Reconcile IPv6 addresses on group membership changes

* Add network map benchmark and correctness test files

* Add tests for network map components correctness and edge cases

* Skip benchmarks in CI and enhance network map test coverage with new helper functions

* Remove legacy network map benchmarks and tests; refactor components-based test coverage for clarity and scalability.

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,14 @@
+blank_issues_enabled: true
+contact_links:
+  - name: Community Support
+    url: https://forum.netbird.io/
+    about: Community support forum
+  - name: Cloud Support
+    url: https://docs.netbird.io/help/report-bug-issues
+    about: Contact us for support
+  - name: Client/Connection Troubleshooting
+    url: https://docs.netbird.io/help/troubleshooting-client
+    about: See our client troubleshooting guide for help addressing common issues
+  - name: Self-host Troubleshooting
+    url: https://docs.netbird.io/selfhosted/troubleshooting
+    about: See our self-host troubleshooting guide for help addressing common issues

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

@@ -31,7 +31,7 @@ jobs:
           while IFS= read -r dir; do
             echo "=== Checking $dir ==="
             # Search for problematic imports, excluding test files
-            RESULTS=$(grep -r "github.com/netbirdio/netbird/\(management\|signal\|relay\|proxy\)" "$dir" --include="*.go" 2>/dev/null | grep -v "_test.go" | grep -v "test_" | grep -v "/test/" || true)
+            RESULTS=$(grep -r "github.com/netbirdio/netbird/\(management\|signal\|relay\|proxy\)" "$dir" --include="*.go" 2>/dev/null | grep -v "_test.go" | grep -v "test_" | grep -v "/test/" | grep -v "tools/idp-migrate/" || true)
             if [ -n "$RESULTS" ]; then
               echo "❌ Found problematic dependencies:"
               echo "$RESULTS"
@@ -88,7 +88,7 @@ jobs:
               IMPORTERS=$(go list -json -deps ./... 2>/dev/null | jq -r "select(.Imports[]? == \"$package\") | .ImportPath")
 
               # Check if any importer is NOT in management/signal/relay
-              BSD_IMPORTER=$(echo "$IMPORTERS" | grep -v "github.com/netbirdio/netbird/\(management\|signal\|relay\|proxy\|combined\)" | head -1)
+              BSD_IMPORTER=$(echo "$IMPORTERS" | grep -v "github.com/netbirdio/netbird/\(management\|signal\|relay\|proxy\|combined\|tools/idp-migrate\)" | head -1)
 
               if [ -n "$BSD_IMPORTER" ]; then
                 echo "❌ $package ($license) is imported by BSD-licensed code: $BSD_IMPORTER"

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

@@ -63,10 +63,15 @@ jobs:
       - run: PsExec64 -s -w ${{ github.workspace }} C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe env -w GOMODCACHE=${{ env.cache }}
       - run: PsExec64 -s -w ${{ github.workspace }} C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe env -w GOCACHE=${{ env.modcache }}
       - run: PsExec64 -s -w ${{ github.workspace }} C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe mod tidy
-      - run: echo "files=$(go list ./... | ForEach-Object { $_ } | Where-Object { $_ -notmatch '/management' } | Where-Object { $_ -notmatch '/relay' } | Where-Object { $_ -notmatch '/signal' } | Where-Object { $_ -notmatch '/proxy' } | Where-Object { $_ -notmatch '/combined' })" >> $env:GITHUB_ENV
+      - name: Generate test script
+        run: |
+          $packages = go list ./... | Where-Object { $_ -notmatch '/management' } | Where-Object { $_ -notmatch '/relay' } | Where-Object { $_ -notmatch '/signal' } | Where-Object { $_ -notmatch '/proxy' } | Where-Object { $_ -notmatch '/combined' }
+          $goExe = "C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe"
+          $cmd = "$goExe test -tags=devcert -timeout 10m -p 1 $($packages -join ' ') > test-out.txt 2>&1"
+          Set-Content -Path "${{ github.workspace }}\run-tests.cmd" -Value $cmd
 
       - name: test
-        run: PsExec64 -s -w ${{ github.workspace }} cmd.exe /c "C:\hostedtoolcache\windows\go\${{ steps.go.outputs.go-version }}\x64\bin\go.exe test -tags=devcert -timeout 10m -p 1 ${{ env.files }} > test-out.txt 2>&1"
+        run: PsExec64 -s -w ${{ github.workspace }} cmd.exe /c "${{ github.workspace }}\run-tests.cmd"
       - name: test output
         if: ${{ always() }}
         run: Get-Content test-out.txt

.github/workflows/golangci-lint.yml

@@ -19,7 +19,7 @@ jobs:
       - name: codespell
         uses: codespell-project/actions-codespell@v2
         with:
-          ignore_words_list: erro,clienta,hastable,iif,groupd,testin,groupe,cros,ans,deriver
+          ignore_words_list: erro,clienta,hastable,iif,groupd,testin,groupe,cros,ans,deriver,te,userA
           skip: go.mod,go.sum,**/proxy/web/**
   golangci:
     strategy:

.github/workflows/pr-title-check.yml

@@ -0,0 +1,51 @@
+name: PR Title Check
+
+on:
+  pull_request:
+    types: [opened, edited, synchronize, reopened]
+
+jobs:
+  check-title:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Validate PR title prefix
+        uses: actions/github-script@v7
+        with:
+          script: |
+            const title = context.payload.pull_request.title;
+            const allowedTags = [
+              'management',
+              'client',
+              'signal',
+              'proxy',
+              'relay',
+              'misc',
+              'infrastructure',
+              'self-hosted',
+              'doc',
+            ];
+
+            const pattern = /^\[([^\]]+)\]\s+.+/;
+            const match = title.match(pattern);
+
+            if (!match) {
+              core.setFailed(
+                `PR title must start with a tag in brackets.\n` +
+                `Example: [client] fix something\n` +
+                `Allowed tags: ${allowedTags.join(', ')}`
+              );
+              return;
+            }
+
+            const tags = match[1].split(',').map(t => t.trim().toLowerCase());
+
+            const invalid = tags.filter(t => !allowedTags.includes(t));
+            if (invalid.length > 0) {
+              core.setFailed(
+                `Invalid tag(s): ${invalid.join(', ')}\n` +
+                `Allowed tags: ${allowedTags.join(', ')}`
+              );
+              return;
+            }
+
+            console.log(`Valid PR title tags: [${tags.join(', ')}]`);

.github/workflows/release.yml

@@ -10,7 +10,7 @@ on:
 
 env:
   SIGN_PIPE_VER: "v0.1.1"
-  GORELEASER_VER: "v2.3.2"
+  GORELEASER_VER: "v2.14.3"
   PRODUCT_NAME: "NetBird"
   COPYRIGHT: "NetBird GmbH"
 
@@ -169,6 +169,14 @@ jobs:
       - name: Install OS build dependencies
         run: sudo apt update && sudo apt install -y -q gcc-arm-linux-gnueabihf gcc-aarch64-linux-gnu
 
+      - name: Decode GPG signing key
+        if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository
+        env:
+          GPG_RPM_PRIVATE_KEY: ${{ secrets.GPG_RPM_PRIVATE_KEY }}
+        run: |
+          echo "$GPG_RPM_PRIVATE_KEY" | base64 -d > /tmp/gpg-rpm-signing-key.asc
+          echo "GPG_RPM_KEY_FILE=/tmp/gpg-rpm-signing-key.asc" >> $GITHUB_ENV
+
       - name: Install goversioninfo
         run: go install github.com/josephspurrier/goversioninfo/cmd/goversioninfo@233067e
       - name: Generate windows syso amd64
@@ -186,18 +194,54 @@ jobs:
           HOMEBREW_TAP_GITHUB_TOKEN: ${{ secrets.HOMEBREW_TAP_GITHUB_TOKEN }}
           UPLOAD_DEBIAN_SECRET: ${{ secrets.PKG_UPLOAD_SECRET }}
           UPLOAD_YUM_SECRET: ${{ secrets.PKG_UPLOAD_SECRET }}
-      - name: Tag and push PR images (amd64 only)
-        if: github.event_name == 'pull_request' && github.event.pull_request.head.repo.full_name == github.repository
+          GPG_RPM_KEY_FILE: ${{ env.GPG_RPM_KEY_FILE }}
+          NFPM_NETBIRD_RPM_PASSPHRASE: ${{ secrets.GPG_RPM_PASSPHRASE }}
+      - name: Verify RPM signatures
         run: |
-          PR_TAG="pr-${{ github.event.pull_request.number }}"
+          docker run --rm -v $(pwd)/dist:/dist fedora:41 bash -c '
+            dnf install -y -q rpm-sign curl >/dev/null 2>&1
+            curl -sSL https://pkgs.netbird.io/yum/repodata/repomd.xml.key -o /tmp/rpm-pub.key
+            rpm --import /tmp/rpm-pub.key
+            echo "=== Verifying RPM signatures ==="
+            for rpm_file in /dist/*amd64*.rpm; do
+              [ -f "$rpm_file" ] || continue
+              echo "--- $(basename $rpm_file) ---"
+              rpm -K "$rpm_file"
+            done
+          '
+      - name: Clean up GPG key
+        if: always()
+        run: rm -f /tmp/gpg-rpm-signing-key.asc
+      - name: Tag and push images (amd64 only)
+        if: |
+          (github.event_name == 'pull_request' && github.event.pull_request.head.repo.full_name == github.repository) ||
+          (github.event_name == 'push' && github.ref == 'refs/heads/main')
+        run: |
+          resolve_tags() {
+            if [[ "${{ github.event_name }}" == "pull_request" ]]; then
+              echo "pr-${{ github.event.pull_request.number }}"
+            else
+              echo "main sha-$(git rev-parse --short HEAD)"
+            fi
+          }
+
+          tag_and_push() {
+            local src="$1" img_name tag dst
+            img_name="${src%%:*}"
+            for tag in $(resolve_tags); do
+              dst="${img_name}:${tag}"
+              echo "Tagging ${src} -> ${dst}"
+              docker tag "$src" "$dst"
+              docker push "$dst"
+            done
+          }
+
+          export -f tag_and_push resolve_tags
+
           echo '${{ steps.goreleaser.outputs.artifacts }}' | \
             jq -r '.[] | select(.type == "Docker Image") | select(.goarch == "amd64") | .name' | \
             grep '^ghcr.io/' | while read -r SRC; do
-              IMG_NAME="${SRC%%:*}"
-              DST="${IMG_NAME}:${PR_TAG}"
-              echo "Tagging ${SRC} -> ${DST}"
-              docker tag "$SRC" "$DST"
-              docker push "$DST"
+              tag_and_push "$SRC"
             done
       - name: upload non tags for debug purposes
         uses: actions/upload-artifact@v4
@@ -265,6 +309,14 @@ jobs:
       - name: Install dependencies
         run: sudo apt update && sudo apt install -y -q libappindicator3-dev gir1.2-appindicator3-0.1 libxxf86vm-dev gcc-mingw-w64-x86-64
 
+      - name: Decode GPG signing key
+        if: github.event_name != 'pull_request' || github.event.pull_request.head.repo.full_name == github.repository
+        env:
+          GPG_RPM_PRIVATE_KEY: ${{ secrets.GPG_RPM_PRIVATE_KEY }}
+        run: |
+          echo "$GPG_RPM_PRIVATE_KEY" | base64 -d > /tmp/gpg-rpm-signing-key.asc
+          echo "GPG_RPM_KEY_FILE=/tmp/gpg-rpm-signing-key.asc" >> $GITHUB_ENV
+
       - name: Install LLVM-MinGW for ARM64 cross-compilation
         run: |
           cd /tmp
@@ -289,6 +341,24 @@ jobs:
           HOMEBREW_TAP_GITHUB_TOKEN: ${{ secrets.HOMEBREW_TAP_GITHUB_TOKEN }}
           UPLOAD_DEBIAN_SECRET: ${{ secrets.PKG_UPLOAD_SECRET }}
           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 }}
+      - name: Verify RPM signatures
+        run: |
+          docker run --rm -v $(pwd)/dist:/dist fedora:41 bash -c '
+            dnf install -y -q rpm-sign curl >/dev/null 2>&1
+            curl -sSL https://pkgs.netbird.io/yum/repodata/repomd.xml.key -o /tmp/rpm-pub.key
+            rpm --import /tmp/rpm-pub.key
+            echo "=== Verifying RPM signatures ==="
+            for rpm_file in /dist/*.rpm; do
+              [ -f "$rpm_file" ] || continue
+              echo "--- $(basename $rpm_file) ---"
+              rpm -K "$rpm_file"
+            done
+          '
+      - name: Clean up GPG key
+        if: always()
+        run: rm -f /tmp/gpg-rpm-signing-key.asc
       - name: upload non tags for debug purposes
         uses: actions/upload-artifact@v4
         with:

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

@@ -61,8 +61,8 @@ jobs:
 
           echo "Size: ${SIZE} bytes (${SIZE_MB} MB)"
 
-          if [ ${SIZE} -gt 57671680 ]; then
-            echo "Wasm binary size (${SIZE_MB}MB) exceeds 55MB limit!"
+          if [ ${SIZE} -gt 58720256 ]; then
+            echo "Wasm binary size (${SIZE_MB}MB) exceeds 56MB limit!"
             exit 1
           fi
 

.gitignore

@@ -33,3 +33,4 @@ infrastructure_files/setup-*.env
 vendor/
 /netbird
 client/netbird-electron/
+management/server/types/testdata/

.goreleaser.yaml

@@ -154,6 +154,26 @@ builds:
       - -s -w -X main.Version={{.Version}} -X main.Commit={{.Commit}} -X main.BuildDate={{.CommitDate}}
     mod_timestamp: "{{ .CommitTimestamp }}"
 
+  - id: netbird-idp-migrate
+    dir: tools/idp-migrate
+    env:
+      - CGO_ENABLED=1
+      - >-
+        {{- if eq .Runtime.Goos "linux" }}
+          {{- if eq .Arch "arm64"}}CC=aarch64-linux-gnu-gcc{{- end }}
+          {{- if eq .Arch "arm"}}CC=arm-linux-gnueabihf-gcc{{- end }}
+        {{- end }}
+    binary: netbird-idp-migrate
+    goos:
+      - linux
+    goarch:
+      - amd64
+      - arm64
+      - arm
+    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 }}"
+
 universal_binaries:
   - id: netbird
 
@@ -166,18 +186,22 @@ archives:
       - netbird-wasm
     name_template: "{{ .ProjectName }}_{{ .Version }}"
     format: binary
+  - id: netbird-idp-migrate
+    builds:
+      - netbird-idp-migrate
+    name_template: "netbird-idp-migrate_{{ .Version }}_{{ .Os }}_{{ .Arch }}"
 
 nfpms:
   - maintainer: Netbird <dev@netbird.io>
     description: Netbird client.
     homepage: https://netbird.io/
-    id: netbird-deb
+    license: BSD-3-Clause
+    id: netbird_deb
     bindir: /usr/bin
     builds:
       - netbird
     formats:
       - deb
-
     scripts:
       postinstall: "release_files/post_install.sh"
       preremove: "release_files/pre_remove.sh"
@@ -185,16 +209,19 @@ nfpms:
   - maintainer: Netbird <dev@netbird.io>
     description: Netbird client.
     homepage: https://netbird.io/
-    id: netbird-rpm
+    license: BSD-3-Clause
+    id: netbird_rpm
     bindir: /usr/bin
     builds:
       - netbird
     formats:
       - rpm
-
     scripts:
       postinstall: "release_files/post_install.sh"
       preremove: "release_files/pre_remove.sh"
+    rpm:
+      signature:
+        key_file: '{{ if index .Env "GPG_RPM_KEY_FILE" }}{{ .Env.GPG_RPM_KEY_FILE }}{{ end }}'
 dockers:
   - image_templates:
       - netbirdio/netbird:{{ .Version }}-amd64
@@ -876,7 +903,7 @@ brews:
 uploads:
   - name: debian
     ids:
-      - netbird-deb
+      - netbird_deb
     mode: archive
     target: https://pkgs.wiretrustee.com/debian/pool/{{ .ArtifactName }};deb.distribution=stable;deb.component=main;deb.architecture={{ if .Arm }}armhf{{ else }}{{ .Arch }}{{ end }};deb.package=
     username: dev@wiretrustee.com
@@ -884,7 +911,7 @@ uploads:
 
   - name: yum
     ids:
-      - netbird-rpm
+      - netbird_rpm
     mode: archive
     target: https://pkgs.wiretrustee.com/yum/{{ .Arch }}{{ if .Arm }}{{ .Arm }}{{ end }}
     username: dev@wiretrustee.com

.goreleaser_ui.yaml

@@ -61,7 +61,7 @@ nfpms:
   - maintainer: Netbird <dev@netbird.io>
     description: Netbird client UI.
     homepage: https://netbird.io/
-    id: netbird-ui-deb
+    id: netbird_ui_deb
     package_name: netbird-ui
     builds:
       - netbird-ui
@@ -80,7 +80,7 @@ nfpms:
   - maintainer: Netbird <dev@netbird.io>
     description: Netbird client UI.
     homepage: https://netbird.io/
-    id: netbird-ui-rpm
+    id: netbird_ui_rpm
     package_name: netbird-ui
     builds:
       - netbird-ui
@@ -95,11 +95,14 @@ nfpms:
         dst: /usr/share/pixmaps/netbird.png
     dependencies:
       - netbird
+    rpm:
+      signature:
+        key_file: '{{ if index .Env "GPG_RPM_KEY_FILE" }}{{ .Env.GPG_RPM_KEY_FILE }}{{ end }}'
 
 uploads:
   - name: debian
     ids:
-      - netbird-ui-deb
+      - netbird_ui_deb
     mode: archive
     target: https://pkgs.wiretrustee.com/debian/pool/{{ .ArtifactName }};deb.distribution=stable;deb.component=main;deb.architecture={{ if .Arm }}armhf{{ else }}{{ .Arch }}{{ end }};deb.package=
     username: dev@wiretrustee.com
@@ -107,7 +110,7 @@ uploads:
 
   - name: yum
     ids:
-      - netbird-ui-rpm
+      - netbird_ui_rpm
     mode: archive
     target: https://pkgs.wiretrustee.com/yum/{{ .Arch }}{{ if .Arm }}{{ .Arm }}{{ end }}
     username: dev@wiretrustee.com

CONTRIBUTOR_LICENSE_AGREEMENT.md

@@ -1,7 +1,7 @@
 ##  Contributor License Agreement
 
 This Contributor License Agreement (referred to as the "Agreement") is entered into by the individual 
-submitting this Agreement and NetBird GmbH, c/o Max-Beer-Straße 2-4 Münzstraße 12 10178 Berlin, Germany, 
+submitting this Agreement and NetBird GmbH, Brunnenstraße 196, 10119 Berlin, Germany, 
 referred to as "NetBird" (collectively, the "Parties"). The Agreement outlines the terms and conditions 
 under which NetBird may utilize software contributions provided by the Contributor for inclusion in 
 its software development projects. By submitting this Agreement, the Contributor confirms their acceptance 

README.md

@@ -126,6 +126,7 @@ See a complete [architecture overview](https://docs.netbird.io/about-netbird/how
 ### Community projects
 -  [NetBird installer script](https://github.com/physk/netbird-installer)
 -  [NetBird ansible collection by Dominion Solutions](https://galaxy.ansible.com/ui/repo/published/dominion_solutions/netbird/)
+-  [netbird-tui](https://github.com/n0pashkov/netbird-tui) — terminal UI for managing NetBird peers, routes, and settings
 
 **Note**: The `main` branch may be in an *unstable or even broken state* during development.
 For stable versions, see [releases](https://github.com/netbirdio/netbird/releases).

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:3.23.2
+FROM alpine:3.23.3
 # iproute2: busybox doesn't display ip rules properly
 RUN apk add --no-cache \
     bash \
@@ -17,8 +17,7 @@ ENV \
     NETBIRD_BIN="/usr/local/bin/netbird" \
     NB_LOG_FILE="console,/var/log/netbird/client.log" \
     NB_DAEMON_ADDR="unix:///var/run/netbird.sock" \
-    NB_ENTRYPOINT_SERVICE_TIMEOUT="5" \
-    NB_ENTRYPOINT_LOGIN_TIMEOUT="5"
+    NB_ENTRYPOINT_SERVICE_TIMEOUT="30"
 
 ENTRYPOINT [ "/usr/local/bin/netbird-entrypoint.sh" ]
 

client/Dockerfile-rootless

@@ -23,8 +23,7 @@ ENV \
     NB_DAEMON_ADDR="unix:///var/lib/netbird/netbird.sock" \
     NB_LOG_FILE="console,/var/lib/netbird/client.log" \
     NB_DISABLE_DNS="true" \
-    NB_ENTRYPOINT_SERVICE_TIMEOUT="5" \
-    NB_ENTRYPOINT_LOGIN_TIMEOUT="1"
+    NB_ENTRYPOINT_SERVICE_TIMEOUT="30"
 
 ENTRYPOINT [ "/usr/local/bin/netbird-entrypoint.sh" ]
 

client/android/client.go

@@ -124,7 +124,7 @@ func (c *Client) Run(platformFiles PlatformFiles, urlOpener URLOpener, isAndroid
 
 	// todo do not throw error in case of cancelled context
 	ctx = internal.CtxInitState(ctx)
-	c.connectClient = internal.NewConnectClient(ctx, cfg, c.recorder, false)
+	c.connectClient = internal.NewConnectClient(ctx, cfg, c.recorder)
 	return c.connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile)
 }
 
@@ -157,7 +157,7 @@ func (c *Client) RunWithoutLogin(platformFiles PlatformFiles, dns *DNSList, dnsR
 
 	// todo do not throw error in case of cancelled context
 	ctx = internal.CtxInitState(ctx)
-	c.connectClient = internal.NewConnectClient(ctx, cfg, c.recorder, false)
+	c.connectClient = internal.NewConnectClient(ctx, cfg, c.recorder)
 	return c.connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile)
 }
 
@@ -203,10 +203,11 @@ func (c *Client) PeersList() *PeerInfoArray {
 	peerInfos := make([]PeerInfo, len(fullStatus.Peers))
 	for n, p := range fullStatus.Peers {
 		pi := PeerInfo{
-			p.IP,
-			p.FQDN,
-			p.ConnStatus.String(),
-			PeerRoutes{routes: maps.Keys(p.GetRoutes())},
+			IP:         p.IP,
+			IPv6:       p.IPv6,
+			FQDN:       p.FQDN,
+			ConnStatus: int(p.ConnStatus),
+			Routes:     PeerRoutes{routes: maps.Keys(p.GetRoutes())},
 		}
 		peerInfos[n] = pi
 	}
@@ -237,43 +238,84 @@ func (c *Client) Networks() *NetworkArray {
 		return nil
 	}
 
+	routesMap := routeManager.GetClientRoutesWithNetID()
+	v6Merged := route.V6ExitMergeSet(routesMap)
+	resolvedDomains := c.recorder.GetResolvedDomainsStates()
+
 	networkArray := &NetworkArray{
 		items: make([]Network, 0),
 	}
 
-	resolvedDomains := c.recorder.GetResolvedDomainsStates()
-
-	for id, routes := range routeManager.GetClientRoutesWithNetID() {
+	for id, routes := range routesMap {
 		if len(routes) == 0 {
 			continue
 		}
-
-		r := routes[0]
-		domains := c.getNetworkDomainsFromRoute(r, resolvedDomains)
-		netStr := r.Network.String()
-
-		if r.IsDynamic() {
-			netStr = r.Domains.SafeString()
-		}
-
-		routePeer, err := c.recorder.GetPeer(routes[0].Peer)
-		if err != nil {
-			log.Errorf("could not get peer info for %s: %v", routes[0].Peer, err)
+		if _, skip := v6Merged[id]; skip {
 			continue
 		}
-		network := Network{
-			Name:       string(id),
-			Network:    netStr,
-			Peer:       routePeer.FQDN,
-			Status:     routePeer.ConnStatus.String(),
-			IsSelected: routeSelector.IsSelected(id),
-			Domains:    domains,
+
+		network := c.buildNetwork(id, routes, routeSelector.IsSelected(id), resolvedDomains, v6Merged)
+		if network == nil {
+			continue
 		}
-		networkArray.Add(network)
+		networkArray.Add(*network)
 	}
 	return networkArray
 }
 
