, 0x72, 0x79, 0x70, 0x74, 0x65, 0x64, 0x4d, 0x65, 0x73, 0x73, 0x61, 0x67, 0x65, 0x22,
+	0x00, 0x28, 0x01, 0x30, 0x01, 0x42, 0x08, 0x5a, 0x06, 0x2f, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x62,
+	0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
 }
 
 var (

shared/signal/proto/signalexchange.proto

@@ -48,6 +48,7 @@ message Body {
     CANDIDATE = 2;
     MODE = 4;
     GO_IDLE = 5;
+    HEARTBEAT = 6;
   }
   Type type = 1;
   string payload = 2;

signal/Dockerfile

@@ -1,4 +1,5 @@
 FROM gcr.io/distroless/base:debug
 ENTRYPOINT [ "/go/bin/netbird-signal","run" ]
 CMD ["--log-file", "console"]
-COPY netbird-signal /go/bin/netbird-signal
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-signal /go/bin/netbird-signal

signal/peer/peer.go

@@ -26,6 +26,10 @@ type Peer struct {
 
 	// a gRpc connection stream to the Peer
 	Stream proto.SignalExchange_ConnectStreamServer
+	// sendMu serializes writes to Stream. gRPC forbids concurrent SendMsg on
+	// the same ServerStream, and a peer can be the target of many senders at
+	// once.
+	sendMu sync.Mutex
 
 	// registration time
 	RegisteredAt time.Time
@@ -33,6 +37,13 @@ type Peer struct {
 	Cancel context.CancelFunc
 }
 
+// Send writes a message to the peer's stream, serializing concurrent senders.
+func (p *Peer) Send(msg *proto.EncryptedMessage) error {
+	p.sendMu.Lock()
+	defer p.sendMu.Unlock()
+	return p.Stream.Send(msg)
+}
+
 // NewPeer creates a new instance of a connected Peer
 func NewPeer(id string, stream proto.SignalExchange_ConnectStreamServer, cancel context.CancelFunc) *Peer {
 	return &Peer{

signal/server/concurrent_send_test.go

@@ -0,0 +1,67 @@
+package server
+
+import (
+	"context"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/require"
+	"go.opentelemetry.io/otel"
+
+	"github.com/netbirdio/netbird/shared/signal/proto"
+	"github.com/netbirdio/netbird/signal/peer"
+)
+
+// concurrencyCheckStream records the maximum number of Send calls in flight at
+// once. gRPC forbids concurrent SendMsg on the same ServerStream, so a correct
+// server must never have more than one in flight per peer.
+type concurrencyCheckStream struct {
+	proto.SignalExchange_ConnectStreamServer
+	ctx      context.Context
+	inflight atomic.Int32
+	maxSeen  atomic.Int32
+}
+
+func (s *concurrencyCheckStream) Send(*proto.EncryptedMessage) error {
+	n := s.inflight.Add(1)
+	for {
+		old := s.maxSeen.Load()
+		if n <= old || s.maxSeen.CompareAndSwap(old, n) {
+			break
+		}
+	}
+	// Widen the window so overlapping callers are reliably observed.
+	time.Sleep(time.Millisecond)
+	s.inflight.Add(-1)
+	return nil
+}
+
+func (s *concurrencyCheckStream) Context() context.Context { return s.ctx }
+
+// TestForwardMessageToPeerSerializesSend verifies that concurrent forwards to the
+// same peer never call Stream.Send concurrently, which would violate the gRPC
+// ServerStream contract.
+func TestForwardMessageToPeerSerializesSend(t *testing.T) {
+	s, err := NewServer(context.Background(), otel.Meter(""))
+	require.NoError(t, err)
+
+	const peerID = "peerX"
+	stream := &concurrencyCheckStream{ctx: context.Background()}
+	_, cancel := context.WithCancel(context.Background())
+	t.Cleanup(cancel)
+	require.NoError(t, s.registry.Register(peer.NewPeer(peerID, stream, cancel)))
+
+	var wg sync.WaitGroup
+	for i := 0; i < 50; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			s.forwardMessageToPeer(context.Background(), &proto.EncryptedMessage{Key: "sender", RemoteKey: peerID})
+		}()
+	}
+	wg.Wait()
+
+	require.Equal(t, int32(1), stream.maxSeen.Load(), "Stream.Send must never run concurrently on the same peer stream")
+}

signal/server/signal.go

@@ -179,7 +179,7 @@ func (s *Server) forwardMessageToPeer(ctx context.Context, msg *proto.EncryptedM
 	sendResultChan := make(chan error, 1)
 	go func() {
 		select {
-		case sendResultChan <- dstPeer.Stream.Send(msg):
+		case sendResultChan <- dstPeer.Send(msg):
 			return
 		case <-dstPeer.Stream.Context().Done():
 			return

upload-server/Dockerfile

@@ -1,3 +1,4 @@
 FROM gcr.io/distroless/base:debug
 ENTRYPOINT [ "/go/bin/netbird-upload" ]
-COPY netbird-upload /go/bin/netbird-upload
+ARG TARGETPLATFORM
+COPY ${TARGETPLATFORM}/netbird-upload /go/bin/netbird-upload

util/log.go

@@ -140,7 +140,12 @@ func newRotatedOutput(logPath string) io.Writer {
 func setGRPCLibLogger(logger *log.Logger) {
 	logOut := logger.Writer()
 	if os.Getenv("GRPC_GO_LOG_SEVERITY_LEVEL") != "info" {
-		grpclog.SetLoggerV2(grpclog.NewLoggerV2(io.Discard, logOut, logOut))
+		// Discard grpc info AND warning logs by default — the warning stream is
+		// dominated by benign connection-retry noise ("addrConn.createTransport
+		// failed", "transport is closing") that surfaces e.g. when the CLI dials
+		// a daemon that is still starting or already gone. Errors are kept. Set
+		// GRPC_GO_LOG_SEVERITY_LEVEL=info to get the full verbose grpc logging.
+		grpclog.SetLoggerV2(grpclog.NewLoggerV2(io.Discard, io.Discard, logOut))
 		return
 	}