+func (c *Client) buildNetwork(id route.NetID, routes []*route.Route, selected bool, resolvedDomains map[domain.Domain]peer.ResolvedDomainInfo, v6Merged map[route.NetID]struct{}) *Network {
+	r := routes[0]
+	netStr := r.Network.String()
+	if r.IsDynamic() {
+		netStr = r.Domains.SafeString()
+	}
+
+	routePeer, err := c.findBestRoutePeer(routes)
+	if err != nil {
+		log.Errorf("could not get peer info for route %s: %v", id, err)
+		return nil
+	}
+
+	network := &Network{
+		Name:       string(id),
+		Network:    netStr,
+		Peer:       routePeer.FQDN,
+		Status:     routePeer.ConnStatus.String(),
+		IsSelected: selected,
+		Domains:    c.getNetworkDomainsFromRoute(r, resolvedDomains),
+	}
+
+	if route.IsV4DefaultRoute(r.Network) && route.HasV6ExitPair(id, v6Merged) {
+		network.Network = "0.0.0.0/0, ::/0"
+	}
+
+	return network
+}
+
+// findBestRoutePeer returns the peer actively routing traffic for the given
+// HA route group. Falls back to the first connected peer, then the first peer.
+func (c *Client) findBestRoutePeer(routes []*route.Route) (peer.State, error) {
+	netStr := routes[0].Network.String()
+
+	fullStatus := c.recorder.GetFullStatus()
+	for _, p := range fullStatus.Peers {
+		if _, ok := p.GetRoutes()[netStr]; ok {
+			return p, nil
+		}
+	}
+
+	for _, r := range routes {
+		p, err := c.recorder.GetPeer(r.Peer)
+		if err != nil {
+			continue
+		}
+		if p.ConnStatus == peer.StatusConnected {
+			return p, nil
+		}
+	}
+	return c.recorder.GetPeer(routes[0].Peer)
+}
+
 // OnUpdatedHostDNS update the DNS servers addresses for root zones
 func (c *Client) OnUpdatedHostDNS(list *DNSList) error {
 	dnsServer, err := dns.GetServerDns()

client/android/peer_notifier.go

@@ -2,11 +2,21 @@
 
 package android
 
+import "github.com/netbirdio/netbird/client/internal/peer"
+
+// Connection status constants exported via gomobile.
+const (
+	ConnStatusIdle       = int(peer.StatusIdle)
+	ConnStatusConnecting = int(peer.StatusConnecting)
+	ConnStatusConnected  = int(peer.StatusConnected)
+)
+
 // PeerInfo describe information about the peers. It designed for the UI usage
 type PeerInfo struct {
 	IP         string
+	IPv6       string
 	FQDN       string
-	ConnStatus string // Todo replace to enum
+	ConnStatus int
 	Routes     PeerRoutes
 }
 

client/android/preferences.go

@@ -307,6 +307,24 @@ func (p *Preferences) SetBlockInbound(block bool) {
 	p.configInput.BlockInbound = &block
 }
 
+// GetDisableIPv6 reads disable IPv6 setting from config file
+func (p *Preferences) GetDisableIPv6() (bool, error) {
+	if p.configInput.DisableIPv6 != nil {
+		return *p.configInput.DisableIPv6, nil
+	}
+
+	cfg, err := profilemanager.ReadConfig(p.configInput.ConfigPath)
+	if err != nil {
+		return false, err
+	}
+	return cfg.DisableIPv6, err
+}
+
+// SetDisableIPv6 stores the given value and waits for commit
+func (p *Preferences) SetDisableIPv6(disable bool) {
+	p.configInput.DisableIPv6 = &disable
+}
+
 // Commit writes out the changes to the config file
 func (p *Preferences) Commit() error {
 	_, err := profilemanager.UpdateOrCreateConfig(p.configInput)

client/android/route_command.go

@@ -18,9 +18,12 @@ func executeRouteToggle(id string, manager routemanager.Manager,
 	netID := route.NetID(id)
 	routes := []route.NetID{netID}
 
-	log.Debugf("%s with id: %s", operationName, id)
+	routesMap := manager.GetClientRoutesWithNetID()
+	routes = route.ExpandV6ExitPairs(routes, routesMap)
 
-	if err := routeOperation(routes, maps.Keys(manager.GetClientRoutesWithNetID())); err != nil {
+	log.Debugf("%s with ids: %v", operationName, routes)
+
+	if err := routeOperation(routes, maps.Keys(routesMap)); err != nil {
 		log.Debugf("error when %s: %s", operationName, err)
 		return fmt.Errorf("error %s: %w", operationName, err)
 	}

client/anonymize/anonymize.go

@@ -9,6 +9,7 @@ import (
 	"net/url"
 	"regexp"
 	"slices"
+	"strconv"
 	"strings"
 )
 
@@ -26,8 +27,9 @@ type Anonymizer struct {
 }
 
 func DefaultAddresses() (netip.Addr, netip.Addr) {
-	// 198.51.100.0, 100::
-	return netip.AddrFrom4([4]byte{198, 51, 100, 0}), netip.AddrFrom16([16]byte{0x01})
+	// 198.51.100.0 (RFC 5737 TEST-NET-2), 2001:db8:ffff:: (RFC 3849 documentation, last /48)
+	// The old start 100:: (discard, RFC 6666) is now used for fake IPs on Android.
+	return netip.AddrFrom4([4]byte{198, 51, 100, 0}), netip.MustParseAddr("2001:db8:ffff::")
 }
 
 func NewAnonymizer(startIPv4, startIPv6 netip.Addr) *Anonymizer {
@@ -96,6 +98,11 @@ func (a *Anonymizer) isInAnonymizedRange(ip netip.Addr) bool {
 }
 
 func (a *Anonymizer) AnonymizeIPString(ip string) string {
+	// Handle CIDR notation (e.g. "2001:db8::/32")
+	if prefix, err := netip.ParsePrefix(ip); err == nil {
+		return a.AnonymizeIP(prefix.Addr()).String() + "/" + strconv.Itoa(prefix.Bits())
+	}
+
 	addr, err := netip.ParseAddr(ip)
 	if err != nil {
 		return ip

client/anonymize/anonymize_test.go

@@ -13,7 +13,7 @@ import (
 
 func TestAnonymizeIP(t *testing.T) {
 	startIPv4 := netip.MustParseAddr("198.51.100.0")
-	startIPv6 := netip.MustParseAddr("100::")
+	startIPv6 := netip.MustParseAddr("2001:db8:ffff::")
 	anonymizer := anonymize.NewAnonymizer(startIPv4, startIPv6)
 
 	tests := []struct {
@@ -26,9 +26,9 @@ func TestAnonymizeIP(t *testing.T) {
 		{"Second Public IPv4", "4.3.2.1", "198.51.100.1"},
 		{"Repeated IPv4", "1.2.3.4", "198.51.100.0"},
 		{"Private IPv4", "192.168.1.1", "192.168.1.1"},
-		{"First Public IPv6", "2607:f8b0:4005:805::200e", "100::"},
-		{"Second Public IPv6", "a::b", "100::1"},
-		{"Repeated IPv6", "2607:f8b0:4005:805::200e", "100::"},
+		{"First Public IPv6", "2607:f8b0:4005:805::200e", "2001:db8:ffff::"},
+		{"Second Public IPv6", "a::b", "2001:db8:ffff::1"},
+		{"Repeated IPv6", "2607:f8b0:4005:805::200e", "2001:db8:ffff::"},
 		{"Private IPv6", "fe80::1", "fe80::1"},
 		{"In Range IPv4", "198.51.100.2", "198.51.100.2"},
 	}
@@ -274,17 +274,17 @@ func TestAnonymizeString_IPAddresses(t *testing.T) {
 		{
 			name:   "IPv6 Address",
 			input:  "Access attempted from 2001:db8::ff00:42",
-			expect: "Access attempted from 100::",
+			expect: "Access attempted from 2001:db8:ffff::",
 		},
 		{
 			name:   "IPv6 Address with Port",
 			input:  "Access attempted from [2001:db8::ff00:42]:8080",
-			expect: "Access attempted from [100::]:8080",
+			expect: "Access attempted from [2001:db8:ffff::]:8080",
 		},
 		{
 			name:   "Both IPv4 and IPv6",
 			input:  "IPv4: 142.108.0.1 and IPv6: 2001:db8::ff00:43",
-			expect: "IPv4: 198.51.100.1 and IPv6: 100::1",
+			expect: "IPv4: 198.51.100.1 and IPv6: 2001:db8:ffff::1",
 		},
 	}
 

client/cmd/debug.go

@@ -181,10 +181,11 @@ func runForDuration(cmd *cobra.Command, args []string) error {
 
 	if stateWasDown {
 		if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
-			return fmt.Errorf("failed to up: %v", status.Convert(err).Message())
+			cmd.PrintErrf("Failed to bring service up: %v\n", status.Convert(err).Message())
+		} else {
+			cmd.Println("netbird up")
+			time.Sleep(time.Second * 10)
 		}
-		cmd.Println("netbird up")
-		time.Sleep(time.Second * 10)
 	}
 
 	initialLevelTrace := initialLogLevel.GetLevel() >= proto.LogLevel_TRACE
@@ -198,10 +199,13 @@ func runForDuration(cmd *cobra.Command, args []string) error {
 		cmd.Println("Log level set to trace.")
 	}
 
+	needsRestoreUp := false
 	if _, err := client.Down(cmd.Context(), &proto.DownRequest{}); err != nil {
-		return fmt.Errorf("failed to down: %v", status.Convert(err).Message())
+		cmd.PrintErrf("Failed to bring service down: %v\n", status.Convert(err).Message())
+	} else {
+		needsRestoreUp = !stateWasDown
+		cmd.Println("netbird down")
 	}
-	cmd.Println("netbird down")
 
 	time.Sleep(1 * time.Second)
 
@@ -209,13 +213,15 @@ func runForDuration(cmd *cobra.Command, args []string) error {
 	if _, err := client.SetSyncResponsePersistence(cmd.Context(), &proto.SetSyncResponsePersistenceRequest{
 		Enabled: true,
 	}); err != nil {
-		return fmt.Errorf("failed to enable sync response persistence: %v", status.Convert(err).Message())
+		cmd.PrintErrf("Failed to enable sync response persistence: %v\n", status.Convert(err).Message())
 	}
 
 	if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
-		return fmt.Errorf("failed to up: %v", status.Convert(err).Message())
+		cmd.PrintErrf("Failed to bring service up: %v\n", status.Convert(err).Message())
+	} else {
+		needsRestoreUp = false
+		cmd.Println("netbird up")
 	}
-	cmd.Println("netbird up")
 
 	time.Sleep(3 * time.Second)
 
@@ -261,18 +267,28 @@ func runForDuration(cmd *cobra.Command, args []string) error {
 		return fmt.Errorf("failed to bundle debug: %v", status.Convert(err).Message())
 	}
 
+	if needsRestoreUp {
+		if _, err := client.Up(cmd.Context(), &proto.UpRequest{}); err != nil {
+			cmd.PrintErrf("Failed to restore service up state: %v\n", status.Convert(err).Message())
+		} else {
+			cmd.Println("netbird up (restored)")
+		}
+	}
+
 	if stateWasDown {
 		if _, err := client.Down(cmd.Context(), &proto.DownRequest{}); err != nil {
-			return fmt.Errorf("failed to down: %v", status.Convert(err).Message())
+			cmd.PrintErrf("Failed to restore service down state: %v\n", status.Convert(err).Message())
+		} else {
+			cmd.Println("netbird down")
 		}
-		cmd.Println("netbird down")
 	}
 
 	if !initialLevelTrace {
 		if _, err := client.SetLogLevel(cmd.Context(), &proto.SetLogLevelRequest{Level: initialLogLevel.GetLevel()}); err != nil {
-			return fmt.Errorf("failed to restore log level: %v", status.Convert(err).Message())
+			cmd.PrintErrf("Failed to restore log level: %v\n", status.Convert(err).Message())
+		} else {
+			cmd.Println("Log level restored to", initialLogLevel.GetLevel())
 		}
-		cmd.Println("Log level restored to", initialLogLevel.GetLevel())
 	}
 
 	cmd.Printf("Local file:\n%s\n", resp.GetPath())

client/cmd/expose.go

@@ -0,0 +1,287 @@
+package cmd
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"os"
+	"os/signal"
+	"regexp"
+	"strconv"
+	"strings"
+	"syscall"
+
+	log "github.com/sirupsen/logrus"
+	"github.com/spf13/cobra"
+	"google.golang.org/grpc/status"
+
+	"github.com/netbirdio/netbird/client/internal/expose"
+	"github.com/netbirdio/netbird/client/proto"
+	"github.com/netbirdio/netbird/util"
+)
+
+var pinRegexp = regexp.MustCompile(`^\d{6}$`)
+
+var (
+	exposePin          string
+	exposePassword     string
+	exposeUserGroups   []string
+	exposeDomain       string
+	exposeNamePrefix   string
+	exposeProtocol     string
+	exposeExternalPort uint16
+)
+
+var exposeCmd = &cobra.Command{
+	Use:   "expose <port>",
+	Short: "Expose a local port via the NetBird reverse proxy",
+	Args:  cobra.ExactArgs(1),
+	Example: `  netbird expose --with-password safe-pass 8080
+  netbird expose --protocol tcp 5432
+  netbird expose --protocol tcp --with-external-port 5433 5432
+  netbird expose --protocol tls --with-custom-domain tls.example.com 4443`,
+	RunE: exposeFn,
+}
+
+func init() {
+	exposeCmd.Flags().StringVar(&exposePin, "with-pin", "", "Protect the exposed service with a 6-digit PIN (e.g. --with-pin 123456)")
+	exposeCmd.Flags().StringVar(&exposePassword, "with-password", "", "Protect the exposed service with a password (e.g. --with-password my-secret)")
+	exposeCmd.Flags().StringSliceVar(&exposeUserGroups, "with-user-groups", nil, "Restrict access to specific user groups with SSO (e.g. --with-user-groups devops,Backend)")
+	exposeCmd.Flags().StringVar(&exposeDomain, "with-custom-domain", "", "Custom domain for the exposed service, must be configured to your account (e.g. --with-custom-domain myapp.example.com)")
+	exposeCmd.Flags().StringVar(&exposeNamePrefix, "with-name-prefix", "", "Prefix for the generated service name (e.g. --with-name-prefix my-app)")
+	exposeCmd.Flags().StringVar(&exposeProtocol, "protocol", "http", "Protocol to use: http, https, tcp, udp, or tls (e.g. --protocol tcp)")
+	exposeCmd.Flags().Uint16Var(&exposeExternalPort, "with-external-port", 0, "Public-facing external port on the proxy cluster (defaults to the target port for L4)")
+}
+
+// isClusterProtocol returns true for L4/TLS protocols that reject HTTP-style auth flags.
+func isClusterProtocol(protocol string) bool {
+	switch strings.ToLower(protocol) {
+	case "tcp", "udp", "tls":
+		return true
+	default:
+		return false
+	}
+}
+
+// isPortBasedProtocol returns true for pure port-based protocols (TCP/UDP)
+// where domain display doesn't apply. TLS uses SNI so it has a domain.
+func isPortBasedProtocol(protocol string) bool {
+	switch strings.ToLower(protocol) {
+	case "tcp", "udp":
+		return true
+	default:
+		return false
+	}
+}
+
+// extractPort returns the port portion of a URL like "tcp://host:12345", or
+// falls back to the given default formatted as a string.
+func extractPort(serviceURL string, fallback uint16) string {
+	u := serviceURL
+	if idx := strings.Index(u, "://"); idx != -1 {
+		u = u[idx+3:]
+	}
+	if i := strings.LastIndex(u, ":"); i != -1 {
+		if p := u[i+1:]; p != "" {
+			return p
+		}
+	}
+	return strconv.FormatUint(uint64(fallback), 10)
+}
+
+// resolveExternalPort returns the effective external port, defaulting to the target port.
+func resolveExternalPort(targetPort uint64) uint16 {
+	if exposeExternalPort != 0 {
+		return exposeExternalPort
+	}
+	return uint16(targetPort)
+}
+
+func validateExposeFlags(cmd *cobra.Command, portStr string) (uint64, error) {
+	port, err := strconv.ParseUint(portStr, 10, 32)
+	if err != nil {
+		return 0, fmt.Errorf("invalid port number: %s", portStr)
+	}
+	if port == 0 || port > 65535 {
+		return 0, fmt.Errorf("invalid port number: must be between 1 and 65535")
+	}
+
+	if !isProtocolValid(exposeProtocol) {
+		return 0, fmt.Errorf("unsupported protocol %q: must be http, https, tcp, udp, or tls", exposeProtocol)
+	}
+
+	if isClusterProtocol(exposeProtocol) {
+		if exposePin != "" || exposePassword != "" || len(exposeUserGroups) > 0 {
+			return 0, fmt.Errorf("auth flags (--with-pin, --with-password, --with-user-groups) are not supported for %s protocol", exposeProtocol)
+		}
+	} else if cmd.Flags().Changed("with-external-port") {
+		return 0, fmt.Errorf("--with-external-port is not supported for %s protocol", exposeProtocol)
+	}
+
+	if exposePin != "" && !pinRegexp.MatchString(exposePin) {
+		return 0, fmt.Errorf("invalid pin: must be exactly 6 digits")
+	}
+
+	if cmd.Flags().Changed("with-password") && exposePassword == "" {
+		return 0, fmt.Errorf("password cannot be empty")
+	}
+
+	if cmd.Flags().Changed("with-user-groups") && len(exposeUserGroups) == 0 {
+		return 0, fmt.Errorf("user groups cannot be empty")
+	}
+
+	return port, nil
+}
+
+func isProtocolValid(exposeProtocol string) bool {
+	switch strings.ToLower(exposeProtocol) {
+	case "http", "https", "tcp", "udp", "tls":
+		return true
+	default:
+		return false
+	}
+}
+
+func exposeFn(cmd *cobra.Command, args []string) error {
+	SetFlagsFromEnvVars(rootCmd)
+
+	if err := util.InitLog(logLevel, util.LogConsole); err != nil {
+		log.Errorf("failed initializing log %v", err)
+		return err
+	}
+
+	cmd.Root().SilenceUsage = false
+
+	port, err := validateExposeFlags(cmd, args[0])
+	if err != nil {
+		return err
+	}
+
+	cmd.Root().SilenceUsage = true
+
+	ctx, cancel := context.WithCancel(cmd.Context())
+	defer cancel()
+
+	sigCh := make(chan os.Signal, 1)
+	signal.Notify(sigCh, syscall.SIGINT, syscall.SIGTERM)
+	go func() {
+		<-sigCh
+		cancel()
+	}()
+
+	conn, err := DialClientGRPCServer(ctx, daemonAddr)
+	if err != nil {
+		return fmt.Errorf("connect to daemon: %w", err)
+	}
+	defer func() {
+		if err := conn.Close(); err != nil {
+			log.Debugf("failed to close daemon connection: %v", err)
+		}
+	}()
+
+	client := proto.NewDaemonServiceClient(conn)
+
+	protocol, err := toExposeProtocol(exposeProtocol)
+	if err != nil {
+		return err
+	}
+
+	req := &proto.ExposeServiceRequest{
+		Port:       uint32(port),
+		Protocol:   protocol,
+		Pin:        exposePin,
+		Password:   exposePassword,
+		UserGroups: exposeUserGroups,
+		Domain:     exposeDomain,
+		NamePrefix: exposeNamePrefix,
+	}
+	if isClusterProtocol(exposeProtocol) {
+		req.ListenPort = uint32(resolveExternalPort(port))
+	}
+
+	stream, err := client.ExposeService(ctx, req)
+	if err != nil {
+		return fmt.Errorf("expose service: %v", status.Convert(err).Message())
+	}
+
+	if err := handleExposeReady(cmd, stream, port); err != nil {
+		return err
+	}
+
+	return waitForExposeEvents(cmd, ctx, stream)
+}
+
+func toExposeProtocol(exposeProtocol string) (proto.ExposeProtocol, error) {
+	p, err := expose.ParseProtocolType(exposeProtocol)
+	if err != nil {
+		return 0, fmt.Errorf("invalid protocol: %w", err)
+	}
+
+	switch p {
+	case expose.ProtocolHTTP:
+		return proto.ExposeProtocol_EXPOSE_HTTP, nil
+	case expose.ProtocolHTTPS:
+		return proto.ExposeProtocol_EXPOSE_HTTPS, nil
+	case expose.ProtocolTCP:
+		return proto.ExposeProtocol_EXPOSE_TCP, nil
+	case expose.ProtocolUDP:
+		return proto.ExposeProtocol_EXPOSE_UDP, nil
+	case expose.ProtocolTLS:
+		return proto.ExposeProtocol_EXPOSE_TLS, nil
+	default:
+		return 0, fmt.Errorf("unhandled protocol type: %d", p)
+	}
+}
+
+func handleExposeReady(cmd *cobra.Command, stream proto.DaemonService_ExposeServiceClient, port uint64) error {
+	event, err := stream.Recv()
+	if err != nil {
+		return fmt.Errorf("receive expose event: %v", status.Convert(err).Message())
+	}
+
+	ready, ok := event.Event.(*proto.ExposeServiceEvent_Ready)
+	if !ok {
+		return fmt.Errorf("unexpected expose event: %T", event.Event)
+	}
+	printExposeReady(cmd, ready.Ready, port)
+	return nil
+}
+
+func printExposeReady(cmd *cobra.Command, r *proto.ExposeServiceReady, port uint64) {
+	cmd.Println("Service exposed successfully!")
+	cmd.Printf("  Name:     %s\n", r.ServiceName)
+	if r.ServiceUrl != "" {
+		cmd.Printf("  URL:      %s\n", r.ServiceUrl)
+	}
+	if r.Domain != "" && !isPortBasedProtocol(exposeProtocol) {
+		cmd.Printf("  Domain:   %s\n", r.Domain)
+	}
+	cmd.Printf("  Protocol: %s\n", exposeProtocol)
+	cmd.Printf("  Internal: %d\n", port)
+	if isClusterProtocol(exposeProtocol) {
+		cmd.Printf("  External: %s\n", extractPort(r.ServiceUrl, resolveExternalPort(port)))
+	}
+	if r.PortAutoAssigned && exposeExternalPort != 0 {
+		cmd.Printf("\n  Note: requested port %d was reassigned\n", exposeExternalPort)
+	}
+	cmd.Println()
+	cmd.Println("Press Ctrl+C to stop exposing.")
+}
+
+func waitForExposeEvents(cmd *cobra.Command, ctx context.Context, stream proto.DaemonService_ExposeServiceClient) error {
+	for {
+		_, err := stream.Recv()
+		if err != nil {
+			if ctx.Err() != nil {
+				cmd.Println("\nService stopped.")
+				//nolint:nilerr
+				return nil
+			}
+			if errors.Is(err, io.EOF) {
+				return fmt.Errorf("connection to daemon closed unexpectedly")
+			}
+			return fmt.Errorf("stream error: %w", err)
+		}
+	}
+}

client/cmd/root.go

@@ -22,6 +22,7 @@ import (
 	"google.golang.org/grpc"
 	"google.golang.org/grpc/credentials/insecure"
 
+	daddr "github.com/netbirdio/netbird/client/internal/daemonaddr"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 )
 
@@ -80,6 +81,15 @@ var (
 		Short:        "",
 		Long:         "",
 		SilenceUsage: true,
+		PersistentPreRunE: func(cmd *cobra.Command, args []string) error {
+			SetFlagsFromEnvVars(cmd.Root())
+
+			// Don't resolve for service commands — they create the socket, not connect to it.
+			if !isServiceCmd(cmd) {
+				daemonAddr = daddr.ResolveUnixDaemonAddr(daemonAddr)
+			}
+			return nil
+		},
 	}
 )
 
@@ -144,6 +154,7 @@ func init() {
 	rootCmd.AddCommand(forwardingRulesCmd)
 	rootCmd.AddCommand(debugCmd)
 	rootCmd.AddCommand(profileCmd)
+	rootCmd.AddCommand(exposeCmd)
 
 	networksCMD.AddCommand(routesListCmd)
 	networksCMD.AddCommand(routesSelectCmd, routesDeselectCmd)
@@ -385,7 +396,6 @@ func migrateToNetbird(oldPath, newPath string) bool {
 }
 
 func getClient(cmd *cobra.Command) (*grpc.ClientConn, error) {
-	SetFlagsFromEnvVars(rootCmd)
 	cmd.SetOut(cmd.OutOrStdout())
 
 	conn, err := DialClientGRPCServer(cmd.Context(), daemonAddr)
@@ -398,3 +408,13 @@ func getClient(cmd *cobra.Command) (*grpc.ClientConn, error) {
 
 	return conn, nil
 }
+
+// isServiceCmd returns true if cmd is the "service" command or a child of it.
+func isServiceCmd(cmd *cobra.Command) bool {
+	for c := cmd; c != nil; c = c.Parent() {
+		if c.Name() == "service" {
+			return true
+		}
+	}
+	return false
+}

client/cmd/service.go

@@ -41,7 +41,7 @@ func init() {
 		defaultServiceName = "Netbird"
 	}
 
-	serviceCmd.AddCommand(runCmd, startCmd, stopCmd, restartCmd, svcStatusCmd, installCmd, uninstallCmd, reconfigureCmd)
+	serviceCmd.AddCommand(runCmd, startCmd, stopCmd, restartCmd, svcStatusCmd, installCmd, uninstallCmd, reconfigureCmd, resetParamsCmd)
 	serviceCmd.PersistentFlags().BoolVar(&profilesDisabled, "disable-profiles", false, "Disables profiles feature. If enabled, the client will not be able to change or edit any profile. To persist this setting, use: netbird service install --disable-profiles")
 	serviceCmd.PersistentFlags().BoolVar(&updateSettingsDisabled, "disable-update-settings", false, "Disables update settings feature. If enabled, the client will not be able to change or edit any settings. To persist this setting, use: netbird service install --disable-update-settings")
 

client/cmd/service_controller.go

@@ -103,7 +103,7 @@ func (p *program) Stop(srv service.Service) error {
 
 // Common setup for service control commands
 func setupServiceControlCommand(cmd *cobra.Command, ctx context.Context, cancel context.CancelFunc) (service.Service, error) {
-	SetFlagsFromEnvVars(rootCmd)
+	// rootCmd env vars are already applied by PersistentPreRunE.
 	SetFlagsFromEnvVars(serviceCmd)
 
 	cmd.SetOut(cmd.OutOrStdout())

client/cmd/service_installer.go

@@ -119,6 +119,10 @@ var installCmd = &cobra.Command{
 			return err
 		}
 
+		if err := loadAndApplyServiceParams(cmd); err != nil {
+			cmd.PrintErrf("Warning: failed to load saved service params: %v\n", err)
+		}
+
 		svcConfig, err := createServiceConfigForInstall()
 		if err != nil {
 			return err
@@ -136,6 +140,10 @@ var installCmd = &cobra.Command{
 			return fmt.Errorf("install service: %w", err)
 		}
 
+		if err := saveServiceParams(currentServiceParams()); err != nil {
+			cmd.PrintErrf("Warning: failed to save service params: %v\n", err)
+		}
+
 		cmd.Println("NetBird service has been installed")
 		return nil
 	},
@@ -187,6 +195,10 @@ This command will temporarily stop the service, update its configuration, and re
 			return err
 		}
 
+		if err := loadAndApplyServiceParams(cmd); err != nil {
+			cmd.PrintErrf("Warning: failed to load saved service params: %v\n", err)
+		}
+
 		wasRunning, err := isServiceRunning()
 		if err != nil && !errors.Is(err, ErrGetServiceStatus) {
 			return fmt.Errorf("check service status: %w", err)
@@ -222,6 +234,10 @@ This command will temporarily stop the service, update its configuration, and re
 			return fmt.Errorf("install service with new config: %w", err)
 		}
 
+		if err := saveServiceParams(currentServiceParams()); err != nil {
+			cmd.PrintErrf("Warning: failed to save service params: %v\n", err)
+		}
+
 		if wasRunning {
 			cmd.Println("Starting NetBird service...")
 			if err := s.Start(); err != nil {

client/cmd/service_params.go

@@ -0,0 +1,201 @@
+//go:build !ios && !android
+
+package cmd
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"maps"
+	"os"
+	"path/filepath"
+
+	"github.com/spf13/cobra"
+
+	"github.com/netbirdio/netbird/client/configs"
+	"github.com/netbirdio/netbird/util"
+)
+
+const serviceParamsFile = "service.json"
+
+// serviceParams holds install-time service parameters that persist across
+// uninstall/reinstall cycles. Saved to <stateDir>/service.json.
+type serviceParams struct {
+	LogLevel              string            `json:"log_level"`
+	DaemonAddr            string            `json:"daemon_addr"`
+	ManagementURL         string            `json:"management_url,omitempty"`
+	ConfigPath            string            `json:"config_path,omitempty"`
+	LogFiles              []string          `json:"log_files,omitempty"`
+	DisableProfiles       bool              `json:"disable_profiles,omitempty"`
+	DisableUpdateSettings bool              `json:"disable_update_settings,omitempty"`
+	ServiceEnvVars        map[string]string `json:"service_env_vars,omitempty"`
+}
+
+// serviceParamsPath returns the path to the service params file.
+func serviceParamsPath() string {
+	return filepath.Join(configs.StateDir, serviceParamsFile)
+}
+
+// loadServiceParams reads saved service parameters from disk.
+// Returns nil with no error if the file does not exist.
+func loadServiceParams() (*serviceParams, error) {
+	path := serviceParamsPath()
+
+	data, err := os.ReadFile(path)
+	if err != nil {
+		if os.IsNotExist(err) {
+			return nil, nil //nolint:nilnil
+		}
+		return nil, fmt.Errorf("read service params %s: %w", path, err)
+	}
+
+	var params serviceParams
+	if err := json.Unmarshal(data, &params); err != nil {
+		return nil, fmt.Errorf("parse service params %s: %w", path, err)
+	}
+
+	return &params, nil
+}
+
+// saveServiceParams writes current service parameters to disk atomically
+// with restricted permissions.
+func saveServiceParams(params *serviceParams) error {
+	path := serviceParamsPath()
+	if err := util.WriteJsonWithRestrictedPermission(context.Background(), path, params); err != nil {
+		return fmt.Errorf("save service params: %w", err)
+	}
+	return nil
+}
+
+// currentServiceParams captures the current state of all package-level
+// variables into a serviceParams struct.
+func currentServiceParams() *serviceParams {
+	params := &serviceParams{
+		LogLevel:              logLevel,
+		DaemonAddr:            daemonAddr,
+		ManagementURL:         managementURL,
+		ConfigPath:            configPath,
+		LogFiles:              logFiles,
+		DisableProfiles:       profilesDisabled,
+		DisableUpdateSettings: updateSettingsDisabled,
+	}
+
+	if len(serviceEnvVars) > 0 {
+		parsed, err := parseServiceEnvVars(serviceEnvVars)
+		if err == nil && len(parsed) > 0 {
+			params.ServiceEnvVars = parsed
+		}
+	}
+
+	return params
+}
+
+// loadAndApplyServiceParams loads saved params from disk and applies them
+// to any flags that were not explicitly set.
+func loadAndApplyServiceParams(cmd *cobra.Command) error {
+	params, err := loadServiceParams()
+	if err != nil {
+		return err
+	}
+	applyServiceParams(cmd, params)
+	return nil
+}
+
+// applyServiceParams merges saved parameters into package-level variables
+// for any flag that was not explicitly set by the user (via CLI or env var).
+// Flags that were Changed() are left untouched.
+func applyServiceParams(cmd *cobra.Command, params *serviceParams) {
+	if params == nil {
+		return
+	}
+
+	// For fields with non-empty defaults (log-level, daemon-addr), keep the
+	// != "" guard so that an older service.json missing the field doesn't
+	// clobber the default with an empty string.
+	if !rootCmd.PersistentFlags().Changed("log-level") && params.LogLevel != "" {
+		logLevel = params.LogLevel
+	}
+
+	if !rootCmd.PersistentFlags().Changed("daemon-addr") && params.DaemonAddr != "" {
+		daemonAddr = params.DaemonAddr
+	}
+
+	// For optional fields where empty means "use default", always apply so
+	// that an explicit clear (--management-url "") persists across reinstalls.
+	if !rootCmd.PersistentFlags().Changed("management-url") {
+		managementURL = params.ManagementURL
+	}
+
+	if !rootCmd.PersistentFlags().Changed("config") {
+		configPath = params.ConfigPath
+	}
+
+	if !rootCmd.PersistentFlags().Changed("log-file") {
+		logFiles = params.LogFiles
+	}
+
+	if !serviceCmd.PersistentFlags().Changed("disable-profiles") {
+		profilesDisabled = params.DisableProfiles
+	}
+
+	if !serviceCmd.PersistentFlags().Changed("disable-update-settings") {
+		updateSettingsDisabled = params.DisableUpdateSettings
+	}
+
+	applyServiceEnvParams(cmd, params)
+}
+
+// applyServiceEnvParams merges saved service environment variables.
+// If --service-env was explicitly set, explicit values win on key conflict
+// but saved keys not in the explicit set are carried over.
+// If --service-env was not set, saved env vars are used entirely.
+func applyServiceEnvParams(cmd *cobra.Command, params *serviceParams) {
+	if len(params.ServiceEnvVars) == 0 {
+		return
+	}
+
+	if !cmd.Flags().Changed("service-env") {
+		// No explicit env vars: rebuild serviceEnvVars from saved params.
+		serviceEnvVars = envMapToSlice(params.ServiceEnvVars)
+		return
+	}
+
+	// Explicit env vars were provided: merge saved values underneath.
+	explicit, err := parseServiceEnvVars(serviceEnvVars)
+	if err != nil {
+		cmd.PrintErrf("Warning: parse explicit service env vars for merge: %v\n", err)
+		return
+	}
+
+	merged := make(map[string]string, len(params.ServiceEnvVars)+len(explicit))
+	maps.Copy(merged, params.ServiceEnvVars)
+	maps.Copy(merged, explicit) // explicit wins on conflict
+	serviceEnvVars = envMapToSlice(merged)
+}
+
+var resetParamsCmd = &cobra.Command{
+	Use:   "reset-params",
+	Short: "Remove saved service install parameters",
+	Long:  "Removes the saved service.json file so the next install uses default parameters.",
+	RunE: func(cmd *cobra.Command, args []string) error {
+		path := serviceParamsPath()
+		if err := os.Remove(path); err != nil {
+			if os.IsNotExist(err) {
+				cmd.Println("No saved service parameters found")
+				return nil
+			}
+			return fmt.Errorf("remove service params: %w", err)
+		}
+		cmd.Printf("Removed saved service parameters (%s)\n", path)
+		return nil
+	},
+}
+
+// envMapToSlice converts a map of env vars to a KEY=VALUE slice.
+func envMapToSlice(m map[string]string) []string {
+	s := make([]string, 0, len(m))
+	for k, v := range m {
+		s = append(s, k+"="+v)
+	}
+	return s
+}

client/cmd/service_params_test.go

@@ -0,0 +1,523 @@
+//go:build !ios && !android
+
+package cmd
+
+import (
+	"encoding/json"
+	"go/ast"
+	"go/parser"
+	"go/token"
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+
+	"github.com/spf13/cobra"
+	"github.com/spf13/pflag"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+
+	"github.com/netbirdio/netbird/client/configs"
+)
+
+func TestServiceParamsPath(t *testing.T) {
+	original := configs.StateDir
+	t.Cleanup(func() { configs.StateDir = original })
+
+	configs.StateDir = "/var/lib/netbird"
+	assert.Equal(t, filepath.Join("/var/lib/netbird", "service.json"), serviceParamsPath())
+
+	configs.StateDir = "/custom/state"
+	assert.Equal(t, filepath.Join("/custom/state", "service.json"), serviceParamsPath())
+}
+
+func TestSaveAndLoadServiceParams(t *testing.T) {
+	tmpDir := t.TempDir()
+
+	original := configs.StateDir
+	t.Cleanup(func() { configs.StateDir = original })
+	configs.StateDir = tmpDir
+
+	params := &serviceParams{
+		LogLevel:              "debug",
+		DaemonAddr:            "unix:///var/run/netbird.sock",
+		ManagementURL:         "https://my.server.com",
+		ConfigPath:            "/etc/netbird/config.json",
+		LogFiles:              []string{"/var/log/netbird/client.log", "console"},
+		DisableProfiles:       true,
+		DisableUpdateSettings: false,
+		ServiceEnvVars:        map[string]string{"NB_LOG_FORMAT": "json", "CUSTOM": "val"},
+	}
+
+	err := saveServiceParams(params)
+	require.NoError(t, err)
+
+	// Verify the file exists and is valid JSON.
+	data, err := os.ReadFile(filepath.Join(tmpDir, "service.json"))
+	require.NoError(t, err)
+	assert.True(t, json.Valid(data))
+
+	loaded, err := loadServiceParams()
+	require.NoError(t, err)
+	require.NotNil(t, loaded)
+
+	assert.Equal(t, params.LogLevel, loaded.LogLevel)
+	assert.Equal(t, params.DaemonAddr, loaded.DaemonAddr)
+	assert.Equal(t, params.ManagementURL, loaded.ManagementURL)
+	assert.Equal(t, params.ConfigPath, loaded.ConfigPath)
+	assert.Equal(t, params.LogFiles, loaded.LogFiles)
+	assert.Equal(t, params.DisableProfiles, loaded.DisableProfiles)
+	assert.Equal(t, params.DisableUpdateSettings, loaded.DisableUpdateSettings)
+	assert.Equal(t, params.ServiceEnvVars, loaded.ServiceEnvVars)
+}
+
+func TestLoadServiceParams_FileNotExists(t *testing.T) {
+	tmpDir := t.TempDir()
+
+	original := configs.StateDir
+	t.Cleanup(func() { configs.StateDir = original })
+	configs.StateDir = tmpDir
+
+	params, err := loadServiceParams()
+	assert.NoError(t, err)
+	assert.Nil(t, params)
+}
+
+func TestLoadServiceParams_InvalidJSON(t *testing.T) {
+	tmpDir := t.TempDir()
+
+	original := configs.StateDir
+	t.Cleanup(func() { configs.StateDir = original })
+	configs.StateDir = tmpDir
+
+	err := os.WriteFile(filepath.Join(tmpDir, "service.json"), []byte("not json"), 0600)
+	require.NoError(t, err)
+
+	params, err := loadServiceParams()
+	assert.Error(t, err)
+	assert.Nil(t, params)
+}
+
+func TestCurrentServiceParams(t *testing.T) {
+	origLogLevel := logLevel
+	origDaemonAddr := daemonAddr
+	origManagementURL := managementURL
+	origConfigPath := configPath
+	origLogFiles := logFiles
+	origProfilesDisabled := profilesDisabled
+	origUpdateSettingsDisabled := updateSettingsDisabled
+	origServiceEnvVars := serviceEnvVars
+	t.Cleanup(func() {
+		logLevel = origLogLevel
+		daemonAddr = origDaemonAddr
+		managementURL = origManagementURL
+		configPath = origConfigPath
+		logFiles = origLogFiles
+		profilesDisabled = origProfilesDisabled
+		updateSettingsDisabled = origUpdateSettingsDisabled
+		serviceEnvVars = origServiceEnvVars
+	})
+
+	logLevel = "trace"
+	daemonAddr = "tcp://127.0.0.1:9999"
+	managementURL = "https://mgmt.example.com"
+	configPath = "/tmp/test-config.json"
+	logFiles = []string{"/tmp/test.log"}
+	profilesDisabled = true
+	updateSettingsDisabled = true
+	serviceEnvVars = []string{"FOO=bar", "BAZ=qux"}
+
+	params := currentServiceParams()
+
+	assert.Equal(t, "trace", params.LogLevel)
+	assert.Equal(t, "tcp://127.0.0.1:9999", params.DaemonAddr)
+	assert.Equal(t, "https://mgmt.example.com", params.ManagementURL)
+	assert.Equal(t, "/tmp/test-config.json", params.ConfigPath)
+	assert.Equal(t, []string{"/tmp/test.log"}, params.LogFiles)
+	assert.True(t, params.DisableProfiles)
+	assert.True(t, params.DisableUpdateSettings)
+	assert.Equal(t, map[string]string{"FOO": "bar", "BAZ": "qux"}, params.ServiceEnvVars)
+}
+
+func TestApplyServiceParams_OnlyUnchangedFlags(t *testing.T) {
+	origLogLevel := logLevel
+	origDaemonAddr := daemonAddr
+	origManagementURL := managementURL
+	origConfigPath := configPath
+	origLogFiles := logFiles
+	origProfilesDisabled := profilesDisabled
+	origUpdateSettingsDisabled := updateSettingsDisabled
+	origServiceEnvVars := serviceEnvVars
+	t.Cleanup(func() {
+		logLevel = origLogLevel
+		daemonAddr = origDaemonAddr
+		managementURL = origManagementURL
+		configPath = origConfigPath
+		logFiles = origLogFiles
+		profilesDisabled = origProfilesDisabled
+		updateSettingsDisabled = origUpdateSettingsDisabled
+		serviceEnvVars = origServiceEnvVars
+	})
+
+	// Reset all flags to defaults.
+	logLevel = "info"
+	daemonAddr = "unix:///var/run/netbird.sock"
+	managementURL = ""
+	configPath = "/etc/netbird/config.json"
+	logFiles = []string{"/var/log/netbird/client.log"}
+	profilesDisabled = false
+	updateSettingsDisabled = false
+	serviceEnvVars = nil
+
+	// Reset Changed state on all relevant flags.
+	rootCmd.PersistentFlags().VisitAll(func(f *pflag.Flag) {
+		f.Changed = false
+	})
+	serviceCmd.PersistentFlags().VisitAll(func(f *pflag.Flag) {
+		f.Changed = false
+	})
+
+	// Simulate user explicitly setting --log-level via CLI.
+	logLevel = "warn"
+	require.NoError(t, rootCmd.PersistentFlags().Set("log-level", "warn"))
+
+	saved := &serviceParams{
+		LogLevel:              "debug",
+		DaemonAddr:            "tcp://127.0.0.1:5555",
+		ManagementURL:         "https://saved.example.com",
+		ConfigPath:            "/saved/config.json",
+		LogFiles:              []string{"/saved/client.log"},
+		DisableProfiles:       true,
+		DisableUpdateSettings: true,
+		ServiceEnvVars:        map[string]string{"SAVED_KEY": "saved_val"},
+	}
+
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+	applyServiceParams(cmd, saved)
+
+	// log-level was Changed, so it should keep "warn", not use saved "debug".
+	assert.Equal(t, "warn", logLevel)
+
+	// All other fields were not Changed, so they should use saved values.
+	assert.Equal(t, "tcp://127.0.0.1:5555", daemonAddr)
+	assert.Equal(t, "https://saved.example.com", managementURL)
+	assert.Equal(t, "/saved/config.json", configPath)
+	assert.Equal(t, []string{"/saved/client.log"}, logFiles)
+	assert.True(t, profilesDisabled)
+	assert.True(t, updateSettingsDisabled)
+	assert.Equal(t, []string{"SAVED_KEY=saved_val"}, serviceEnvVars)
+}
+
+func TestApplyServiceParams_BooleanRevertToFalse(t *testing.T) {
+	origProfilesDisabled := profilesDisabled
+	origUpdateSettingsDisabled := updateSettingsDisabled
+	t.Cleanup(func() {
+		profilesDisabled = origProfilesDisabled
+		updateSettingsDisabled = origUpdateSettingsDisabled
+	})
+
+	// Simulate current state where booleans are true (e.g. set by previous install).
+	profilesDisabled = true
+	updateSettingsDisabled = true
+
+	// Reset Changed state so flags appear unset.
+	serviceCmd.PersistentFlags().VisitAll(func(f *pflag.Flag) {
+		f.Changed = false
+	})
+
+	// Saved params have both as false.
+	saved := &serviceParams{
+		DisableProfiles:       false,
+		DisableUpdateSettings: false,
+	}
+
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+	applyServiceParams(cmd, saved)
+
+	assert.False(t, profilesDisabled, "saved false should override current true")
+	assert.False(t, updateSettingsDisabled, "saved false should override current true")
+}
+
+func TestApplyServiceParams_ClearManagementURL(t *testing.T) {
+	origManagementURL := managementURL
+	t.Cleanup(func() { managementURL = origManagementURL })
+
+	managementURL = "https://leftover.example.com"
+
+	// Simulate saved params where management URL was explicitly cleared.
+	saved := &serviceParams{
+		LogLevel:   "info",
+		DaemonAddr: "unix:///var/run/netbird.sock",
+		// ManagementURL intentionally empty: was cleared with --management-url "".
+	}
+
+	rootCmd.PersistentFlags().VisitAll(func(f *pflag.Flag) {
+		f.Changed = false
+	})
+
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+	applyServiceParams(cmd, saved)
+
+	assert.Equal(t, "", managementURL, "saved empty management URL should clear the current value")
+}
+
+func TestApplyServiceParams_NilParams(t *testing.T) {
+	origLogLevel := logLevel
+	t.Cleanup(func() { logLevel = origLogLevel })
+
+	logLevel = "info"
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+
+	// Should be a no-op.
+	applyServiceParams(cmd, nil)
+	assert.Equal(t, "info", logLevel)
+}
+
+func TestApplyServiceEnvParams_MergeExplicitAndSaved(t *testing.T) {
+	origServiceEnvVars := serviceEnvVars
+	t.Cleanup(func() { serviceEnvVars = origServiceEnvVars })
+
+	// Set up a command with --service-env marked as Changed.
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+	require.NoError(t, cmd.Flags().Set("service-env", "EXPLICIT=yes,OVERLAP=explicit"))
+
+	serviceEnvVars = []string{"EXPLICIT=yes", "OVERLAP=explicit"}
+
+	saved := &serviceParams{
+		ServiceEnvVars: map[string]string{
+			"SAVED":   "val",
+			"OVERLAP": "saved",
+		},
+	}
+
+	applyServiceEnvParams(cmd, saved)
+
+	// Parse result for easier assertion.
+	result, err := parseServiceEnvVars(serviceEnvVars)
+	require.NoError(t, err)
+
+	assert.Equal(t, "yes", result["EXPLICIT"])
+	assert.Equal(t, "val", result["SAVED"])
+	// Explicit wins on conflict.
+	assert.Equal(t, "explicit", result["OVERLAP"])
+}
+
+func TestApplyServiceEnvParams_NotChanged(t *testing.T) {
+	origServiceEnvVars := serviceEnvVars
+	t.Cleanup(func() { serviceEnvVars = origServiceEnvVars })
+
+	serviceEnvVars = nil
+
+	cmd := &cobra.Command{}
+	cmd.Flags().StringSlice("service-env", nil, "")
+
+	saved := &serviceParams{
+		ServiceEnvVars: map[string]string{"FROM_SAVED": "val"},
+	}
+
+	applyServiceEnvParams(cmd, saved)
+
+	result, err := parseServiceEnvVars(serviceEnvVars)
+	require.NoError(t, err)
+	assert.Equal(t, map[string]string{"FROM_SAVED": "val"}, result)
+}
+
+// TestServiceParams_FieldsCoveredInFunctions ensures that all serviceParams fields are
+// referenced in both currentServiceParams() and applyServiceParams(). If a new field is
+// added to serviceParams but not wired into these functions, this test fails.
+func TestServiceParams_FieldsCoveredInFunctions(t *testing.T) {
+	fset := token.NewFileSet()
+	file, err := parser.ParseFile(fset, "service_params.go", nil, 0)
+	require.NoError(t, err)
+
+	// Collect all JSON field names from the serviceParams struct.
+	structFields := extractStructJSONFields(t, file, "serviceParams")
+	require.NotEmpty(t, structFields, "failed to find serviceParams struct fields")
+
+	// Collect field names referenced in currentServiceParams and applyServiceParams.
+	currentFields := extractFuncFieldRefs(t, file, "currentServiceParams", structFields)
+	applyFields := extractFuncFieldRefs(t, file, "applyServiceParams", structFields)
+	// applyServiceEnvParams handles ServiceEnvVars indirectly.
+	applyEnvFields := extractFuncFieldRefs(t, file, "applyServiceEnvParams", structFields)
+	for k, v := range applyEnvFields {
+		applyFields[k] = v
+	}
+
+	for _, field := range structFields {
+		assert.Contains(t, currentFields, field,
+			"serviceParams field %q is not captured in currentServiceParams()", field)
+		assert.Contains(t, applyFields, field,
+			"serviceParams field %q is not restored in applyServiceParams()/applyServiceEnvParams()", field)
+	}
+}
+
+// TestServiceParams_BuildArgsCoversAllFlags ensures that buildServiceArguments references
+// all serviceParams fields that should become CLI args. ServiceEnvVars is excluded because
+// it flows through newSVCConfig() EnvVars, not CLI args.
+func TestServiceParams_BuildArgsCoversAllFlags(t *testing.T) {
+	fset := token.NewFileSet()
+	file, err := parser.ParseFile(fset, "service_params.go", nil, 0)
+	require.NoError(t, err)
+
+	structFields := extractStructJSONFields(t, file, "serviceParams")
+	require.NotEmpty(t, structFields)
+
+	installerFile, err := parser.ParseFile(fset, "service_installer.go", nil, 0)
+	require.NoError(t, err)
+
+	// Fields that are handled outside of buildServiceArguments (env vars go through newSVCConfig).
+	fieldsNotInArgs := map[string]bool{
+		"ServiceEnvVars": true,
+	}
+
+	buildFields := extractFuncGlobalRefs(t, installerFile, "buildServiceArguments")
+
+	// Forward: every struct field must appear in buildServiceArguments.
+	for _, field := range structFields {
+		if fieldsNotInArgs[field] {
+			continue
+		}
+		globalVar := fieldToGlobalVar(field)
+		assert.Contains(t, buildFields, globalVar,
+			"serviceParams field %q (global %q) is not referenced in buildServiceArguments()", field, globalVar)
+	}
+
+	// Reverse: every service-related global used in buildServiceArguments must
+	// have a corresponding serviceParams field. This catches a developer adding
+	// a new flag to buildServiceArguments without adding it to the struct.
+	globalToField := make(map[string]string, len(structFields))
+	for _, field := range structFields {
+		globalToField[fieldToGlobalVar(field)] = field
+	}
+	// Identifiers in buildServiceArguments that are not service params
+	// (builtins, boilerplate, loop variables).
+	nonParamGlobals := map[string]bool{
+		"args": true, "append": true, "string": true, "_": true,
+		"logFile": true, // range variable over logFiles
+	}
+	for ref := range buildFields {
+		if nonParamGlobals[ref] {
+			continue
+		}
+		_, inStruct := globalToField[ref]
+		assert.True(t, inStruct,
+			"buildServiceArguments() references global %q which has no corresponding serviceParams field", ref)
+	}
+}
+
+// extractStructJSONFields returns field names from a named struct type.
+func extractStructJSONFields(t *testing.T, file *ast.File, structName string) []string {
+	t.Helper()
+	var fields []string
+	ast.Inspect(file, func(n ast.Node) bool {
+		ts, ok := n.(*ast.TypeSpec)
+		if !ok || ts.Name.Name != structName {
+			return true
+		}
+		st, ok := ts.Type.(*ast.StructType)
+		if !ok {
+			return false
+		}
+		for _, f := range st.Fields.List {
+			if len(f.Names) > 0 {
+				fields = append(fields, f.Names[0].Name)
+			}
+		}
+		return false
+	})
+	return fields
+}
+
+// extractFuncFieldRefs returns which of the given field names appear inside the
+// named function, either as selector expressions (params.FieldName) or as
+// composite literal keys (&serviceParams{FieldName: ...}).
+func extractFuncFieldRefs(t *testing.T, file *ast.File, funcName string, fields []string) map[string]bool {
+	t.Helper()
+	fieldSet := make(map[string]bool, len(fields))
+	for _, f := range fields {
+		fieldSet[f] = true
+	}
+
+	found := make(map[string]bool)
+	fn := findFuncDecl(file, funcName)
+	require.NotNil(t, fn, "function %s not found", funcName)
+
+	ast.Inspect(fn.Body, func(n ast.Node) bool {
+		switch v := n.(type) {
+		case *ast.SelectorExpr:
+			if fieldSet[v.Sel.Name] {
+				found[v.Sel.Name] = true
+			}
+		case *ast.KeyValueExpr:
+			if ident, ok := v.Key.(*ast.Ident); ok && fieldSet[ident.Name] {
+				found[ident.Name] = true
+			}
+		}
+		return true
+	})
+	return found
+}
+
+// extractFuncGlobalRefs returns all identifier names referenced in the named function body.
+func extractFuncGlobalRefs(t *testing.T, file *ast.File, funcName string) map[string]bool {
+	t.Helper()
+	fn := findFuncDecl(file, funcName)
+	require.NotNil(t, fn, "function %s not found", funcName)
+
+	refs := make(map[string]bool)
+	ast.Inspect(fn.Body, func(n ast.Node) bool {
+		if ident, ok := n.(*ast.Ident); ok {
+			refs[ident.Name] = true
+		}
+		return true
+	})
+	return refs
+}
+
+func findFuncDecl(file *ast.File, name string) *ast.FuncDecl {
+	for _, decl := range file.Decls {
+		fn, ok := decl.(*ast.FuncDecl)
+		if ok && fn.Name.Name == name {
+			return fn
+		}
+	}
+	return nil
+}
+
+// fieldToGlobalVar maps serviceParams field names to the package-level variable
+// names used in buildServiceArguments and applyServiceParams.
+func fieldToGlobalVar(field string) string {
+	m := map[string]string{
+		"LogLevel":              "logLevel",
+		"DaemonAddr":            "daemonAddr",
+		"ManagementURL":         "managementURL",
+		"ConfigPath":            "configPath",
+		"LogFiles":              "logFiles",
+		"DisableProfiles":       "profilesDisabled",
+		"DisableUpdateSettings": "updateSettingsDisabled",
+		"ServiceEnvVars":        "serviceEnvVars",
+	}
+	if v, ok := m[field]; ok {
+		return v
+	}
+	// Default: lowercase first letter.
+	return strings.ToLower(field[:1]) + field[1:]
+}
+
+func TestEnvMapToSlice(t *testing.T) {
+	m := map[string]string{"A": "1", "B": "2"}
+	s := envMapToSlice(m)
+	assert.Len(t, s, 2)
+	assert.Contains(t, s, "A=1")
+	assert.Contains(t, s, "B=2")
+}
+
+func TestEnvMapToSlice_Empty(t *testing.T) {
+	s := envMapToSlice(map[string]string{})
+	assert.Empty(t, s)
+}

client/cmd/service_test.go

@@ -4,7 +4,9 @@ import (
 	"context"
 	"fmt"
 	"os"
+	"os/signal"
 	"runtime"
+	"syscall"
 	"testing"
 	"time"
 
@@ -13,6 +15,22 @@ import (
 	"github.com/stretchr/testify/require"
 )
 
+// TestMain intercepts when this test binary is run as a daemon subprocess.
+// On FreeBSD, the rc.d service script runs the binary via daemon(8) -r with
+// "service run ..." arguments. Since the test binary can't handle cobra CLI
+// args, it exits immediately, causing daemon -r to respawn rapidly until
+// hitting the rate limit and exiting. This makes service restart unreliable.
+// Blocking here keeps the subprocess alive until the init system sends SIGTERM.
+func TestMain(m *testing.M) {
+	if len(os.Args) > 2 && os.Args[1] == "service" && os.Args[2] == "run" {
+		sig := make(chan os.Signal, 1)
+		signal.Notify(sig, syscall.SIGTERM, os.Interrupt)
+		<-sig
+		return
+	}
+	os.Exit(m.Run())
+}
+
 const (
 	serviceStartTimeout = 10 * time.Second
 	serviceStopTimeout  = 5 * time.Second
@@ -79,6 +97,34 @@ func TestServiceLifecycle(t *testing.T) {
 	logLevel = "info"
 	daemonAddr = fmt.Sprintf("unix://%s/netbird-test.sock", tempDir)
 
+	// Ensure cleanup even if a subtest fails and Stop/Uninstall subtests don't run.
+	t.Cleanup(func() {
+		cfg, err := newSVCConfig()
+		if err != nil {
+			t.Errorf("cleanup: create service config: %v", err)
+			return
+		}
+		ctxSvc, cancel := context.WithCancel(context.Background())
+		defer cancel()
+		s, err := newSVC(newProgram(ctxSvc, cancel), cfg)
+		if err != nil {
+			t.Errorf("cleanup: create service: %v", err)
+			return
+		}
+
+		// If the subtests already cleaned up, there's nothing to do.
+		if _, err := s.Status(); err != nil {
+			return
+		}
+
+		if err := s.Stop(); err != nil {
+			t.Errorf("cleanup: stop service: %v", err)
+		}
+		if err := s.Uninstall(); err != nil {
+			t.Errorf("cleanup: uninstall service: %v", err)
+		}
+	})
+
 	ctx := context.Background()
 
 	t.Run("Install", func(t *testing.T) {

client/cmd/signer/artifactkey.go

@@ -7,7 +7,7 @@ import (
 
 	"github.com/spf13/cobra"
 
-	"github.com/netbirdio/netbird/client/internal/updatemanager/reposign"
+	"github.com/netbirdio/netbird/client/internal/updater/reposign"
 )
 
 var (

client/cmd/signer/artifactsign.go

@@ -6,7 +6,7 @@ import (
 
 	"github.com/spf13/cobra"
 
-	"github.com/netbirdio/netbird/client/internal/updatemanager/reposign"
+	"github.com/netbirdio/netbird/client/internal/updater/reposign"
 )
 
 const (

client/cmd/signer/revocation.go

@@ -7,7 +7,7 @@ import (
 
 	"github.com/spf13/cobra"
 
-	"github.com/netbirdio/netbird/client/internal/updatemanager/reposign"
+	"github.com/netbirdio/netbird/client/internal/updater/reposign"
 )
 
 const (

client/cmd/signer/rootkey.go

@@ -7,7 +7,7 @@ import (
 
 	"github.com/spf13/cobra"
 
-	"github.com/netbirdio/netbird/client/internal/updatemanager/reposign"
+	"github.com/netbirdio/netbird/client/internal/updater/reposign"
 )
 
 var (

client/cmd/ssh.go

@@ -787,10 +787,10 @@ func isUnixSocket(path string) bool {
 	return strings.HasPrefix(path, "/") || strings.HasPrefix(path, "./")
 }
 
-// normalizeLocalHost converts "*" to "0.0.0.0" for binding to all interfaces.
+// normalizeLocalHost converts "*" to "" for binding to all interfaces (dual-stack).
 func normalizeLocalHost(host string) string {
 	if host == "*" {
-		return "0.0.0.0"
+		return ""
 	}
 	return host
 }

client/cmd/ssh_test.go

@@ -527,10 +527,10 @@ func TestParsePortForward(t *testing.T) {
 		{
 			name:           "wildcard bind all interfaces",
 			spec:           "*:8080:localhost:80",
-			expectedLocal:  "0.0.0.0:8080",
+			expectedLocal:  ":8080",
 			expectedRemote: "localhost:80",
 			expectError:    false,
-			description:    "Wildcard * should bind to all interfaces (0.0.0.0)",
+			description:    "Wildcard * should bind to all interfaces (dual-stack)",
 		},
 		{
 			name:           "wildcard for port only",

client/cmd/status.go

@@ -20,6 +20,7 @@ import (
 var (
 	detailFlag           bool
 	ipv4Flag             bool
+	ipv6Flag             bool
 	jsonFlag             bool
 	yamlFlag             bool
 	ipsFilter            []string
@@ -28,6 +29,7 @@ var (
 	ipsFilterMap         map[string]struct{}
 	prefixNamesFilterMap map[string]struct{}
 	connectionTypeFilter string
+	checkFlag            string
 )
 
 var statusCmd = &cobra.Command{
@@ -44,11 +46,13 @@ func init() {
 	statusCmd.PersistentFlags().BoolVar(&jsonFlag, "json", false, "display detailed status information in json format")
 	statusCmd.PersistentFlags().BoolVar(&yamlFlag, "yaml", false, "display detailed status information in yaml format")
 	statusCmd.PersistentFlags().BoolVar(&ipv4Flag, "ipv4", false, "display only NetBird IPv4 of this peer, e.g., --ipv4 will output 100.64.0.33")
-	statusCmd.MarkFlagsMutuallyExclusive("detail", "json", "yaml", "ipv4")
-	statusCmd.PersistentFlags().StringSliceVar(&ipsFilter, "filter-by-ips", []string{}, "filters the detailed output by a list of one or more IPs, e.g., --filter-by-ips 100.64.0.100,100.64.0.200")
+	statusCmd.PersistentFlags().BoolVar(&ipv6Flag, "ipv6", false, "display only NetBird IPv6 of this peer")
+	statusCmd.MarkFlagsMutuallyExclusive("detail", "json", "yaml", "ipv4", "ipv6")
+	statusCmd.PersistentFlags().StringSliceVar(&ipsFilter, "filter-by-ips", []string{}, "filters the detailed output by a list of one or more IPs (v4 or v6), e.g., --filter-by-ips 100.64.0.100,fd00::1")
 	statusCmd.PersistentFlags().StringSliceVar(&prefixNamesFilter, "filter-by-names", []string{}, "filters the detailed output by a list of one or more peer FQDN or hostnames, e.g., --filter-by-names peer-a,peer-b.netbird.cloud")
 	statusCmd.PersistentFlags().StringVar(&statusFilter, "filter-by-status", "", "filters the detailed output by connection status(idle|connecting|connected), e.g., --filter-by-status connected")
 	statusCmd.PersistentFlags().StringVar(&connectionTypeFilter, "filter-by-connection-type", "", "filters the detailed output by connection type (P2P|Relayed), e.g., --filter-by-connection-type P2P")
+	statusCmd.PersistentFlags().StringVar(&checkFlag, "check", "", "run a health check and exit with code 0 on success, 1 on failure (live|ready|startup)")
 }
 
 func statusFunc(cmd *cobra.Command, args []string) error {
@@ -56,6 +60,10 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 
 	cmd.SetOut(cmd.OutOrStdout())
 
+	if checkFlag != "" {
+		return runHealthCheck(cmd)
+	}
+
 	err := parseFilters()
 	if err != nil {
 		return err
@@ -68,15 +76,17 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 
 	ctx := internal.CtxInitState(cmd.Context())
 
-	resp, err := getStatus(ctx, false)
+	resp, err := getStatus(ctx, true, false)
 	if err != nil {
 		return err
 	}
 
 	status := resp.GetStatus()
 
-	if status == string(internal.StatusNeedsLogin) || status == string(internal.StatusLoginFailed) ||
-		status == string(internal.StatusSessionExpired) {
+	needsAuth := status == string(internal.StatusNeedsLogin) || status == string(internal.StatusLoginFailed) ||
+		status == string(internal.StatusSessionExpired)
+
+	if needsAuth && !jsonFlag && !yamlFlag {
 		cmd.Printf("Daemon status: %s\n\n"+
 			"Run UP command to log in with SSO (interactive login):\n\n"+
 			" netbird up \n\n"+
@@ -93,13 +103,31 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 		return nil
 	}
 
+	if ipv6Flag {
+		ipv6 := resp.GetFullStatus().GetLocalPeerState().GetIpv6()
+		if ipv6 != "" {
+			cmd.Print(parseInterfaceIP(ipv6))
+		}
+		return nil
+	}
+
 	pm := profilemanager.NewProfileManager()
 	var profName string
 	if activeProf, err := pm.GetActiveProfile(); err == nil {
 		profName = activeProf.Name
 	}
 
-	var outputInformationHolder = nbstatus.ConvertToStatusOutputOverview(resp.GetFullStatus(), anonymizeFlag, resp.GetDaemonVersion(), statusFilter, prefixNamesFilter, prefixNamesFilterMap, ipsFilterMap, connectionTypeFilter, profName)
+	var outputInformationHolder = nbstatus.ConvertToStatusOutputOverview(resp.GetFullStatus(), nbstatus.ConvertOptions{
+		Anonymize:            anonymizeFlag,
+		DaemonVersion:        resp.GetDaemonVersion(),
+		DaemonStatus:         nbstatus.ParseDaemonStatus(status),
+		StatusFilter:         statusFilter,
+		PrefixNamesFilter:    prefixNamesFilter,
+		PrefixNamesFilterMap: prefixNamesFilterMap,
+		IPsFilter:            ipsFilterMap,
+		ConnectionTypeFilter: connectionTypeFilter,
+		ProfileName:          profName,
+	})
 	var statusOutputString string
 	switch {
 	case detailFlag:
@@ -121,7 +149,7 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 	return nil
 }
 
-func getStatus(ctx context.Context, shouldRunProbes bool) (*proto.StatusResponse, error) {
+func getStatus(ctx context.Context, fullPeerStatus bool, shouldRunProbes bool) (*proto.StatusResponse, error) {
 	conn, err := DialClientGRPCServer(ctx, daemonAddr)
 	if err != nil {
 		//nolint
@@ -131,7 +159,7 @@ func getStatus(ctx context.Context, shouldRunProbes bool) (*proto.StatusResponse
 	}
 	defer conn.Close()
 
-	resp, err := proto.NewDaemonServiceClient(conn).Status(ctx, &proto.StatusRequest{GetFullPeerStatus: true, ShouldRunProbes: shouldRunProbes})
+	resp, err := proto.NewDaemonServiceClient(conn).Status(ctx, &proto.StatusRequest{GetFullPeerStatus: fullPeerStatus, ShouldRunProbes: shouldRunProbes})
 	if err != nil {
 		return nil, fmt.Errorf("status failed: %v", status.Convert(err).Message())
 	}
@@ -185,6 +213,83 @@ func enableDetailFlagWhenFilterFlag() {
 	}
 }
 
+func runHealthCheck(cmd *cobra.Command) error {
+	check := strings.ToLower(checkFlag)
+	switch check {
+	case "live", "ready", "startup":
+	default:
+		return fmt.Errorf("unknown check %q, must be one of: live, ready, startup", checkFlag)
+	}
+
+	if err := util.InitLog(logLevel, util.LogConsole); err != nil {
+		return fmt.Errorf("init log: %w", err)
+	}
+
+	ctx := internal.CtxInitState(cmd.Context())
+
+	isStartup := check == "startup"
+	resp, err := getStatus(ctx, isStartup, false)
+	if err != nil {
+		return err
+	}
+
+	switch check {
+	case "live":
+		return nil
+	case "ready":
+		return checkReadiness(resp)
+	case "startup":
+		return checkStartup(resp)
+	default:
+		return nil
+	}
+}
+
+func checkReadiness(resp *proto.StatusResponse) error {
+	daemonStatus := internal.StatusType(resp.GetStatus())
+	switch daemonStatus {
+	case internal.StatusIdle, internal.StatusConnecting, internal.StatusConnected:
+		return nil
+	case internal.StatusNeedsLogin, internal.StatusLoginFailed, internal.StatusSessionExpired:
+		return fmt.Errorf("readiness check: daemon status is %s", daemonStatus)
+	default:
+		return fmt.Errorf("readiness check: unexpected daemon status %q", daemonStatus)
+	}
+}
+
+func checkStartup(resp *proto.StatusResponse) error {
+	fullStatus := resp.GetFullStatus()
+	if fullStatus == nil {
+		return fmt.Errorf("startup check: no full status available")
+	}
+
+	if !fullStatus.GetManagementState().GetConnected() {
+		return fmt.Errorf("startup check: management not connected")
+	}
+
+	if !fullStatus.GetSignalState().GetConnected() {
+		return fmt.Errorf("startup check: signal not connected")
+	}
+
+	var relayCount, relaysConnected int
+	for _, r := range fullStatus.GetRelays() {
+		uri := r.GetURI()
+		if !strings.HasPrefix(uri, "rel://") && !strings.HasPrefix(uri, "rels://") {
+			continue
+		}
+		relayCount++
+		if r.GetAvailable() {
+			relaysConnected++
+		}
+	}
+
+	if relayCount > 0 && relaysConnected == 0 {
+		return fmt.Errorf("startup check: no relay servers available (0/%d connected)", relayCount)
+	}
+
+	return nil
+}
+
 func parseInterfaceIP(interfaceIP string) string {
 	ip, _, err := net.ParseCIDR(interfaceIP)
 	if err != nil {

client/cmd/system.go

@@ -8,6 +8,7 @@ const (
 	disableFirewallFlag     = "disable-firewall"
 	blockLANAccessFlag      = "block-lan-access"
 	blockInboundFlag        = "block-inbound"
+	disableIPv6Flag         = "disable-ipv6"
 )
 
 var (
@@ -17,6 +18,7 @@ var (
 	disableFirewall     bool
 	blockLANAccess      bool
 	blockInbound        bool
+	disableIPv6         bool
 )
 
 func init() {
@@ -39,4 +41,7 @@ func init() {
 	upCmd.PersistentFlags().BoolVar(&blockInbound, blockInboundFlag, false,
 		"Block inbound connections. If enabled, the client will not allow any inbound connections to the local machine nor routed networks.\n"+
 			"This overrides any policies received from the management service.")
+
+	upCmd.PersistentFlags().BoolVar(&disableIPv6, disableIPv6Flag, false,
+		"Disable IPv6 overlay. If enabled, the client won't request or use an IPv6 overlay address.")
 }

client/cmd/up.go

@@ -197,7 +197,7 @@ func runInForegroundMode(ctx context.Context, cmd *cobra.Command, activeProf *pr
 	r := peer.NewRecorder(config.ManagementURL.String())
 	r.GetFullStatus()
 
-	connectClient := internal.NewConnectClient(ctx, config, r, false)
+	connectClient := internal.NewConnectClient(ctx, config, r)
 	SetupDebugHandler(ctx, config, r, connectClient, "")
 
 	return connectClient.Run(nil, util.FindFirstLogPath(logFiles))
@@ -430,6 +430,10 @@ func setupSetConfigReq(customDNSAddressConverted []byte, cmd *cobra.Command, pro
 		req.BlockInbound = &blockInbound
 	}
 
+	if cmd.Flag(disableIPv6Flag).Changed {
+		req.DisableIpv6 = &disableIPv6
+	}
+
 	if cmd.Flag(enableLazyConnectionFlag).Changed {
 		req.LazyConnectionEnabled = &lazyConnEnabled
 	}
@@ -547,6 +551,10 @@ func setupConfig(customDNSAddressConverted []byte, cmd *cobra.Command, configFil
 		ic.BlockInbound = &blockInbound
 	}
 
+	if cmd.Flag(disableIPv6Flag).Changed {
+		ic.DisableIPv6 = &disableIPv6
+	}
+
 	if cmd.Flag(enableLazyConnectionFlag).Changed {
 		ic.LazyConnectionEnabled = &lazyConnEnabled
 	}
@@ -661,6 +669,10 @@ func setupLoginRequest(providedSetupKey string, customDNSAddressConverted []byte
 		loginRequest.BlockInbound = &blockInbound
 	}
 
+	if cmd.Flag(disableIPv6Flag).Changed {
+		loginRequest.DisableIpv6 = &disableIPv6
+	}
+
 	if cmd.Flag(enableLazyConnectionFlag).Changed {
 		loginRequest.LazyConnectionEnabled = &lazyConnEnabled
 	}

client/cmd/update_supported.go

@@ -11,7 +11,7 @@ import (
 	log "github.com/sirupsen/logrus"
 	"github.com/spf13/cobra"
 
-	"github.com/netbirdio/netbird/client/internal/updatemanager/installer"
+	"github.com/netbirdio/netbird/client/internal/updater/installer"
 	"github.com/netbirdio/netbird/util"
 )
 

client/embed/embed.go

@@ -14,6 +14,7 @@ import (
 	"github.com/sirupsen/logrus"
 	wgnetstack "golang.zx2c4.com/wireguard/tun/netstack"
 
+	"github.com/netbirdio/netbird/client/iface"
 	"github.com/netbirdio/netbird/client/iface/netstack"
 	"github.com/netbirdio/netbird/client/internal"
 	"github.com/netbirdio/netbird/client/internal/auth"
@@ -21,6 +22,7 @@ import (
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	sshcommon "github.com/netbirdio/netbird/client/ssh"
 	"github.com/netbirdio/netbird/client/system"
+	"github.com/netbirdio/netbird/shared/management/domain"
 	mgmProto "github.com/netbirdio/netbird/shared/management/proto"
 )
 
@@ -31,14 +33,14 @@ var (
 	ErrConfigNotInitialized = errors.New("config not initialized")
 )
 
-// PeerConnStatus is a peer's connection status.
-type PeerConnStatus = peer.ConnStatus
-
 const (
 	// PeerStatusConnected indicates the peer is in connected state.
 	PeerStatusConnected = peer.StatusConnected
 )
 
+// PeerConnStatus is a peer's connection status.
+type PeerConnStatus = peer.ConnStatus
+
 // Client manages a netbird embedded client instance.
 type Client struct {
 	deviceName string
@@ -77,10 +79,20 @@ type Options struct {
 	StatePath string
 	// DisableClientRoutes disables the client routes
 	DisableClientRoutes bool
+	// DisableIPv6 disables IPv6 overlay addressing
+	DisableIPv6 bool
 	// BlockInbound blocks all inbound connections from peers
 	BlockInbound bool
 	// WireguardPort is the port for the WireGuard interface. Use 0 for a random port.
 	WireguardPort *int
+	// MTU is the MTU for the WireGuard interface.
+	// Valid values are in the range 576..8192 bytes.
+	// If non-nil, this value overrides any value stored in the config file.
+	// If nil, the existing config MTU (if non-zero) is preserved; otherwise it defaults to 1280.
+	// Set to a higher value (e.g. 1400) if carrying QUIC or other protocols that require larger datagrams.
+	MTU *uint16
+	// DNSLabels defines additional DNS labels configured in the peer.
+	DNSLabels []string
 }
 
 // validateCredentials checks that exactly one credential type is provided
@@ -112,6 +124,12 @@ func New(opts Options) (*Client, error) {
 		return nil, err
 	}
 
+	if opts.MTU != nil {
+		if err := iface.ValidateMTU(*opts.MTU); err != nil {
+			return nil, fmt.Errorf("invalid MTU: %w", err)
+		}
+	}
+
 	if opts.LogOutput != nil {
 		logrus.SetOutput(opts.LogOutput)
 	}
@@ -140,17 +158,25 @@ func New(opts Options) (*Client, error) {
 		}
 	}
 
+	var err error
+	var parsedLabels domain.List
+	if parsedLabels, err = domain.FromStringList(opts.DNSLabels); err != nil {
+		return nil, fmt.Errorf("invalid dns labels: %w", err)
+	}
+
 	t := true
 	var config *profilemanager.Config
-	var err error
 	input := profilemanager.ConfigInput{
 		ConfigPath:          opts.ConfigPath,
 		ManagementURL:       opts.ManagementURL,
 		PreSharedKey:        &opts.PreSharedKey,
 		DisableServerRoutes: &t,
 		DisableClientRoutes: &opts.DisableClientRoutes,
+		DisableIPv6:         &opts.DisableIPv6,
 		BlockInbound:        &opts.BlockInbound,
 		WireguardPort:       opts.WireguardPort,
+		MTU:                 opts.MTU,
+		DNSLabels:           parsedLabels,
 	}
 	if opts.ConfigPath != "" {
 		config, err = profilemanager.UpdateOrCreateConfig(input)
@@ -202,7 +228,7 @@ 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, false)
+	client := internal.NewConnectClient(ctx, c.config, c.recorder)
 	client.SetSyncResponsePersistence(true)
 
 	// either startup error (permanent backoff err) or nil err (successful engine up)
@@ -352,6 +378,32 @@ func (c *Client) NewHTTPClient() *http.Client {
 	}
 }
 
+// Expose exposes a local service via the NetBird reverse proxy, making it accessible through a public URL.
+// It returns an ExposeSession. Call Wait on the session to keep it alive.
+func (c *Client) Expose(ctx context.Context, req ExposeRequest) (*ExposeSession, error) {
+	engine, err := c.getEngine()
+	if err != nil {
+		return nil, err
+	}
+
+	mgr := engine.GetExposeManager()
+	if mgr == nil {
+		return nil, fmt.Errorf("expose manager not available")
+	}
+
+	resp, err := mgr.Expose(ctx, req)
+	if err != nil {
+		return nil, fmt.Errorf("expose: %w", err)
+	}
+
+	return &ExposeSession{
+		Domain:      resp.Domain,
+		ServiceName: resp.ServiceName,
+		ServiceURL:  resp.ServiceURL,
+		mgr:         mgr,
+	}, nil
+}
+
 // Status returns the current status of the client.
 func (c *Client) Status() (peer.FullStatus, error) {
 	c.mu.Lock()

client/embed/expose.go

@@ -0,0 +1,45 @@
+package embed
+
+import (
+	"context"
+	"errors"
+
+	"github.com/netbirdio/netbird/client/internal/expose"
+)
+
+const (
+	// ExposeProtocolHTTP exposes the service as HTTP.
+	ExposeProtocolHTTP = expose.ProtocolHTTP
+	// ExposeProtocolHTTPS exposes the service as HTTPS.
+	ExposeProtocolHTTPS = expose.ProtocolHTTPS
+	// ExposeProtocolTCP exposes the service as TCP.
+	ExposeProtocolTCP = expose.ProtocolTCP
+	// ExposeProtocolUDP exposes the service as UDP.
+	ExposeProtocolUDP = expose.ProtocolUDP
+	// ExposeProtocolTLS exposes the service as TLS.
+	ExposeProtocolTLS = expose.ProtocolTLS
+)
+
+// ExposeRequest is a request to expose a local service via the NetBird reverse proxy.
+type ExposeRequest = expose.Request
+
+// ExposeProtocolType represents the protocol used for exposing a service.
+type ExposeProtocolType = expose.ProtocolType
+
+// ExposeSession represents an active expose session. Use Wait to block until the session ends.
+type ExposeSession struct {
+	Domain      string
+	ServiceName string
+	ServiceURL  string
+
+	mgr *expose.Manager
+}
+
+// Wait blocks while keeping the expose session alive.
+// It returns when ctx is cancelled or a keep-alive error occurs, then terminates the session.
+func (s *ExposeSession) Wait(ctx context.Context) error {
+	if s == nil || s.mgr == nil {
+		return errors.New("expose session is not initialized")
+	}
+	return s.mgr.KeepAlive(ctx, s.Domain)
+}

client/firewall/iptables/acl_linux.go

@@ -36,6 +36,7 @@ type aclManager struct {
 	entries         aclEntries
 	optionalEntries map[string][]entry
 	ipsetStore      *ipsetStore
+	v6              bool
 
 	stateManager *statemanager.Manager
 }
@@ -47,6 +48,7 @@ func newAclManager(iptablesClient *iptables.IPTables, wgIface iFaceMapper) (*acl
 		entries:         make(map[string][][]string),
 		optionalEntries: make(map[string][]entry),
 		ipsetStore:      newIpsetStore(),
+		v6:              iptablesClient.Proto() == iptables.ProtocolIPv6,
 	}, nil
 }
 
@@ -81,7 +83,11 @@ func (m *aclManager) AddPeerFiltering(
 	chain := chainNameInputRules
 
 	ipsetName = transformIPsetName(ipsetName, sPort, dPort, action)
-	specs := filterRuleSpecs(ip, string(protocol), sPort, dPort, action, ipsetName)
+	if m.v6 && ipsetName != "" {
+		ipsetName += "-v6"
+	}
+	proto := protoForFamily(protocol, m.v6)
+	specs := filterRuleSpecs(ip, proto, sPort, dPort, action, ipsetName)
 
 	mangleSpecs := slices.Clone(specs)
 	mangleSpecs = append(mangleSpecs,
@@ -105,6 +111,7 @@ func (m *aclManager) AddPeerFiltering(
 				ip:        ip.String(),
 				chain:     chain,
 				specs:     specs,
+				v6:        m.v6,
 			}}, nil
 		}
 
@@ -157,6 +164,7 @@ func (m *aclManager) AddPeerFiltering(
 		ipsetName:   ipsetName,
 		ip:          ip.String(),
 		chain:       chain,
+		v6:          m.v6,
 	}
 
 	m.updateState()
@@ -376,8 +384,13 @@ func (m *aclManager) updateState() {
 	currentState.Lock()
 	defer currentState.Unlock()
 
-	currentState.ACLEntries = m.entries
-	currentState.ACLIPsetStore = m.ipsetStore
+	if m.v6 {
+		currentState.ACLEntries6 = m.entries
+		currentState.ACLIPsetStore6 = m.ipsetStore
+	} else {
+		currentState.ACLEntries = m.entries
+		currentState.ACLIPsetStore = m.ipsetStore
+	}
 
 	if err := m.stateManager.UpdateState(currentState); err != nil {
 		log.Errorf("failed to update state: %v", err)
@@ -385,6 +398,15 @@ func (m *aclManager) updateState() {
 }
 
 // filterRuleSpecs returns the specs of a filtering rule
+// protoForFamily translates ICMP to ICMPv6 for ip6tables.
+// ip6tables requires "ipv6-icmp" (or "icmpv6") instead of "icmp".
+func protoForFamily(protocol firewall.Protocol, v6 bool) string {
+	if v6 && protocol == firewall.ProtocolICMP {
+		return "ipv6-icmp"
+	}
+	return string(protocol)
+}
+
 func filterRuleSpecs(ip net.IP, protocol string, sPort, dPort *firewall.Port, action firewall.Action, ipsetName string) (specs []string) {
 	// don't use IP matching if IP is 0.0.0.0
 	matchByIP := !ip.IsUnspecified()
@@ -437,6 +459,9 @@ func (m *aclManager) createIPSet(name string) error {
 	opts := ipset.CreateOptions{
 		Replace: true,
 	}
+	if m.v6 {
+		opts.Family = ipset.FamilyIPV6
+	}
 
 	if err := ipset.Create(name, ipset.TypeHashNet, opts); err != nil {
 		return fmt.Errorf("create ipset %s: %w", name, err)

client/firewall/iptables/manager_linux.go

@@ -17,15 +17,25 @@ import (
 	"github.com/netbirdio/netbird/client/internal/statemanager"
 )
 
+type resetter interface {
+	Reset() error
+}
+
 // Manager of iptables firewall
 type Manager struct {
 	mutex sync.Mutex
 
 	wgIface iFaceMapper
 
-	ipv4Client *iptables.IPTables
-	aclMgr     *aclManager
-	router     *router
+	ipv4Client   *iptables.IPTables
+	aclMgr       *aclManager
+	router       *router
+	rawSupported bool
+
+	// IPv6 counterparts, nil when no v6 overlay
+	ipv6Client *iptables.IPTables
+	aclMgr6    *aclManager
+	router6    *router
 }
 
 // iFaceMapper defines subset methods of interface required for manager
@@ -57,9 +67,43 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
 		return nil, fmt.Errorf("create acl manager: %w", err)
 	}
 
+	if wgIface.Address().HasIPv6() {
+		if err := m.createIPv6Components(wgIface, mtu); err != nil {
+			return nil, fmt.Errorf("create IPv6 firewall: %w", err)
+		}
+	}
+
 	return m, nil
 }
 
+func (m *Manager) createIPv6Components(wgIface iFaceMapper, mtu uint16) error {
+	ip6Client, err := iptables.NewWithProtocol(iptables.ProtocolIPv6)
+	if err != nil {
+		return fmt.Errorf("init ip6tables: %w", err)
+	}
+	m.ipv6Client = ip6Client
+
+	m.router6, err = newRouter(ip6Client, wgIface, mtu)
+	if err != nil {
+		return fmt.Errorf("create v6 router: %w", err)
+	}
+
+	// Share the same IP forwarding state with the v4 router, since
+	// EnableIPForwarding controls both v4 and v6 sysctls.
+	m.router6.ipFwdState = m.router.ipFwdState
+
+	m.aclMgr6, err = newAclManager(ip6Client, wgIface)
+	if err != nil {
+		return fmt.Errorf("create v6 acl manager: %w", err)
+	}
+
+	return nil
+}
+
+func (m *Manager) hasIPv6() bool {
+	return m.ipv6Client != nil
+}
+
 func (m *Manager) Init(stateManager *statemanager.Manager) error {
 	state := &ShutdownState{
 		InterfaceState: &InterfaceState{
@@ -74,17 +118,12 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
 		log.Errorf("failed to update state: %v", err)
 	}
 
-	if err := m.router.init(stateManager); err != nil {
-		return fmt.Errorf("router init: %w", err)
-	}
-
-	if err := m.aclMgr.init(stateManager); err != nil {
-		// TODO: cleanup router
-		return fmt.Errorf("acl manager init: %w", err)
+	if err := m.initChains(stateManager); err != nil {
+		return err
 	}
 
 	if err := m.initNoTrackChain(); err != nil {
-		return fmt.Errorf("init notrack chain: %w", err)
+		log.Warnf("raw table not available, notrack rules will be disabled: %v", err)
 	}
 
 	// persist early to ensure cleanup of chains
@@ -97,6 +136,41 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
 	return nil
 }
 
+// initChains initializes router and ACL chains for both address families,
+// rolling back on failure.
+func (m *Manager) initChains(stateManager *statemanager.Manager) error {
+	type initStep struct {
+		name string
+		init func(*statemanager.Manager) error
+		mgr  resetter
+	}
+
+	steps := []initStep{
+		{"router", m.router.init, m.router},
+		{"acl manager", m.aclMgr.init, m.aclMgr},
+	}
+	if m.hasIPv6() {
+		steps = append(steps,
+			initStep{"v6 router", m.router6.init, m.router6},
+			initStep{"v6 acl manager", m.aclMgr6.init, m.aclMgr6},
+		)
+	}
+
+	var initialized []initStep
+	for _, s := range steps {
+		if err := s.init(stateManager); err != nil {
+			for i := len(initialized) - 1; i >= 0; i-- {
+				if rerr := initialized[i].mgr.Reset(); rerr != nil {
+					log.Warnf("rollback %s: %v", initialized[i].name, rerr)
+				}
+			}
+			return fmt.Errorf("%s init: %w", s.name, err)
+		}
+		initialized = append(initialized, s)
+	}
+	return nil
+}
+
 // AddPeerFiltering adds a rule to the firewall
 //
 // Comment will be ignored because some system this feature is not supported
@@ -112,7 +186,13 @@ func (m *Manager) AddPeerFiltering(
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.aclMgr.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
+	if ip.To4() != nil {
+		return m.aclMgr.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
+	}
+	if !m.hasIPv6() {
+		return nil, fmt.Errorf("IPv6 not initialized, cannot add rule for %s", ip)
+	}
+	return m.aclMgr6.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
 }
 
 func (m *Manager) AddRouteFiltering(
@@ -126,25 +206,48 @@ func (m *Manager) AddRouteFiltering(
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	if destination.IsPrefix() && !destination.Prefix.Addr().Is4() {
-		return nil, fmt.Errorf("unsupported IP version: %s", destination.Prefix.Addr().String())
+	if isIPv6RouteRule(sources, destination) {
+		if !m.hasIPv6() {
+			return nil, fmt.Errorf("IPv6 not initialized, cannot add route rule")
+		}
+		return m.router6.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
 	}
 
 	return m.router.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
 }
 
+func isIPv6RouteRule(sources []netip.Prefix, destination firewall.Network) bool {
+	if destination.IsPrefix() {
+		return destination.Prefix.Addr().Is6()
+	}
+	return len(sources) > 0 && sources[0].Addr().Is6()
+}
+
 // DeletePeerRule from the firewall by rule definition
 func (m *Manager) DeletePeerRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && isIPv6IptRule(rule) {
+		return m.aclMgr6.DeletePeerRule(rule)
+	}
 	return m.aclMgr.DeletePeerRule(rule)
 }
 
+func isIPv6IptRule(rule firewall.Rule) bool {
+	r, ok := rule.(*Rule)
+	return ok && r.v6
+}
+
+// DeleteRouteRule deletes a routing rule.
+// Route rules are keyed by content hash. Check v4 first, try v6 if not found.
 func (m *Manager) DeleteRouteRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && !m.router.hasRule(rule.ID()) {
+		return m.router6.DeleteRouteRule(rule)
+	}
 	return m.router.DeleteRouteRule(rule)
 }
 
@@ -160,18 +263,63 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.AddNatRule(pair)
+	if pair.Destination.IsPrefix() && pair.Destination.Prefix.Addr().Is6() {
+		if !m.hasIPv6() {
+			return fmt.Errorf("IPv6 not initialized, cannot add NAT rule")
+		}
+		return m.router6.AddNatRule(pair)
+	}
+
+	if err := m.router.AddNatRule(pair); err != nil {
+		return err
+	}
+
+	// Dynamic routes need NAT in both tables
+	if m.hasIPv6() && pair.Destination.IsSet() {
+		v6Pair := pair
+		v6Pair.Source = firewall.Network{Prefix: netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
+		if err := m.router6.AddNatRule(v6Pair); err != nil {
+			return fmt.Errorf("add v6 NAT rule: %w", err)
+		}
+	}
+
+	return nil
 }
 
 func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.RemoveNatRule(pair)
+	if pair.Destination.IsPrefix() && pair.Destination.Prefix.Addr().Is6() {
+		if !m.hasIPv6() {
+			return nil
+		}
+		return m.router6.RemoveNatRule(pair)
+	}
+
+	if err := m.router.RemoveNatRule(pair); err != nil {
+		return err
+	}
+
+	if m.hasIPv6() && pair.Destination.IsSet() {
+		v6Pair := pair
+		v6Pair.Source = firewall.Network{Prefix: netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
+		if err := m.router6.RemoveNatRule(v6Pair); err != nil {
+			return fmt.Errorf("remove v6 NAT rule: %w", err)
+		}
+	}
+
+	return nil
 }
 
 func (m *Manager) SetLegacyManagement(isLegacy bool) error {
-	return firewall.SetLegacyManagement(m.router, isLegacy)
+	if err := firewall.SetLegacyManagement(m.router, isLegacy); err != nil {
+		return err
+	}
+	if m.hasIPv6() {
+		return firewall.SetLegacyManagement(m.router6, isLegacy)
+	}
+	return nil
 }
 
 // Reset firewall to the default state
@@ -185,6 +333,15 @@ func (m *Manager) Close(stateManager *statemanager.Manager) error {
 		merr = multierror.Append(merr, fmt.Errorf("cleanup notrack chain: %w", err))
 	}
 
+	if m.hasIPv6() {
+		if err := m.aclMgr6.Reset(); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("reset v6 acl manager: %w", err))
+		}
+		if err := m.router6.Reset(); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("reset v6 router: %w", err))
+		}
+	}
+
 	if err := m.aclMgr.Reset(); err != nil {
 		merr = multierror.Append(merr, fmt.Errorf("reset acl manager: %w", err))
 	}
@@ -208,19 +365,16 @@ func (m *Manager) AllowNetbird() error {
 		return nil
 	}
 
-	_, err := m.AddPeerFiltering(
-		nil,
-		net.IP{0, 0, 0, 0},
-		firewall.ProtocolALL,
-		nil,
-		nil,
-		firewall.ActionAccept,
-		"",
-	)
-	if err != nil {
-		return fmt.Errorf("allow netbird interface traffic: %w", err)
+	var merr *multierror.Error
+	if _, err := m.aclMgr.AddPeerFiltering(nil, net.IP{0, 0, 0, 0}, firewall.ProtocolALL, nil, nil, firewall.ActionAccept, ""); err != nil {
+		merr = multierror.Append(merr, fmt.Errorf("allow netbird interface traffic: %w", err))
 	}
-	return nil
+	if m.hasIPv6() {
+		if _, err := m.aclMgr6.AddPeerFiltering(nil, net.IPv6zero, firewall.ProtocolALL, nil, nil, firewall.ActionAccept, ""); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("allow v6 netbird interface traffic: %w", err))
+		}
+	}
+	return nberrors.FormatErrorOrNil(merr)
 }
 
 // Flush doesn't need to be implemented for this manager
@@ -250,6 +404,9 @@ func (m *Manager) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error)
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && rule.TranslatedAddress.Is6() {
+		return m.router6.AddDNATRule(rule)
+	}
 	return m.router.AddDNATRule(rule)
 }
 
@@ -258,6 +415,9 @@ func (m *Manager) DeleteDNATRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && !m.router.hasRule(rule.ID()+dnatSuffix) {
+		return m.router6.DeleteDNATRule(rule)
+	}
 	return m.router.DeleteDNATRule(rule)
 }
 
@@ -266,7 +426,26 @@ func (m *Manager) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.UpdateSet(set, prefixes)
+	var v4Prefixes, v6Prefixes []netip.Prefix
+	for _, p := range prefixes {
+		if p.Addr().Is6() {
+			v6Prefixes = append(v6Prefixes, p)
+		} else {
+			v4Prefixes = append(v4Prefixes, p)
+		}
+	}
+
+	if err := m.router.UpdateSet(set, v4Prefixes); err != nil {
+		return err
+	}
+
+	if m.hasIPv6() && len(v6Prefixes) > 0 {
+		if err := m.router6.UpdateSet(set, v6Prefixes); err != nil {
+			return fmt.Errorf("update v6 set: %w", err)
+		}
+	}
+
+	return nil
 }
 
 // AddInboundDNAT adds an inbound DNAT rule redirecting traffic from NetBird peers to local services.
@@ -274,6 +453,9 @@ func (m *Manager) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protoco
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && localAddr.Is6() {
+		return m.router6.AddInboundDNAT(localAddr, protocol, sourcePort, targetPort)
+	}
 	return m.router.AddInboundDNAT(localAddr, protocol, sourcePort, targetPort)
 }
 
@@ -282,9 +464,28 @@ func (m *Manager) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Prot
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && localAddr.Is6() {
+		return m.router6.RemoveInboundDNAT(localAddr, protocol, sourcePort, targetPort)
+	}
 	return m.router.RemoveInboundDNAT(localAddr, protocol, sourcePort, targetPort)
 }
 
+// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
+func (m *Manager) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	m.mutex.Lock()
+	defer m.mutex.Unlock()
+
+	return m.router.AddOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
+// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
+func (m *Manager) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	m.mutex.Lock()
+	defer m.mutex.Unlock()
+
+	return m.router.RemoveOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
 const (
 	chainNameRaw = "NETBIRD-RAW"
 	chainOUTPUT  = "OUTPUT"
@@ -318,6 +519,10 @@ func (m *Manager) SetupEBPFProxyNoTrack(proxyPort, wgPort uint16) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if !m.rawSupported {
+		return fmt.Errorf("raw table not available")
+	}
+
 	wgPortStr := fmt.Sprintf("%d", wgPort)
 	proxyPortStr := fmt.Sprintf("%d", proxyPort)
 
@@ -375,12 +580,16 @@ func (m *Manager) initNoTrackChain() error {
 		return fmt.Errorf("add prerouting jump rule: %w", err)
 	}
 
+	m.rawSupported = true
 	return nil
 }
 
 func (m *Manager) cleanupNoTrackChain() error {
 	exists, err := m.ipv4Client.ChainExists(tableRaw, chainNameRaw)
 	if err != nil {
+		if !m.rawSupported {
+			return nil
+		}
 		return fmt.Errorf("check chain exists: %w", err)
 	}
 	if !exists {
@@ -401,6 +610,7 @@ func (m *Manager) cleanupNoTrackChain() error {
 		return fmt.Errorf("clear and delete chain: %w", err)
 	}
 
+	m.rawSupported = false
 	return nil
 }
 

client/firewall/iptables/router_linux.go

@@ -36,6 +36,7 @@ const (
 	chainRTFWDOUT           = "NETBIRD-RT-FWD-OUT"
 	chainRTPRE              = "NETBIRD-RT-PRE"
 	chainRTRDR              = "NETBIRD-RT-RDR"
+	chainNATOutput          = "NETBIRD-NAT-OUTPUT"
 	chainRTMSSCLAMP         = "NETBIRD-RT-MSSCLAMP"
 	routingFinalForwardJump = "ACCEPT"
 	routingFinalNatJump     = "MASQUERADE"
@@ -43,6 +44,7 @@ const (
 	jumpManglePre  = "jump-mangle-pre"
 	jumpNatPre     = "jump-nat-pre"
 	jumpNatPost    = "jump-nat-post"
+	jumpNatOutput  = "jump-nat-output"
 	jumpMSSClamp   = "jump-mss-clamp"
 	markManglePre  = "mark-mangle-pre"
 	markManglePost = "mark-mangle-post"
@@ -52,8 +54,10 @@ const (
 	snatSuffix = "_snat"
 	fwdSuffix  = "_fwd"
 
-	// ipTCPHeaderMinSize represents minimum IP (20) + TCP (20) header size for MSS calculation
-	ipTCPHeaderMinSize = 40
+	// ipv4TCPHeaderSize is the minimum IPv4 (20) + TCP (20) header size for MSS calculation.
+	ipv4TCPHeaderSize = 40
+	// ipv6TCPHeaderSize is the minimum IPv6 (40) + TCP (20) header size for MSS calculation.
+	ipv6TCPHeaderSize = 60
 )
 
 type ruleInfo struct {
@@ -84,6 +88,7 @@ type router struct {
 	wgIface          iFaceMapper
 	legacyManagement bool
 	mtu              uint16
+	v6               bool
 
 	stateManager *statemanager.Manager
 	ipFwdState   *ipfwdstate.IPForwardingState
@@ -95,6 +100,7 @@ func newRouter(iptablesClient *iptables.IPTables, wgIface iFaceMapper, mtu uint1
 		rules:          make(map[string][]string),
 		wgIface:        wgIface,
 		mtu:            mtu,
+		v6:             iptablesClient.Proto() == iptables.ProtocolIPv6,
 		ipFwdState:     ipfwdstate.NewIPForwardingState(),
 	}
 
@@ -184,6 +190,11 @@ func (r *router) AddRouteFiltering(
 	return ruleKey, nil
 }
 
+func (r *router) hasRule(id string) bool {
+	_, ok := r.rules[id]
+	return ok
+}
+
 func (r *router) DeleteRouteRule(rule firewall.Rule) error {
 	ruleKey := rule.ID()
 
@@ -387,6 +398,14 @@ func (r *router) cleanUpDefaultForwardRules() error {
 	}
 
 	log.Debug("flushing routing related tables")
+
+	// Remove jump rules from built-in chains before deleting custom chains,
+	// otherwise the chain deletion fails with "device or resource busy".
+	jumpRule := []string{"-j", chainNATOutput}
+	if err := r.iptablesClient.Delete(tableNat, "OUTPUT", jumpRule...); err != nil {
+		log.Debugf("clean OUTPUT jump rule: %v", err)
+	}
+
 	for _, chainInfo := range []struct {
 		chain string
 		table string
@@ -396,6 +415,7 @@ func (r *router) cleanUpDefaultForwardRules() error {
 		{chainRTPRE, tableMangle},
 		{chainRTNAT, tableNat},
 		{chainRTRDR, tableNat},
+		{chainNATOutput, tableNat},
 		{chainRTMSSCLAMP, tableMangle},
 	} {
 		ok, err := r.iptablesClient.ChainExists(chainInfo.table, chainInfo.chain)
@@ -423,6 +443,12 @@ func (r *router) createContainers() error {
 		{chainRTRDR, tableNat},
 		{chainRTMSSCLAMP, tableMangle},
 	} {
+		// Fallback: clear chains that survived an unclean shutdown.
+		if ok, _ := r.iptablesClient.ChainExists(chainInfo.table, chainInfo.chain); ok {
+			if err := r.iptablesClient.ClearAndDeleteChain(chainInfo.table, chainInfo.chain); err != nil {
+				log.Warnf("clear stale chain %s in %s: %v", chainInfo.chain, chainInfo.table, err)
+			}
+		}
 		if err := r.iptablesClient.NewChain(chainInfo.table, chainInfo.chain); err != nil {
 			return fmt.Errorf("create chain %s in table %s: %w", chainInfo.chain, chainInfo.table, err)
 		}
@@ -529,9 +555,12 @@ func (r *router) addPostroutingRules() error {
 }
 
 // addMSSClampingRules adds MSS clamping rules to prevent fragmentation for forwarded traffic.
-// TODO: Add IPv6 support
 func (r *router) addMSSClampingRules() error {
-	mss := r.mtu - ipTCPHeaderMinSize
+	overhead := uint16(ipv4TCPHeaderSize)
+	if r.v6 {
+		overhead = ipv6TCPHeaderSize
+	}
+	mss := r.mtu - overhead
 
 	// Add jump rule from FORWARD chain in mangle table to our custom chain
 	jumpRule := []string{
@@ -716,8 +745,13 @@ func (r *router) updateState() {
 	currentState.Lock()
 	defer currentState.Unlock()
 
-	currentState.RouteRules = r.rules
-	currentState.RouteIPsetCounter = r.ipsetCounter
+	if r.v6 {
+		currentState.RouteRules6 = r.rules
+		currentState.RouteIPsetCounter6 = r.ipsetCounter
+	} else {
+		currentState.RouteRules = r.rules
+		currentState.RouteIPsetCounter = r.ipsetCounter
+	}
 
 	if err := r.stateManager.UpdateState(currentState); err != nil {
 		log.Errorf("failed to update state: %v", err)
@@ -845,7 +879,7 @@ func (r *router) DeleteDNATRule(rule firewall.Rule) error {
 	}
 
 	if fwdRule, exists := r.rules[ruleKey+fwdSuffix]; exists {
-		if err := r.iptablesClient.Delete(tableFilter, chainRTFWDIN, fwdRule...); err != nil {
+		if err := r.iptablesClient.Delete(tableFilter, chainRTFWDOUT, fwdRule...); err != nil {
 			merr = multierror.Append(merr, fmt.Errorf("delete forward rule: %w", err))
 		}
 		delete(r.rules, ruleKey+fwdSuffix)
@@ -872,7 +906,7 @@ func (r *router) genRouteRuleSpec(params routeFilteringRuleParams, sources []net
 	rule = append(rule, destExp...)
 
 	if params.Proto != firewall.ProtocolALL {
-		rule = append(rule, "-p", strings.ToLower(string(params.Proto)))
+		rule = append(rule, "-p", strings.ToLower(protoForFamily(params.Proto, r.v6)))
 		rule = append(rule, applyPort("--sport", params.SPort)...)
 		rule = append(rule, applyPort("--dport", params.DPort)...)
 	}
@@ -889,11 +923,12 @@ func (r *router) applyNetwork(flag string, network firewall.Network, prefixes []
 	}
 
 	if network.IsSet() {
-		if _, err := r.ipsetCounter.Increment(network.Set.HashedName(), prefixes); err != nil {
+		name := r.ipsetName(network.Set.HashedName())
+		if _, err := r.ipsetCounter.Increment(name, prefixes); err != nil {
 			return nil, fmt.Errorf("create or get ipset: %w", err)
 		}
 
-		return []string{"-m", "set", matchSet, network.Set.HashedName(), direction}, nil
+		return []string{"-m", "set", matchSet, name, direction}, nil
 	}
 	if network.IsPrefix() {
 		return []string{flag, network.Prefix.String()}, nil
@@ -904,19 +939,15 @@ func (r *router) applyNetwork(flag string, network firewall.Network, prefixes []
 }
 
 func (r *router) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
+	name := r.ipsetName(set.HashedName())
 	var merr *multierror.Error
 	for _, prefix := range prefixes {
-		// TODO: Implement IPv6 support
-		if prefix.Addr().Is6() {
-			log.Tracef("skipping IPv6 prefix %s: IPv6 support not yet implemented", prefix)
-			continue
-		}
-		if err := r.addPrefixToIPSet(set.HashedName(), prefix); err != nil {
+		if err := r.addPrefixToIPSet(name, prefix); err != nil {
 			merr = multierror.Append(merr, fmt.Errorf("add prefix to ipset: %w", err))
 		}
 	}
 	if merr == nil {
-		log.Debugf("updated set %s with prefixes %v", set.HashedName(), prefixes)
+		log.Debugf("updated set %s with prefixes %v", name, prefixes)
 	}
 
 	return nberrors.FormatErrorOrNil(merr)
@@ -932,7 +963,7 @@ func (r *router) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol
 
 	dnatRule := []string{
 		"-i", r.wgIface.Name(),
-		"-p", strings.ToLower(string(protocol)),
+		"-p", strings.ToLower(protoForFamily(protocol, r.v6)),
 		"--dport", strconv.Itoa(int(sourcePort)),
 		"-d", localAddr.String(),
 		"-m", "addrtype", "--dst-type", "LOCAL",
@@ -970,6 +1001,81 @@ func (r *router) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Proto
 	return nil
 }
 
+// ensureNATOutputChain lazily creates the OUTPUT NAT chain and jump rule on first use.
+func (r *router) ensureNATOutputChain() error {
+	if _, exists := r.rules[jumpNatOutput]; exists {
+		return nil
+	}
+
+	chainExists, err := r.iptablesClient.ChainExists(tableNat, chainNATOutput)
+	if err != nil {
+		return fmt.Errorf("check chain %s: %w", chainNATOutput, err)
+	}
+	if !chainExists {
+		if err := r.iptablesClient.NewChain(tableNat, chainNATOutput); err != nil {
+			return fmt.Errorf("create chain %s: %w", chainNATOutput, err)
+		}
+	}
+
+	jumpRule := []string{"-j", chainNATOutput}
+	if err := r.iptablesClient.Insert(tableNat, "OUTPUT", 1, jumpRule...); err != nil {
+		if !chainExists {
+			if delErr := r.iptablesClient.ClearAndDeleteChain(tableNat, chainNATOutput); delErr != nil {
+				log.Warnf("failed to rollback chain %s: %v", chainNATOutput, delErr)
+			}
+		}
+		return fmt.Errorf("add OUTPUT jump rule: %w", err)
+	}
+	r.rules[jumpNatOutput] = jumpRule
+
+	r.updateState()
+	return nil
+}
+
+// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
+func (r *router) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	ruleID := fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, sourcePort, targetPort)
+
+	if _, exists := r.rules[ruleID]; exists {
+		return nil
+	}
+
+	if err := r.ensureNATOutputChain(); err != nil {
+		return err
+	}
+
+	dnatRule := []string{
+		"-p", strings.ToLower(string(protocol)),
+		"--dport", strconv.Itoa(int(sourcePort)),
+		"-d", localAddr.String(),
+		"-j", "DNAT",
+		"--to-destination", ":" + strconv.Itoa(int(targetPort)),
+	}
+
+	if err := r.iptablesClient.Append(tableNat, chainNATOutput, dnatRule...); err != nil {
+		return fmt.Errorf("add output DNAT rule: %w", err)
+	}
+	r.rules[ruleID] = dnatRule
+
+	r.updateState()
+	return nil
+}
+
+// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
+func (r *router) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	ruleID := fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, sourcePort, targetPort)
+
+	if dnatRule, exists := r.rules[ruleID]; exists {
+		if err := r.iptablesClient.Delete(tableNat, chainNATOutput, dnatRule...); err != nil {
+			return fmt.Errorf("delete output DNAT rule: %w", err)
+		}
+		delete(r.rules, ruleID)
+	}
+
+	r.updateState()
+	return nil
+}
+
 func applyPort(flag string, port *firewall.Port) []string {
 	if port == nil {
 		return nil
@@ -990,10 +1096,22 @@ func applyPort(flag string, port *firewall.Port) []string {
 	return []string{flag, strconv.Itoa(int(port.Values[0]))}
 }
 
+// ipsetName returns the ipset name, suffixed with "-v6" for the v6 router
+// to avoid collisions since ipsets are global in the kernel.
+func (r *router) ipsetName(name string) string {
+	if r.v6 {
+		return name + "-v6"
+	}
+	return name
+}
+
 func (r *router) createIPSet(name string) error {
 	opts := ipset.CreateOptions{
 		Replace: true,
 	}
+	if r.v6 {
+		opts.Family = ipset.FamilyIPV6
+	}
 
 	if err := ipset.Create(name, ipset.TypeHashNet, opts); err != nil {
 		return fmt.Errorf("create ipset %s: %w", name, err)

client/firewall/iptables/rule.go

@@ -9,6 +9,7 @@ type Rule struct {
 	mangleSpecs []string
 	ip          string
 	chain       string
+	v6          bool
 }
 
 // GetRuleID returns the rule id

client/firewall/iptables/state_linux.go

@@ -4,6 +4,8 @@ import (
 	"fmt"
 	"sync"
 
+	log "github.com/sirupsen/logrus"
+
 	"github.com/netbirdio/netbird/client/iface"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 )
@@ -37,6 +39,12 @@ type ShutdownState struct {
 
 	ACLEntries    aclEntries  `json:"acl_entries,omitempty"`
 	ACLIPsetStore *ipsetStore `json:"acl_ipset_store,omitempty"`
+
+	// IPv6 counterparts
+	RouteRules6        routeRules    `json:"route_rules_v6,omitempty"`
+	RouteIPsetCounter6 *ipsetCounter `json:"route_ipset_counter_v6,omitempty"`
+	ACLEntries6        aclEntries    `json:"acl_entries_v6,omitempty"`
+	ACLIPsetStore6     *ipsetStore   `json:"acl_ipset_store_v6,omitempty"`
 }
 
 func (s *ShutdownState) Name() string {
@@ -67,6 +75,28 @@ func (s *ShutdownState) Cleanup() error {
 		ipt.aclMgr.ipsetStore = s.ACLIPsetStore
 	}
 
+	// Clean up v6 state even if the current run has no IPv6.
+	// The previous run may have left ip6tables rules behind.
+	if !ipt.hasIPv6() {
+		if err := ipt.createIPv6Components(s.InterfaceState, mtu); err != nil {
+			log.Warnf("failed to create v6 components for cleanup: %v", err)
+		}
+	}
+	if ipt.hasIPv6() {
+		if s.RouteRules6 != nil {
+			ipt.router6.rules = s.RouteRules6
+		}
+		if s.RouteIPsetCounter6 != nil {
+			ipt.router6.ipsetCounter.LoadData(s.RouteIPsetCounter6)
+		}
+		if s.ACLEntries6 != nil {
+			ipt.aclMgr6.entries = s.ACLEntries6
+		}
+		if s.ACLIPsetStore6 != nil {
+			ipt.aclMgr6.ipsetStore = s.ACLIPsetStore6
+		}
+	}
+
 	if err := ipt.Close(nil); err != nil {
 		return fmt.Errorf("reset iptables manager: %w", err)
 	}

client/firewall/manager/firewall.go

@@ -169,6 +169,14 @@ type Manager interface {
 	// RemoveInboundDNAT removes inbound DNAT rule
 	RemoveInboundDNAT(localAddr netip.Addr, protocol Protocol, sourcePort, targetPort uint16) error
 
+	// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
+	// localAddr must be IPv4; the underlying iptables/nftables backends are IPv4-only.
+	AddOutputDNAT(localAddr netip.Addr, protocol Protocol, sourcePort, targetPort uint16) error
+
+	// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
+	// localAddr must be IPv4; the underlying iptables/nftables backends are IPv4-only.
+	RemoveOutputDNAT(localAddr netip.Addr, protocol Protocol, sourcePort, targetPort uint16) error
+
 	// SetupEBPFProxyNoTrack creates static notrack rules for eBPF proxy loopback traffic.
 	// This prevents conntrack from interfering with WireGuard proxy communication.
 	SetupEBPFProxyNoTrack(proxyPort, wgPort uint16) error

client/firewall/nftables/acl_linux.go

@@ -33,15 +33,12 @@ const (
 
 const flushError = "flush: %w"
 
-var (
-	anyIP = []byte{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
-)
-
 type AclManager struct {
 	rConn              *nftables.Conn
 	sConn              *nftables.Conn
 	wgIface            iFaceMapper
 	routingFwChainName string
+	af                 addrFamily
 
 	workTable       *nftables.Table
 	chainInputRules *nftables.Chain
@@ -67,6 +64,7 @@ func newAclManager(table *nftables.Table, wgIface iFaceMapper, routingFwChainNam
 		wgIface:            wgIface,
 		workTable:          table,
 		routingFwChainName: routingFwChainName,
+		af:                 familyForAddr(table.Family == nftables.TableFamilyIPv4),
 
 		ipsetStore: newIpsetStore(),
 		rules:      make(map[string]*Rule),
@@ -145,7 +143,7 @@ func (m *AclManager) DeletePeerRule(rule firewall.Rule) error {
 	}
 
 	if _, ok := ips[r.ip.String()]; ok {
-		err := m.sConn.SetDeleteElements(r.nftSet, []nftables.SetElement{{Key: r.ip.To4()}})
+		err := m.sConn.SetDeleteElements(r.nftSet, []nftables.SetElement{{Key: ipToBytes(r.ip, m.af)}})
 		if err != nil {
 			log.Errorf("delete elements for set %q: %v", r.nftSet.Name, err)
 		}
@@ -254,11 +252,11 @@ func (m *AclManager) addIOFiltering(
 		expressions = append(expressions, &expr.Payload{
 			DestRegister: 1,
 			Base:         expr.PayloadBaseNetworkHeader,
-			Offset:       uint32(9),
+			Offset:       m.af.protoOffset,
 			Len:          uint32(1),
 		})
 
-		protoData, err := protoToInt(proto)
+		protoData, err := m.af.protoNum(proto)
 		if err != nil {
 			return nil, fmt.Errorf("convert protocol to number: %v", err)
 		}
@@ -270,19 +268,16 @@ func (m *AclManager) addIOFiltering(
 		})
 	}
 
-	rawIP := ip.To4()
+	rawIP := ipToBytes(ip, m.af)
 	// check if rawIP contains zeroed IPv4 0.0.0.0 value
 	// in that case not add IP match expression into the rule definition
-	if !bytes.HasPrefix(anyIP, rawIP) {
-		// source address position
-		addrOffset := uint32(12)
-
+	if slices.ContainsFunc(rawIP, func(v byte) bool { return v != 0 }) {
 		expressions = append(expressions,
 			&expr.Payload{
 				DestRegister: 1,
 				Base:         expr.PayloadBaseNetworkHeader,
-				Offset:       addrOffset,
-				Len:          4,
+				Offset:       m.af.srcAddrOffset,
+				Len:          m.af.addrLen,
 			},
 		)
 		// add individual IP for match if no ipset defined
@@ -587,7 +582,7 @@ func (m *AclManager) addJumpRule(chain *nftables.Chain, to string, ifaceKey expr
 
 func (m *AclManager) addIpToSet(ipsetName string, ip net.IP) (*nftables.Set, error) {
 	ipset, err := m.rConn.GetSetByName(m.workTable, ipsetName)
-	rawIP := ip.To4()
+	rawIP := ipToBytes(ip, m.af)
 	if err != nil {
 		if ipset, err = m.createSet(m.workTable, ipsetName); err != nil {
 			return nil, fmt.Errorf("get set name: %v", err)
@@ -619,7 +614,7 @@ func (m *AclManager) createSet(table *nftables.Table, name string) (*nftables.Se
 		Name:    name,
 		Table:   table,
 		Dynamic: true,
-		KeyType: nftables.TypeIPAddr,
+		KeyType: m.af.setKeyType,
 	}
 
 	if err := m.rConn.AddSet(ipset, nil); err != nil {
@@ -707,15 +702,12 @@ func ifname(n string) []byte {
 	return b
 }
 
-func protoToInt(protocol firewall.Protocol) (uint8, error) {
-	switch protocol {
-	case firewall.ProtocolTCP:
-		return unix.IPPROTO_TCP, nil
-	case firewall.ProtocolUDP:
-		return unix.IPPROTO_UDP, nil
-	case firewall.ProtocolICMP:
-		return unix.IPPROTO_ICMP, nil
-	}
 
-	return 0, fmt.Errorf("unsupported protocol: %s", protocol)
+// ipToBytes converts net.IP to the correct byte length for the address family.
+func ipToBytes(ip net.IP, af addrFamily) []byte {
+	if af.addrLen == 4 {
+		return ip.To4()
+	}
+	return ip.To16()
 }
+

client/firewall/nftables/addr_family_linux.go

@@ -0,0 +1,81 @@
+package nftables
+
+import (
+	"fmt"
+	"net"
+
+	"github.com/google/nftables"
+	"golang.org/x/sys/unix"
+
+	firewall "github.com/netbirdio/netbird/client/firewall/manager"
+)
+
+var (
+	// afIPv4 defines IPv4 header layout and nftables types.
+	afIPv4 = addrFamily{
+		protoOffset:   9,
+		srcAddrOffset: 12,
+		dstAddrOffset: 16,
+		addrLen:       net.IPv4len,
+		totalBits:     8 * net.IPv4len,
+		setKeyType:    nftables.TypeIPAddr,
+		tableFamily:   nftables.TableFamilyIPv4,
+		icmpProto:     unix.IPPROTO_ICMP,
+	}
+	// afIPv6 defines IPv6 header layout and nftables types.
+	afIPv6 = addrFamily{
+		protoOffset:   6,
+		srcAddrOffset: 8,
+		dstAddrOffset: 24,
+		addrLen:       net.IPv6len,
+		totalBits:     8 * net.IPv6len,
+		setKeyType:    nftables.TypeIP6Addr,
+		tableFamily:   nftables.TableFamilyIPv6,
+		icmpProto:     unix.IPPROTO_ICMPV6,
+	}
+)
+
+// addrFamily holds protocol-specific constants for nftables expression building.
+type addrFamily struct {
+	// protoOffset is the IP header offset for the protocol/next-header field (9 for v4, 6 for v6)
+	protoOffset uint32
+	// srcAddrOffset is the IP header offset for the source address (12 for v4, 8 for v6)
+	srcAddrOffset uint32
+	// dstAddrOffset is the IP header offset for the destination address (16 for v4, 24 for v6)
+	dstAddrOffset uint32
+	// addrLen is the byte length of addresses (4 for v4, 16 for v6)
+	addrLen uint32
+	// totalBits is the address size in bits (32 for v4, 128 for v6)
+	totalBits int
+	// setKeyType is the nftables set data type for addresses
+	setKeyType nftables.SetDatatype
+	// tableFamily is the nftables table family
+	tableFamily nftables.TableFamily
+	// icmpProto is the ICMP protocol number for this family (1 for v4, 58 for v6)
+	icmpProto uint8
+}
+
+// familyForAddr returns the address family for the given IP.
+func familyForAddr(is4 bool) addrFamily {
+	if is4 {
+		return afIPv4
+	}
+	return afIPv6
+}
+
+// protoNum converts a firewall protocol to the IP protocol number,
+// using the correct ICMP variant for the address family.
+func (af addrFamily) protoNum(protocol firewall.Protocol) (uint8, error) {
+	switch protocol {
+	case firewall.ProtocolTCP:
+		return unix.IPPROTO_TCP, nil
+	case firewall.ProtocolUDP:
+		return unix.IPPROTO_UDP, nil
+	case firewall.ProtocolICMP:
+		return af.icmpProto, nil
+	case firewall.ProtocolALL:
+		return 0, nil
+	default:
+		return 0, fmt.Errorf("unsupported protocol: %s", protocol)
+	}
+}

client/firewall/nftables/manager_linux.go

@@ -11,9 +11,11 @@ import (
 	"github.com/google/nftables"
 	"github.com/google/nftables/binaryutil"
 	"github.com/google/nftables/expr"
+	"github.com/hashicorp/go-multierror"
 	log "github.com/sirupsen/logrus"
 	"golang.org/x/sys/unix"
 
+	nberrors "github.com/netbirdio/netbird/client/errors"
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/internal/statemanager"
@@ -49,8 +51,13 @@ type Manager struct {
 	rConn   *nftables.Conn
 	wgIface iFaceMapper
 
-	router                 *router
-	aclManager             *AclManager
+	router     *router
+	aclManager *AclManager
+
+	// IPv6 counterparts, nil when no v6 overlay
+	router6     *router
+	aclManager6 *AclManager
+
 	notrackOutputChain     *nftables.Chain
 	notrackPreroutingChain *nftables.Chain
 }
@@ -62,7 +69,8 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
 		wgIface: wgIface,
 	}
 
-	workTable := &nftables.Table{Name: getTableName(), Family: nftables.TableFamilyIPv4}
+	tableName := getTableName()
+	workTable := &nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv4}
 
 	var err error
 	m.router, err = newRouter(workTable, wgIface, mtu)
@@ -75,11 +83,70 @@ func Create(wgIface iFaceMapper, mtu uint16) (*Manager, error) {
 		return nil, fmt.Errorf("create acl manager: %w", err)
 	}
 
+	if wgIface.Address().HasIPv6() {
+		if err := m.createIPv6Components(tableName, wgIface, mtu); err != nil {
+			return nil, fmt.Errorf("create IPv6 firewall: %w", err)
+		}
+	}
+
 	return m, nil
 }
 
+func (m *Manager) createIPv6Components(tableName string, wgIface iFaceMapper, mtu uint16) error {
+	workTable6 := &nftables.Table{Name: tableName, Family: nftables.TableFamilyIPv6}
+
+	var err error
+	m.router6, err = newRouter(workTable6, wgIface, mtu)
+	if err != nil {
+		return fmt.Errorf("create v6 router: %w", err)
+	}
+
+	// Share the same IP forwarding state with the v4 router, since
+	// EnableIPForwarding controls both v4 and v6 sysctls.
+	m.router6.ipFwdState = m.router.ipFwdState
+
+	m.aclManager6, err = newAclManager(workTable6, wgIface, chainNameRoutingFw)
+	if err != nil {
+		return fmt.Errorf("create v6 acl manager: %w", err)
+	}
+
+	return nil
+}
+
+// hasIPv6 reports whether the manager has IPv6 components initialized.
+func (m *Manager) hasIPv6() bool {
+	return m.router6 != nil
+}
+
+func (m *Manager) initIPv6() error {
+	workTable6, err := m.createWorkTableFamily(nftables.TableFamilyIPv6)
+	if err != nil {
+		return fmt.Errorf("create v6 work table: %w", err)
+	}
+
+	if err := m.router6.init(workTable6); err != nil {
+		return fmt.Errorf("v6 router init: %w", err)
+	}
+
+	if err := m.aclManager6.init(workTable6); err != nil {
+		return fmt.Errorf("v6 acl manager init: %w", err)
+	}
+
+	return nil
+}
+
 // Init nftables firewall manager
 func (m *Manager) Init(stateManager *statemanager.Manager) error {
+	if err := m.initFirewall(); err != nil {
+		return err
+	}
+
+	m.persistState(stateManager)
+
+	return nil
+}
+
+func (m *Manager) initFirewall() error {
 	workTable, err := m.createWorkTable()
 	if err != nil {
 		return fmt.Errorf("create work table: %w", err)
@@ -90,20 +157,32 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
 	}
 
 	if err := m.aclManager.init(workTable); err != nil {
-		// TODO: cleanup router
+		m.rollbackInit()
 		return fmt.Errorf("acl manager init: %w", err)
 	}
 
-	if err := m.initNoTrackChains(workTable); err != nil {
-		return fmt.Errorf("init notrack chains: %w", err)
+	if m.hasIPv6() {
+		if err := m.initIPv6(); err != nil {
+			// Peer has a v6 address: v6 firewall MUST work or we risk fail-open.
+			m.rollbackInit()
+			return fmt.Errorf("init IPv6 firewall (required because peer has IPv6 address): %w", err)
+		}
 	}
 
+	if err := m.initNoTrackChains(workTable); err != nil {
+		log.Warnf("raw priority chains not available, notrack rules will be disabled: %v", err)
+	}
+
+	return nil
+}
+
+// persistState saves the current interface state for potential recreation on restart.
+// Unlike iptables, which requires tracking individual rules, nftables maintains
+// a known state (our netbird table plus a few static rules). This allows for easy
+// cleanup using Close() without needing to store specific rules.
+func (m *Manager) persistState(stateManager *statemanager.Manager) {
 	stateManager.RegisterState(&ShutdownState{})
 
-	// We only need to record minimal interface state for potential recreation.
-	// Unlike iptables, which requires tracking individual rules, nftables maintains
-	// a known state (our netbird table plus a few static rules). This allows for easy
-	// cleanup using Close() without needing to store specific rules.
 	if err := stateManager.UpdateState(&ShutdownState{
 		InterfaceState: &InterfaceState{
 			NameStr:       m.wgIface.Name(),
@@ -115,14 +194,29 @@ func (m *Manager) Init(stateManager *statemanager.Manager) error {
 		log.Errorf("failed to update state: %v", err)
 	}
 
-	// persist early
 	go func() {
 		if err := stateManager.PersistState(context.Background()); err != nil {
 			log.Errorf("failed to persist state: %v", err)
 		}
 	}()
+}
 
-	return nil
+// rollbackInit performs best-effort cleanup of already-initialized state when Init fails partway through.
+func (m *Manager) rollbackInit() {
+	if err := m.router.Reset(); err != nil {
+		log.Warnf("rollback router: %v", err)
+	}
+	if m.hasIPv6() {
+		if err := m.router6.Reset(); err != nil {
+			log.Warnf("rollback v6 router: %v", err)
+		}
+	}
+	if err := m.cleanupNetbirdTables(); err != nil {
+		log.Warnf("cleanup tables: %v", err)
+	}
+	if err := m.rConn.Flush(); err != nil {
+		log.Warnf("flush: %v", err)
+	}
 }
 
 // AddPeerFiltering rule to the firewall
@@ -141,12 +235,14 @@ func (m *Manager) AddPeerFiltering(
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	rawIP := ip.To4()
-	if rawIP == nil {
-		return nil, fmt.Errorf("unsupported IP version: %s", ip.String())
+	if ip.To4() != nil {
+		return m.aclManager.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
 	}
 
-	return m.aclManager.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
+	if !m.hasIPv6() {
+		return nil, fmt.Errorf("IPv6 not initialized, cannot add rule for %s", ip)
+	}
+	return m.aclManager6.AddPeerFiltering(id, ip, proto, sPort, dPort, action, ipsetName)
 }
 
 func (m *Manager) AddRouteFiltering(
@@ -160,8 +256,11 @@ func (m *Manager) AddRouteFiltering(
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	if destination.IsPrefix() && !destination.Prefix.Addr().Is4() {
-		return nil, fmt.Errorf("unsupported IP version: %s", destination.Prefix.Addr().String())
+	if isIPv6RouteRule(sources, destination) {
+		if !m.hasIPv6() {
+			return nil, fmt.Errorf("IPv6 not initialized, cannot add route rule")
+		}
+		return m.router6.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
 	}
 
 	return m.router.AddRouteFiltering(id, sources, destination, proto, sPort, dPort, action)
@@ -172,14 +271,38 @@ func (m *Manager) DeletePeerRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && isIPv6Rule(rule) {
+		return m.aclManager6.DeletePeerRule(rule)
+	}
 	return m.aclManager.DeletePeerRule(rule)
 }
 
-// DeleteRouteRule deletes a routing rule
+func isIPv6Rule(rule firewall.Rule) bool {
+	r, ok := rule.(*Rule)
+	return ok && r.nftRule != nil && r.nftRule.Table != nil && r.nftRule.Table.Family == nftables.TableFamilyIPv6
+}
+
+// isIPv6RouteRule determines whether a route rule belongs to the v6 table.
+// For static routes, the destination prefix determines the family. For dynamic
+// routes (DomainSet), the sources determine the family since management
+// duplicates dynamic rules per family.
+func isIPv6RouteRule(sources []netip.Prefix, destination firewall.Network) bool {
+	if destination.IsPrefix() {
+		return destination.Prefix.Addr().Is6()
+	}
+	return len(sources) > 0 && sources[0].Addr().Is6()
+}
+
+// DeleteRouteRule deletes a routing rule.
+// Route rules are keyed by content hash, so the rule exists in exactly one
+// router. We check v4 first; if the key isn't there, try v6.
 func (m *Manager) DeleteRouteRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && !m.router.hasRule(rule.ID()) {
+		return m.router6.DeleteRouteRule(rule)
+	}
 	return m.router.DeleteRouteRule(rule)
 }
 
@@ -195,17 +318,63 @@ func (m *Manager) AddNatRule(pair firewall.RouterPair) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.AddNatRule(pair)
+	if pair.Destination.IsPrefix() && pair.Destination.Prefix.Addr().Is6() {
+		if !m.hasIPv6() {
+			return fmt.Errorf("IPv6 not initialized, cannot add NAT rule")
+		}
+		return m.router6.AddNatRule(pair)
+	}
+
+	if err := m.router.AddNatRule(pair); err != nil {
+		return err
+	}
+
+	// Dynamic routes (DomainSet) need NAT in both tables since resolved IPs
+	// can be either v4 or v6.
+	if m.hasIPv6() && pair.Destination.IsSet() {
+		v6Pair := pair
+		v6Pair.Source = firewall.Network{Prefix: netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
+		if err := m.router6.AddNatRule(v6Pair); err != nil {
+			return fmt.Errorf("add v6 NAT rule: %w", err)
+		}
+	}
+
+	return nil
 }
 
 func (m *Manager) RemoveNatRule(pair firewall.RouterPair) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.RemoveNatRule(pair)
+	if pair.Destination.IsPrefix() && pair.Destination.Prefix.Addr().Is6() {
+		if !m.hasIPv6() {
+			return nil
+		}
+		return m.router6.RemoveNatRule(pair)
+	}
+
+	if err := m.router.RemoveNatRule(pair); err != nil {
+		return err
+	}
+
+	if m.hasIPv6() && pair.Destination.IsSet() {
+		v6Pair := pair
+		v6Pair.Source = firewall.Network{Prefix: netip.PrefixFrom(netip.IPv6Unspecified(), 0)}
+		if err := m.router6.RemoveNatRule(v6Pair); err != nil {
+			return fmt.Errorf("remove v6 NAT rule: %w", err)
+		}
+	}
+
+	return nil
 }
 
-// AllowNetbird allows netbird interface traffic
+// AllowNetbird allows netbird interface traffic.
+// TODO: In USP mode this only adds ACCEPT to the netbird table's own chains,
+// which doesn't override DROP rules in external tables (e.g. firewalld).
+// Should add passthrough rules to external chains (like the native mode router's
+// addExternalChainsRules does) for both the netbird table family and inet tables.
+// The netbird table itself is fine (routing chains already exist there), but
+// non-netbird tables with INPUT/FORWARD hooks can still DROP our WG traffic.
 func (m *Manager) AllowNetbird() error {
 	if !m.wgIface.IsUserspaceBind() {
 		return nil
@@ -217,6 +386,11 @@ func (m *Manager) AllowNetbird() error {
 	if err := m.aclManager.createDefaultAllowRules(); err != nil {
 		return fmt.Errorf("create default allow rules: %w", err)
 	}
+	if m.hasIPv6() {
+		if err := m.aclManager6.createDefaultAllowRules(); err != nil {
+			return fmt.Errorf("create v6 default allow rules: %w", err)
+		}
+	}
 	if err := m.rConn.Flush(); err != nil {
 		return fmt.Errorf("flush allow input netbird rules: %w", err)
 	}
@@ -226,7 +400,13 @@ func (m *Manager) AllowNetbird() error {
 
 // SetLegacyManagement sets the route manager to use legacy management
 func (m *Manager) SetLegacyManagement(isLegacy bool) error {
-	return firewall.SetLegacyManagement(m.router, isLegacy)
+	if err := firewall.SetLegacyManagement(m.router, isLegacy); err != nil {
+		return err
+	}
+	if m.hasIPv6() {
+		return firewall.SetLegacyManagement(m.router6, isLegacy)
+	}
+	return nil
 }
 
 // Close closes the firewall manager
@@ -234,23 +414,31 @@ func (m *Manager) Close(stateManager *statemanager.Manager) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	var merr *multierror.Error
+
 	if err := m.router.Reset(); err != nil {
-		return fmt.Errorf("reset router: %v", err)
+		merr = multierror.Append(merr, fmt.Errorf("reset router: %v", err))
+	}
+
+	if m.hasIPv6() {
+		if err := m.router6.Reset(); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("reset v6 router: %v", err))
+		}
 	}
 
 	if err := m.cleanupNetbirdTables(); err != nil {
-		return fmt.Errorf("cleanup netbird tables: %v", err)
+		merr = multierror.Append(merr, fmt.Errorf("cleanup netbird tables: %v", err))
 	}
 
 	if err := m.rConn.Flush(); err != nil {
-		return fmt.Errorf(flushError, err)
+		merr = multierror.Append(merr, fmt.Errorf(flushError, err))
 	}
 
 	if err := stateManager.DeleteState(&ShutdownState{}); err != nil {
-		return fmt.Errorf("delete state: %v", err)
+		merr = multierror.Append(merr, fmt.Errorf("delete state: %v", err))
 	}
 
-	return nil
+	return nberrors.FormatErrorOrNil(merr)
 }
 
 func (m *Manager) cleanupNetbirdTables() error {
@@ -299,6 +487,12 @@ func (m *Manager) Flush() error {
 		return err
 	}
 
+	if m.hasIPv6() {
+		if err := m.aclManager6.Flush(); err != nil {
+			return fmt.Errorf("flush v6 acl: %w", err)
+		}
+	}
+
 	if err := m.refreshNoTrackChains(); err != nil {
 		log.Errorf("failed to refresh notrack chains: %v", err)
 	}
@@ -311,6 +505,9 @@ func (m *Manager) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error)
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && rule.TranslatedAddress.Is6() {
+		return m.router6.AddDNATRule(rule)
+	}
 	return m.router.AddDNATRule(rule)
 }
 
@@ -319,6 +516,9 @@ func (m *Manager) DeleteDNATRule(rule firewall.Rule) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && !m.router.hasDNATRule(rule.ID()) {
+		return m.router6.DeleteDNATRule(rule)
+	}
 	return m.router.DeleteDNATRule(rule)
 }
 
@@ -327,7 +527,26 @@ func (m *Manager) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	return m.router.UpdateSet(set, prefixes)
+	var v4Prefixes, v6Prefixes []netip.Prefix
+	for _, p := range prefixes {
+		if p.Addr().Is6() {
+			v6Prefixes = append(v6Prefixes, p)
+		} else {
+			v4Prefixes = append(v4Prefixes, p)
+		}
+	}
+
+	if err := m.router.UpdateSet(set, v4Prefixes); err != nil {
+		return err
+	}
+
+	if m.hasIPv6() && len(v6Prefixes) > 0 {
+		if err := m.router6.UpdateSet(set, v6Prefixes); err != nil {
+			return fmt.Errorf("update v6 set: %w", err)
+		}
+	}
+
+	return nil
 }
 
 // AddInboundDNAT adds an inbound DNAT rule redirecting traffic from NetBird peers to local services.
@@ -335,6 +554,9 @@ func (m *Manager) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protoco
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && localAddr.Is6() {
+		return m.router6.AddInboundDNAT(localAddr, protocol, sourcePort, targetPort)
+	}
 	return m.router.AddInboundDNAT(localAddr, protocol, sourcePort, targetPort)
 }
 
@@ -343,9 +565,28 @@ func (m *Manager) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Prot
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
+	if m.hasIPv6() && localAddr.Is6() {
+		return m.router6.RemoveInboundDNAT(localAddr, protocol, sourcePort, targetPort)
+	}
 	return m.router.RemoveInboundDNAT(localAddr, protocol, sourcePort, targetPort)
 }
 
+// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
+func (m *Manager) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	m.mutex.Lock()
+	defer m.mutex.Unlock()
+
+	return m.router.AddOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
+// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
+func (m *Manager) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	m.mutex.Lock()
+	defer m.mutex.Unlock()
+
+	return m.router.RemoveOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
 const (
 	chainNameRawOutput     = "netbird-raw-out"
 	chainNameRawPrerouting = "netbird-raw-pre"
@@ -517,7 +758,11 @@ func (m *Manager) refreshNoTrackChains() error {
 }
 
 func (m *Manager) createWorkTable() (*nftables.Table, error) {
-	tables, err := m.rConn.ListTablesOfFamily(nftables.TableFamilyIPv4)
+	return m.createWorkTableFamily(nftables.TableFamilyIPv4)
+}
+
+func (m *Manager) createWorkTableFamily(family nftables.TableFamily) (*nftables.Table, error) {
+	tables, err := m.rConn.ListTablesOfFamily(family)
 	if err != nil {
 		return nil, fmt.Errorf("list of tables: %w", err)
 	}
@@ -529,7 +774,7 @@ func (m *Manager) createWorkTable() (*nftables.Table, error) {
 		}
 	}
 
-	table := m.rConn.AddTable(&nftables.Table{Name: getTableName(), Family: nftables.TableFamilyIPv4})
+	table := m.rConn.AddTable(&nftables.Table{Name: tableName, Family: family})
 	err = m.rConn.Flush()
 	return table, err
 }

client/firewall/nftables/manager_linux_test.go

@@ -385,10 +385,134 @@ func TestNftablesManagerCompatibilityWithIptables(t *testing.T) {
 	err = manager.AddNatRule(pair)
 	require.NoError(t, err, "failed to add NAT rule")
 
+	dnatRule, err := manager.AddDNATRule(fw.ForwardRule{
+		Protocol:          fw.ProtocolTCP,
+		DestinationPort:   fw.Port{Values: []uint16{8080}},
+		TranslatedAddress: netip.MustParseAddr("100.96.0.2"),
+		TranslatedPort:    fw.Port{Values: []uint16{80}},
+	})
+	require.NoError(t, err, "failed to add DNAT rule")
+
+	t.Cleanup(func() {
+		require.NoError(t, manager.DeleteDNATRule(dnatRule), "failed to delete DNAT rule")
+	})
+
 	stdout, stderr = runIptablesSave(t)
 	verifyIptablesOutput(t, stdout, stderr)
 }
 
+func TestNftablesManagerIPv6CompatibilityWithIp6tables(t *testing.T) {
+	if check() != NFTABLES {
+		t.Skip("nftables not supported on this system")
+	}
+
+	for _, bin := range []string{"ip6tables", "ip6tables-save", "iptables-save"} {
+		if _, err := exec.LookPath(bin); err != nil {
+			t.Skipf("%s not available on this system: %v", bin, err)
+		}
+	}
+
+	// Seed ip6 tables in the nft backend. Docker may not create them.
+	seedIp6tables(t)
+
+	ifaceMockV6 := &iFaceMock{
+		NameFunc: func() string { return "wt-test" },
+		AddressFunc: func() wgaddr.Address {
+			return wgaddr.Address{
+				IP:      netip.MustParseAddr("100.96.0.1"),
+				Network: netip.MustParsePrefix("100.96.0.0/16"),
+				IPv6:    netip.MustParseAddr("fd00::1"),
+				IPv6Net: netip.MustParsePrefix("fd00::/64"),
+			}
+		},
+	}
+
+	manager, err := Create(ifaceMockV6, iface.DefaultMTU)
+	require.NoError(t, err, "create manager")
+	require.NoError(t, manager.Init(nil))
+
+	t.Cleanup(func() {
+		require.NoError(t, manager.Close(nil), "close manager")
+
+		stdout, stderr := runIp6tablesSave(t)
+		verifyIp6tablesOutput(t, stdout, stderr)
+	})
+
+	ip := netip.MustParseAddr("fd00::2")
+	_, err = manager.AddPeerFiltering(nil, ip.AsSlice(), fw.ProtocolTCP, nil, &fw.Port{Values: []uint16{80}}, fw.ActionAccept, "")
+	require.NoError(t, err, "add v6 peer filtering rule")
+
+	_, err = manager.AddRouteFiltering(
+		nil,
+		[]netip.Prefix{netip.MustParsePrefix("fd00:1::/64")},
+		fw.Network{Prefix: netip.MustParsePrefix("2001:db8::/48")},
+		fw.ProtocolTCP,
+		nil,
+		&fw.Port{Values: []uint16{443}},
+		fw.ActionAccept,
+	)
+	require.NoError(t, err, "add v6 route filtering rule")
+
+	err = manager.AddNatRule(fw.RouterPair{
+		Source:      fw.Network{Prefix: netip.MustParsePrefix("fd00::/64")},
+		Destination: fw.Network{Prefix: netip.MustParsePrefix("2001:db8::/48")},
+		Masquerade:  true,
+	})
+	require.NoError(t, err, "add v6 NAT rule")
+
+	dnatRule, err := manager.AddDNATRule(fw.ForwardRule{
+		Protocol:          fw.ProtocolTCP,
+		DestinationPort:   fw.Port{Values: []uint16{8080}},
+		TranslatedAddress: netip.MustParseAddr("fd00::2"),
+		TranslatedPort:    fw.Port{Values: []uint16{80}},
+	})
+	require.NoError(t, err, "add v6 DNAT rule")
+
+	t.Cleanup(func() {
+		require.NoError(t, manager.DeleteDNATRule(dnatRule), "delete v6 DNAT rule")
+	})
+
+	stdout, stderr := runIptablesSave(t)
+	verifyIptablesOutput(t, stdout, stderr)
+
+	stdout, stderr = runIp6tablesSave(t)
+	verifyIp6tablesOutput(t, stdout, stderr)
+}
+
+func seedIp6tables(t *testing.T) {
+	t.Helper()
+	for _, tc := range []struct{ table, chain string }{
+		{"filter", "FORWARD"},
+		{"nat", "POSTROUTING"},
+		{"mangle", "FORWARD"},
+	} {
+		add := exec.Command("ip6tables", "-t", tc.table, "-A", tc.chain, "-j", "ACCEPT")
+		require.NoError(t, add.Run(), "seed ip6tables -t %s", tc.table)
+		del := exec.Command("ip6tables", "-t", tc.table, "-D", tc.chain, "-j", "ACCEPT")
+		require.NoError(t, del.Run(), "unseed ip6tables -t %s", tc.table)
+	}
+}
+
+func runIp6tablesSave(t *testing.T) (string, string) {
+	t.Helper()
+	var stdout, stderr bytes.Buffer
+	cmd := exec.Command("ip6tables-save")
+	cmd.Stdout = &stdout
+	cmd.Stderr = &stderr
+	require.NoError(t, cmd.Run(), "ip6tables-save failed")
+	return stdout.String(), stderr.String()
+}
+
+func verifyIp6tablesOutput(t *testing.T, stdout, stderr string) {
+	t.Helper()
+	require.NotContains(t, stdout, "Table `nat' is incompatible",
+		"ip6tables-save: nat table incompatible. Full output: %s", stdout)
+	require.NotContains(t, stdout, "Table `mangle' is incompatible",
+		"ip6tables-save: mangle table incompatible. Full output: %s", stdout)
+	require.NotContains(t, stdout, "Table `filter' is incompatible",
+		"ip6tables-save: filter table incompatible. Full output: %s", stdout)
+}
+
 func TestNftablesManagerCompatibilityWithIptablesFor6kPrefixes(t *testing.T) {
 	if check() != NFTABLES {
 		t.Skip("nftables not supported on this system")

client/firewall/nftables/router_linux.go

@@ -36,6 +36,7 @@ const (
 	chainNameRoutingFw     = "netbird-rt-fwd"
 	chainNameRoutingNat    = "netbird-rt-postrouting"
 	chainNameRoutingRdr    = "netbird-rt-redirect"
+	chainNameNATOutput     = "netbird-nat-output"
 	chainNameForward       = "FORWARD"
 	chainNameMangleForward = "netbird-mangle-forward"
 
@@ -46,8 +47,10 @@ const (
 	dnatSuffix = "_dnat"
 	snatSuffix = "_snat"
 
-	// ipTCPHeaderMinSize represents minimum IP (20) + TCP (20) header size for MSS calculation
-	ipTCPHeaderMinSize = 40
+	// ipv4TCPHeaderSize is the minimum IPv4 (20) + TCP (20) header size for MSS calculation.
+	ipv4TCPHeaderSize = 40
+	// ipv6TCPHeaderSize is the minimum IPv6 (40) + TCP (20) header size for MSS calculation.
+	ipv6TCPHeaderSize = 60
 
 	// maxPrefixesSet 1638 prefixes start to fail, taking some margin
 	maxPrefixesSet       = 1500
@@ -72,6 +75,7 @@ type router struct {
 	rules        map[string]*nftables.Rule
 	ipsetCounter *refcounter.Counter[string, setInput, *nftables.Set]
 
+	af               addrFamily
 	wgIface          iFaceMapper
 	ipFwdState       *ipfwdstate.IPForwardingState
 	legacyManagement bool
@@ -84,6 +88,7 @@ func newRouter(workTable *nftables.Table, wgIface iFaceMapper, mtu uint16) (*rou
 		workTable:  workTable,
 		chains:     make(map[string]*nftables.Chain),
 		rules:      make(map[string]*nftables.Rule),
+		af:         familyForAddr(workTable.Family == nftables.TableFamilyIPv4),
 		wgIface:    wgIface,
 		ipFwdState: ipfwdstate.NewIPForwardingState(),
 		mtu:        mtu,
@@ -142,7 +147,7 @@ func (r *router) Reset() error {
 func (r *router) removeNatPreroutingRules() error {
 	table := &nftables.Table{
 		Name:   tableNat,
-		Family: nftables.TableFamilyIPv4,
+		Family: r.af.tableFamily,
 	}
 	chain := &nftables.Chain{
 		Name:     chainNameNatPrerouting,
@@ -175,7 +180,7 @@ func (r *router) removeNatPreroutingRules() error {
 }
 
 func (r *router) loadFilterTable() (*nftables.Table, error) {
-	tables, err := r.conn.ListTablesOfFamily(nftables.TableFamilyIPv4)
+	tables, err := r.conn.ListTablesOfFamily(r.af.tableFamily)
 	if err != nil {
 		return nil, fmt.Errorf("list tables: %w", err)
 	}
@@ -407,7 +412,7 @@ func (r *router) AddRouteFiltering(
 
 	// Handle protocol
 	if proto != firewall.ProtocolALL {
-		protoNum, err := protoToInt(proto)
+		protoNum, err := r.af.protoNum(proto)
 		if err != nil {
 			return nil, fmt.Errorf("convert protocol to number: %w", err)
 		}
@@ -467,7 +472,24 @@ func (r *router) getIpSet(set firewall.Set, prefixes []netip.Prefix, isSource bo
 		return nil, fmt.Errorf("create or get ipset: %w", err)
 	}
 
-	return getIpSetExprs(ref, isSource)
+	return r.getIpSetExprs(ref, isSource)
+}
+
+func (r *router) iptablesProto() iptables.Protocol {
+	if r.af.tableFamily == nftables.TableFamilyIPv6 {
+		return iptables.ProtocolIPv6
+	}
+	return iptables.ProtocolIPv4
+}
+
+func (r *router) hasRule(id string) bool {
+	_, ok := r.rules[id]
+	return ok
+}
+
+func (r *router) hasDNATRule(id string) bool {
+	_, ok := r.rules[id+dnatSuffix]
+	return ok
 }
 
 func (r *router) DeleteRouteRule(rule firewall.Rule) error {
@@ -516,10 +538,10 @@ func (r *router) createIpSet(setName string, input setInput) (*nftables.Set, err
 		Table:   r.workTable,
 		// required for prefixes
 		Interval: true,
-		KeyType:  nftables.TypeIPAddr,
+		KeyType:  r.af.setKeyType,
 	}
 
-	elements := convertPrefixesToSet(prefixes)
+	elements := r.convertPrefixesToSet(prefixes)
 	nElements := len(elements)
 
 	maxElements := maxPrefixesSet * 2
@@ -552,23 +574,17 @@ func (r *router) createIpSet(setName string, input setInput) (*nftables.Set, err
 	return nfset, nil
 }
 
-func convertPrefixesToSet(prefixes []netip.Prefix) []nftables.SetElement {
+func (r *router) convertPrefixesToSet(prefixes []netip.Prefix) []nftables.SetElement {
 	var elements []nftables.SetElement
 	for _, prefix := range prefixes {
-		// TODO: Implement IPv6 support
-		if prefix.Addr().Is6() {
-			log.Tracef("skipping IPv6 prefix %s: IPv6 support not yet implemented", prefix)
-			continue
-		}
-
 		// nftables needs half-open intervals [firstIP, lastIP) for prefixes
 		// e.g. 10.0.0.0/24 becomes [10.0.0.0, 10.0.1.0), 10.1.1.1/32 becomes [10.1.1.1, 10.1.1.2) etc
 		firstIP := prefix.Addr()
 		lastIP := calculateLastIP(prefix).Next()
 
 		elements = append(elements,
-			// the nft tool also adds a line like this, see https://github.com/google/nftables/issues/247
-			// nftables.SetElement{Key: []byte{0, 0, 0, 0}, IntervalEnd: true},
+			// the nft tool also adds a zero-address IntervalEnd element, see https://github.com/google/nftables/issues/247
+			// nftables.SetElement{Key: make([]byte, r.af.addrLen), IntervalEnd: true},
 			nftables.SetElement{Key: firstIP.AsSlice()},
 			nftables.SetElement{Key: lastIP.AsSlice(), IntervalEnd: true},
 		)
@@ -578,10 +594,20 @@ func convertPrefixesToSet(prefixes []netip.Prefix) []nftables.SetElement {
 
 // calculateLastIP determines the last IP in a given prefix.
 func calculateLastIP(prefix netip.Prefix) netip.Addr {
-	hostMask := ^uint32(0) >> prefix.Masked().Bits()
-	lastIP := uint32FromNetipAddr(prefix.Addr()) | hostMask
+	masked := prefix.Masked()
+	if masked.Addr().Is4() {
+		hostMask := ^uint32(0) >> masked.Bits()
+		lastIP := uint32FromNetipAddr(masked.Addr()) | hostMask
+		return netip.AddrFrom4(uint32ToBytes(lastIP))
+	}
 
-	return netip.AddrFrom4(uint32ToBytes(lastIP))
+	// IPv6: set host bits to all 1s
+	b := masked.Addr().As16()
+	bits := masked.Bits()
+	for i := bits; i < 128; i++ {
+		b[i/8] |= 1 << (7 - i%8)
+	}
+	return netip.AddrFrom16(b)
 }
 
 // Utility function to convert netip.Addr to uint32.
@@ -833,9 +859,12 @@ func (r *router) addPostroutingRules() {
 }
 
 // addMSSClampingRules adds MSS clamping rules to prevent fragmentation for forwarded traffic.
-// TODO: Add IPv6 support
 func (r *router) addMSSClampingRules() error {
-	mss := r.mtu - ipTCPHeaderMinSize
+	overhead := uint16(ipv4TCPHeaderSize)
+	if r.af.tableFamily == nftables.TableFamilyIPv6 {
+		overhead = ipv6TCPHeaderSize
+	}
+	mss := r.mtu - overhead
 
 	exprsOut := []expr.Any{
 		&expr.Meta{
@@ -1042,17 +1071,22 @@ func (r *router) acceptFilterTableRules() error {
 		log.Debugf("Used %s to add accept forward and input rules", fw)
 	}()
 
-	// Try iptables first and fallback to nftables if iptables is not available
-	ipt, err := iptables.New()
+	// Try iptables first and fallback to nftables if iptables is not available.
+	// Use the correct protocol (iptables vs ip6tables) for the address family.
+	ipt, err := iptables.NewWithProtocol(r.iptablesProto())
 	if err != nil {
-		// iptables is not available but the filter table exists
 		log.Warnf("Will use nftables to manipulate the filter table because iptables is not available: %v", err)
 
 		fw = "nftables"
 		return r.acceptFilterRulesNftables(r.filterTable)
 	}
 
-	return r.acceptFilterRulesIptables(ipt)
+	if err := r.acceptFilterRulesIptables(ipt); err != nil {
+		log.Warnf("iptables failed (table may be incompatible), falling back to nftables: %v", err)
+		fw = "nftables"
+		return r.acceptFilterRulesNftables(r.filterTable)
+	}
+	return nil
 }
 
 func (r *router) acceptFilterRulesIptables(ipt *iptables.IPTables) error {
@@ -1221,13 +1255,17 @@ func (r *router) removeFilterTableRules() error {
 		return nil
 	}
 
-	ipt, err := iptables.New()
+	ipt, err := iptables.NewWithProtocol(r.iptablesProto())
 	if err != nil {
 		log.Debugf("iptables not available, using nftables to remove filter rules: %v", err)
 		return r.removeAcceptRulesFromTable(r.filterTable)
 	}
 
-	return r.removeAcceptFilterRulesIptables(ipt)
+	if err := r.removeAcceptFilterRulesIptables(ipt); err != nil {
+		log.Debugf("iptables removal failed (table may be incompatible), falling back to nftables: %v", err)
+		return r.removeAcceptRulesFromTable(r.filterTable)
+	}
+	return nil
 }
 
 func (r *router) removeAcceptRulesFromTable(table *nftables.Table) error {
@@ -1294,7 +1332,7 @@ func (r *router) removeExternalChainsRules() error {
 func (r *router) findExternalChains() []*nftables.Chain {
 	var chains []*nftables.Chain
 
-	families := []nftables.TableFamily{nftables.TableFamilyIPv4, nftables.TableFamilyINet}
+	families := []nftables.TableFamily{r.af.tableFamily, nftables.TableFamilyINet}
 
 	for _, family := range families {
 		allChains, err := r.conn.ListChainsOfTableFamily(family)
@@ -1318,8 +1356,8 @@ func (r *router) isExternalChain(chain *nftables.Chain) bool {
 		return false
 	}
 
-	// Skip all iptables-managed tables in the ip family
-	if chain.Table.Family == nftables.TableFamilyIPv4 && isIptablesTable(chain.Table.Name) {
+	// Skip iptables/ip6tables-managed tables (adding nft-native rules breaks iptables-save compat)
+	if (chain.Table.Family == nftables.TableFamilyIPv4 || chain.Table.Family == nftables.TableFamilyIPv6) && isIptablesTable(chain.Table.Name) {
 		return false
 	}
 
@@ -1460,7 +1498,7 @@ func (r *router) AddDNATRule(rule firewall.ForwardRule) (firewall.Rule, error) {
 		return rule, nil
 	}
 
-	protoNum, err := protoToInt(rule.Protocol)
+	protoNum, err := r.af.protoNum(rule.Protocol)
 	if err != nil {
 		return nil, fmt.Errorf("convert protocol to number: %w", err)
 	}
@@ -1523,7 +1561,7 @@ func (r *router) addDnatRedirect(rule firewall.ForwardRule, protoNum uint8, rule
 	dnatExprs = append(dnatExprs,
 		&expr.NAT{
 			Type:        expr.NATTypeDestNAT,
-			Family:      uint32(nftables.TableFamilyIPv4),
+			Family:      uint32(r.af.tableFamily),
 			RegAddrMin:  1,
 			RegProtoMin: regProtoMin,
 			RegProtoMax: regProtoMax,
@@ -1619,7 +1657,7 @@ func (r *router) addXTablesRedirect(dnatExprs []expr.Any, ruleKey string, rule f
 	dnatRule := &nftables.Rule{
 		Table: &nftables.Table{
 			Name:   tableNat,
-			Family: nftables.TableFamilyIPv4,
+			Family: r.af.tableFamily,
 		},
 		Chain: &nftables.Chain{
 			Name:     chainNameNatPrerouting,
@@ -1654,8 +1692,8 @@ func (r *router) addDnatMasq(rule firewall.ForwardRule, protoNum uint8, ruleKey
 		&expr.Payload{
 			DestRegister: 1,
 			Base:         expr.PayloadBaseNetworkHeader,
-			Offset:       16,
-			Len:          4,
+			Offset:       r.af.dstAddrOffset,
+			Len:          r.af.addrLen,
 		},
 		&expr.Cmp{
 			Op:       expr.CmpOpEq,
@@ -1733,7 +1771,7 @@ func (r *router) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
 		return fmt.Errorf("get set %s: %w", set.HashedName(), err)
 	}
 
-	elements := convertPrefixesToSet(prefixes)
+	elements := r.convertPrefixesToSet(prefixes)
 	if err := r.conn.SetAddElements(nfset, elements); err != nil {
 		return fmt.Errorf("add elements to set %s: %w", set.HashedName(), err)
 	}
@@ -1755,7 +1793,7 @@ func (r *router) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol
 		return nil
 	}
 
-	protoNum, err := protoToInt(protocol)
+	protoNum, err := r.af.protoNum(protocol)
 	if err != nil {
 		return fmt.Errorf("convert protocol to number: %w", err)
 	}
@@ -1786,7 +1824,11 @@ func (r *router) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol
 		},
 	}
 
-	exprs = append(exprs, applyPrefix(netip.PrefixFrom(localAddr, 32), false)...)
+	bits := 32
+	if localAddr.Is6() {
+		bits = 128
+	}
+	exprs = append(exprs, r.applyPrefix(netip.PrefixFrom(localAddr, bits), false)...)
 
 	exprs = append(exprs,
 		&expr.Immediate{
@@ -1799,7 +1841,7 @@ func (r *router) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol
 		},
 		&expr.NAT{
 			Type:        expr.NATTypeDestNAT,
-			Family:      uint32(nftables.TableFamilyIPv4),
+			Family:      uint32(r.af.tableFamily),
 			RegAddrMin:  1,
 			RegProtoMin: 2,
 			RegProtoMax: 0,
@@ -1853,6 +1895,134 @@ func (r *router) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Proto
 	return nil
 }
 
+// ensureNATOutputChain lazily creates the OUTPUT NAT chain on first use.
+func (r *router) ensureNATOutputChain() error {
+	if _, exists := r.chains[chainNameNATOutput]; exists {
+		return nil
+	}
+
+	r.chains[chainNameNATOutput] = r.conn.AddChain(&nftables.Chain{
+		Name:     chainNameNATOutput,
+		Table:    r.workTable,
+		Hooknum:  nftables.ChainHookOutput,
+		Priority: nftables.ChainPriorityNATDest,
+		Type:     nftables.ChainTypeNAT,
+	})
+
+	if err := r.conn.Flush(); err != nil {
+		delete(r.chains, chainNameNATOutput)
+		return fmt.Errorf("create NAT output chain: %w", err)
+	}
+	return nil
+}
+
+// AddOutputDNAT adds an OUTPUT chain DNAT rule for locally-generated traffic.
+func (r *router) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	ruleID := fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, sourcePort, targetPort)
+
+	if _, exists := r.rules[ruleID]; exists {
+		return nil
+	}
+
+	if err := r.ensureNATOutputChain(); err != nil {
+		return err
+	}
+
+	protoNum, err := r.af.protoNum(protocol)
+	if err != nil {
+		return fmt.Errorf("convert protocol to number: %w", err)
+	}
+
+	exprs := []expr.Any{
+		&expr.Meta{Key: expr.MetaKeyL4PROTO, Register: 1},
+		&expr.Cmp{
+			Op:       expr.CmpOpEq,
+			Register: 1,
+			Data:     []byte{protoNum},
+		},
+		&expr.Payload{
+			DestRegister: 2,
+			Base:         expr.PayloadBaseTransportHeader,
+			Offset:       2,
+			Len:          2,
+		},
+		&expr.Cmp{
+			Op:       expr.CmpOpEq,
+			Register: 2,
+			Data:     binaryutil.BigEndian.PutUint16(sourcePort),
+		},
+	}
+
+	bits := 32
+	if localAddr.Is6() {
+		bits = 128
+	}
+	exprs = append(exprs, r.applyPrefix(netip.PrefixFrom(localAddr, bits), false)...)
+
+	exprs = append(exprs,
+		&expr.Immediate{
+			Register: 1,
+			Data:     localAddr.AsSlice(),
+		},
+		&expr.Immediate{
+			Register: 2,
+			Data:     binaryutil.BigEndian.PutUint16(targetPort),
+		},
+		&expr.NAT{
+			Type:        expr.NATTypeDestNAT,
+			Family:      uint32(r.af.tableFamily),
+			RegAddrMin:  1,
+			RegProtoMin: 2,
+		},
+	)
+
+	dnatRule := &nftables.Rule{
+		Table:    r.workTable,
+		Chain:    r.chains[chainNameNATOutput],
+		Exprs:    exprs,
+		UserData: []byte(ruleID),
+	}
+	r.conn.AddRule(dnatRule)
+
+	if err := r.conn.Flush(); err != nil {
+		return fmt.Errorf("add output DNAT rule: %w", err)
+	}
+
+	r.rules[ruleID] = dnatRule
+
+	return nil
+}
+
+// RemoveOutputDNAT removes an OUTPUT chain DNAT rule.
+func (r *router) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	if err := r.refreshRulesMap(); err != nil {
+		return fmt.Errorf(refreshRulesMapError, err)
+	}
+
+	ruleID := fmt.Sprintf("output-dnat-%s-%s-%d-%d", localAddr.String(), protocol, sourcePort, targetPort)
+
+	rule, exists := r.rules[ruleID]
+	if !exists {
+		return nil
+	}
+
+	if rule.Handle == 0 {
+		log.Warnf("output DNAT rule %s has no handle, removing stale entry", ruleID)
+		delete(r.rules, ruleID)
+		return nil
+	}
+
+	if err := r.conn.DelRule(rule); err != nil {
+		return fmt.Errorf("delete output DNAT rule %s: %w", ruleID, err)
+	}
+	if err := r.conn.Flush(); err != nil {
+		return fmt.Errorf("flush delete output DNAT rule: %w", err)
+	}
+	delete(r.rules, ruleID)
+
+	return nil
+}
+
 // applyNetwork generates nftables expressions for networks (CIDR) or sets
 func (r *router) applyNetwork(
 	network firewall.Network,
@@ -1868,45 +2038,44 @@ func (r *router) applyNetwork(
 	}
 
 	if network.IsPrefix() {
-		return applyPrefix(network.Prefix, isSource), nil
+		return r.applyPrefix(network.Prefix, isSource), nil
 	}
 
 	return nil, nil
 }
 
 // applyPrefix generates nftables expressions for a CIDR prefix
-func applyPrefix(prefix netip.Prefix, isSource bool) []expr.Any {
-	// dst offset
-	offset := uint32(16)
+func (r *router) applyPrefix(prefix netip.Prefix, isSource bool) []expr.Any {
+	// dst offset by default
+	offset := r.af.dstAddrOffset
 	if isSource {
 		// src offset
-		offset = 12
+		offset = r.af.srcAddrOffset
 	}
 
 	ones := prefix.Bits()
-	// 0.0.0.0/0 doesn't need extra expressions
+	// unspecified address (/0) doesn't need extra expressions
 	if ones == 0 {
 		return nil
 	}
 
-	mask := net.CIDRMask(ones, 32)
+	mask := net.CIDRMask(ones, r.af.totalBits)
+	xor := make([]byte, r.af.addrLen)
 
 	return []expr.Any{
 		&expr.Payload{
 			DestRegister: 1,
 			Base:         expr.PayloadBaseNetworkHeader,
 			Offset:       offset,
-			Len:          4,
+			Len:          r.af.addrLen,
 		},
-		// netmask
 		&expr.Bitwise{
 			DestRegister:   1,
 			SourceRegister: 1,
-			Len:            4,
+			Len:            r.af.addrLen,
 			Mask:           mask,
-			Xor:            []byte{0, 0, 0, 0},
+			Xor:            xor,
 		},
-		// net address
 		&expr.Cmp{
 			Op:       expr.CmpOpEq,
 			Register: 1,
@@ -1989,13 +2158,12 @@ func getCtNewExprs() []expr.Any {
 	}
 }
 
-func getIpSetExprs(ref refcounter.Ref[*nftables.Set], isSource bool) ([]expr.Any, error) {
-
-	// dst offset
-	offset := uint32(16)
+func (r *router) getIpSetExprs(ref refcounter.Ref[*nftables.Set], isSource bool) ([]expr.Any, error) {
+	// dst offset by default
+	offset := r.af.dstAddrOffset
 	if isSource {
 		// src offset
-		offset = 12
+		offset = r.af.srcAddrOffset
 	}
 
 	return []expr.Any{
@@ -2003,7 +2171,7 @@ func getIpSetExprs(ref refcounter.Ref[*nftables.Set], isSource bool) ([]expr.Any
 			DestRegister: 1,
 			Base:         expr.PayloadBaseNetworkHeader,
 			Offset:       offset,
-			Len:          4,
+			Len:          r.af.addrLen,
 		},
 		&expr.Lookup{
 			SourceRegister: 1,

client/firewall/nftables/router_linux_test.go

@@ -90,8 +90,9 @@ func TestNftablesManager_AddNatRule(t *testing.T) {
 				}
 
 				// Build CIDR matching expressions
-				sourceExp := applyPrefix(testCase.InputPair.Source.Prefix, true)
-				destExp := applyPrefix(testCase.InputPair.Destination.Prefix, false)
+				testRouter := &router{af: afIPv4}
+				sourceExp := testRouter.applyPrefix(testCase.InputPair.Source.Prefix, true)
+				destExp := testRouter.applyPrefix(testCase.InputPair.Destination.Prefix, false)
 
 				// Combine all expressions in the correct order
 				// nolint:gocritic
@@ -508,6 +509,136 @@ func TestNftablesCreateIpSet(t *testing.T) {
 	}
 }
 
+func TestNftablesCreateIpSet_IPv6(t *testing.T) {
+	if check() != NFTABLES {
+		t.Skip("nftables not supported on this system")
+	}
+
+	workTable, err := createWorkTableIPv6()
+	require.NoError(t, err, "Failed to create v6 work table")
+	defer deleteWorkTableIPv6()
+
+	r, err := newRouter(workTable, ifaceMock, iface.DefaultMTU)
+	require.NoError(t, err, "Failed to create router")
+	require.NoError(t, r.init(workTable))
+	defer func() {
+		require.NoError(t, r.Reset(), "Failed to reset router")
+	}()
+
+	tests := []struct {
+		name     string
+		sources  []netip.Prefix
+		expected []netip.Prefix
+	}{
+		{
+			name:    "Single IPv6",
+			sources: []netip.Prefix{netip.MustParsePrefix("2001:db8::1/128")},
+		},
+		{
+			name: "Multiple IPv6 Subnets",
+			sources: []netip.Prefix{
+				netip.MustParsePrefix("fd00::/64"),
+				netip.MustParsePrefix("2001:db8::/48"),
+				netip.MustParsePrefix("fe80::/10"),
+			},
+		},
+		{
+			name: "Overlapping IPv6",
+			sources: []netip.Prefix{
+				netip.MustParsePrefix("fd00::/48"),
+				netip.MustParsePrefix("fd00::/64"),
+				netip.MustParsePrefix("fd00::1/128"),
+			},
+			expected: []netip.Prefix{
+				netip.MustParsePrefix("fd00::/48"),
+			},
+		},
+		{
+			name: "Mixed prefix lengths",
+			sources: []netip.Prefix{
+				netip.MustParsePrefix("2001:db8:1::/48"),
+				netip.MustParsePrefix("2001:db8:2::1/128"),
+				netip.MustParsePrefix("fd00:abcd::/32"),
+			},
+		},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			setName := firewall.NewPrefixSet(tt.sources).HashedName()
+			set, err := r.createIpSet(setName, setInput{prefixes: tt.sources})
+			require.NoError(t, err, "Failed to create IPv6 set")
+			require.NotNil(t, set)
+
+			assert.Equal(t, setName, set.Name)
+			assert.True(t, set.Interval)
+			assert.Equal(t, nftables.TypeIP6Addr, set.KeyType)
+
+			fetchedSet, err := r.conn.GetSetByName(r.workTable, setName)
+			require.NoError(t, err, "Failed to fetch created set")
+
+			elements, err := r.conn.GetSetElements(fetchedSet)
+			require.NoError(t, err, "Failed to get set elements")
+
+			uniquePrefixes := make(map[string]bool)
+			for _, elem := range elements {
+				if !elem.IntervalEnd && len(elem.Key) == 16 {
+					ip := netip.AddrFrom16([16]byte(elem.Key))
+					uniquePrefixes[ip.String()] = true
+				}
+			}
+
+			expectedCount := len(tt.expected)
+			if expectedCount == 0 {
+				expectedCount = len(tt.sources)
+			}
+			assert.Equal(t, expectedCount, len(uniquePrefixes), "unique prefix count mismatch")
+
+			r.conn.DelSet(set)
+			require.NoError(t, r.conn.Flush())
+		})
+	}
+}
+
+func createWorkTableIPv6() (*nftables.Table, error) {
+	sConn, err := nftables.New(nftables.AsLasting())
+	if err != nil {
+		return nil, err
+	}
+
+	tables, err := sConn.ListTablesOfFamily(nftables.TableFamilyIPv6)
+	if err != nil {
+		return nil, err
+	}
+	for _, t := range tables {
+		if t.Name == tableNameNetbird {
+			sConn.DelTable(t)
+		}
+	}
+
+	table := sConn.AddTable(&nftables.Table{Name: tableNameNetbird, Family: nftables.TableFamilyIPv6})
+	err = sConn.Flush()
+	return table, err
+}
+
+func deleteWorkTableIPv6() {
+	sConn, err := nftables.New(nftables.AsLasting())
+	if err != nil {
+		return
+	}
+
+	tables, err := sConn.ListTablesOfFamily(nftables.TableFamilyIPv6)
+	if err != nil {
+		return
+	}
+	for _, t := range tables {
+		if t.Name == tableNameNetbird {
+			sConn.DelTable(t)
+			_ = sConn.Flush()
+		}
+	}
+}
+
 func verifyRule(t *testing.T, rule *nftables.Rule, sources []netip.Prefix, destination netip.Prefix, proto firewall.Protocol, sPort, dPort *firewall.Port, direction firewall.RuleDirection, action firewall.Action, expectSet bool) {
 	t.Helper()
 
@@ -627,7 +758,7 @@ func containsPort(exprs []expr.Any, port *firewall.Port, isSource bool) bool {
 
 func containsProtocol(exprs []expr.Any, proto firewall.Protocol) bool {
 	var metaFound, cmpFound bool
-	expectedProto, _ := protoToInt(proto)
+	expectedProto, _ := afIPv4.protoNum(proto)
 	for _, e := range exprs {
 		switch ex := e.(type) {
 		case *expr.Meta:
@@ -854,3 +985,55 @@ func TestRouter_AddNatRule_WithStaleEntry(t *testing.T) {
 	}
 	assert.Equal(t, 1, found, "NAT rule should exist in kernel")
 }
+
+func TestCalculateLastIP(t *testing.T) {
+	tests := []struct {
+		prefix string
+		want   string
+	}{
+		{"10.0.0.0/24", "10.0.0.255"},
+		{"10.0.0.0/32", "10.0.0.0"},
+		{"0.0.0.0/0", "255.255.255.255"},
+		{"192.168.1.0/28", "192.168.1.15"},
+		{"fd00::/64", "fd00::ffff:ffff:ffff:ffff"},
+		{"fd00::/128", "fd00::"},
+		{"2001:db8::/48", "2001:db8:0:ffff:ffff:ffff:ffff:ffff"},
+		{"::/0", "ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff"},
+	}
+	for _, tt := range tests {
+		t.Run(tt.prefix, func(t *testing.T) {
+			prefix := netip.MustParsePrefix(tt.prefix)
+			got := calculateLastIP(prefix)
+			assert.Equal(t, tt.want, got.String())
+		})
+	}
+}
+
+func TestConvertPrefixesToSet_IPv6(t *testing.T) {
+	r := &router{af: afIPv6}
+	prefixes := []netip.Prefix{
+		netip.MustParsePrefix("fd00::/64"),
+		netip.MustParsePrefix("2001:db8::1/128"),
+	}
+
+	elements := r.convertPrefixesToSet(prefixes)
+
+	// Each prefix produces 2 elements (start + end)
+	require.Len(t, elements, 4)
+
+	// fd00::/64 start
+	assert.Equal(t, netip.MustParseAddr("fd00::").As16(), [16]byte(elements[0].Key))
+	assert.False(t, elements[0].IntervalEnd)
+
+	// fd00::/64 end (fd00:0:0:1::, one past the last)
+	assert.Equal(t, netip.MustParseAddr("fd00:0:0:1::").As16(), [16]byte(elements[1].Key))
+	assert.True(t, elements[1].IntervalEnd)
+
+	// 2001:db8::1/128 start
+	assert.Equal(t, netip.MustParseAddr("2001:db8::1").As16(), [16]byte(elements[2].Key))
+	assert.False(t, elements[2].IntervalEnd)
+
+	// 2001:db8::1/128 end (2001:db8::2)
+	assert.Equal(t, netip.MustParseAddr("2001:db8::2").As16(), [16]byte(elements[3].Key))
+	assert.True(t, elements[3].IntervalEnd)
+}

client/firewall/uspfilter/allow_netbird_windows.go

@@ -5,8 +5,10 @@ import (
 	"os/exec"
 	"syscall"
 
+	"github.com/hashicorp/go-multierror"
 	log "github.com/sirupsen/logrus"
 
+	nberrors "github.com/netbirdio/netbird/client/errors"
 	"github.com/netbirdio/netbird/client/internal/statemanager"
 )
 
@@ -29,15 +31,20 @@ func (m *Manager) Close(*statemanager.Manager) error {
 		return nil
 	}
 
-	if !isFirewallRuleActive(firewallRuleName) {
-		return nil
+	var merr *multierror.Error
+	if isFirewallRuleActive(firewallRuleName) {
+		if err := manageFirewallRule(firewallRuleName, deleteRule); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("remove windows firewall rule: %w", err))
+		}
 	}
 
-	if err := manageFirewallRule(firewallRuleName, deleteRule); err != nil {
-		return fmt.Errorf("couldn't remove windows firewall: %w", err)
+	if isFirewallRuleActive(firewallRuleName + "-v6") {
+		if err := manageFirewallRule(firewallRuleName+"-v6", deleteRule); err != nil {
+			merr = multierror.Append(merr, fmt.Errorf("remove windows v6 firewall rule: %w", err))
+		}
 	}
 
-	return nil
+	return nberrors.FormatErrorOrNil(merr)
 }
 
 // AllowNetbird allows netbird interface traffic
@@ -46,17 +53,33 @@ func (m *Manager) AllowNetbird() error {
 		return nil
 	}
 
-	if isFirewallRuleActive(firewallRuleName) {
-		return nil
+	if !isFirewallRuleActive(firewallRuleName) {
+		if err := manageFirewallRule(firewallRuleName,
+			addRule,
+			"dir=in",
+			"enable=yes",
+			"action=allow",
+			"profile=any",
+			"localip="+m.wgIface.Address().IP.String(),
+		); err != nil {
+			return err
+		}
 	}
-	return manageFirewallRule(firewallRuleName,
-		addRule,
-		"dir=in",
-		"enable=yes",
-		"action=allow",
-		"profile=any",
-		"localip="+m.wgIface.Address().IP.String(),
-	)
+
+	if v6 := m.wgIface.Address().IPv6; v6.IsValid() && !isFirewallRuleActive(firewallRuleName+"-v6") {
+		if err := manageFirewallRule(firewallRuleName+"-v6",
+			addRule,
+			"dir=in",
+			"enable=yes",
+			"action=allow",
+			"profile=any",
+			"localip="+v6.String(),
+		); err != nil {
+			return err
+		}
+	}
+
+	return nil
 }
 
 func manageFirewallRule(ruleName string, action action, extraArgs ...string) error {

client/firewall/uspfilter/conntrack/common.go

@@ -1,8 +1,9 @@
 package conntrack
 
 import (
-	"fmt"
+	"net"
 	"net/netip"
+	"strconv"
 	"sync/atomic"
 	"time"
 
@@ -64,5 +65,7 @@ type ConnKey struct {
 }
 
 func (c ConnKey) String() string {
-	return fmt.Sprintf("%s:%d → %s:%d", c.SrcIP.Unmap(), c.SrcPort, c.DstIP.Unmap(), c.DstPort)
+	return net.JoinHostPort(c.SrcIP.Unmap().String(), strconv.Itoa(int(c.SrcPort))) +
+		" → " +
+		net.JoinHostPort(c.DstIP.Unmap().String(), strconv.Itoa(int(c.DstPort)))
 }

client/firewall/uspfilter/conntrack/icmp.go

@@ -21,9 +21,10 @@ const (
 	// ICMPCleanupInterval is how often we check for stale ICMP connections
 	ICMPCleanupInterval = 15 * time.Second
 
-	// MaxICMPPayloadLength is the maximum length of ICMP payload we consider for original packet info,
-	// which includes the IP header (20 bytes) and transport header (8 bytes)
-	MaxICMPPayloadLength = 28
+	// MaxICMPPayloadLength is the maximum length of ICMP payload we consider for original packet info.
+	// IPv4: 20-byte header + 8-byte transport = 28 bytes.
+	// IPv6: 40-byte header + 8-byte transport = 48 bytes.
+	MaxICMPPayloadLength = 48
 )
 
 // ICMPConnKey uniquely identifies an ICMP connection
@@ -74,32 +75,64 @@ func (info ICMPInfo) String() string {
 	return info.TypeCode.String()
 }
 
-// isErrorMessage returns true if this ICMP type carries original packet info
+// isErrorMessage returns true if this ICMP type carries original packet info.
+// Covers both ICMPv4 and ICMPv6 error types. Without a family field we match
+// both sets; type 3 overlaps (v4 DestUnreachable / v6 TimeExceeded) so it's
+// kept as a literal.
 func (info ICMPInfo) isErrorMessage() bool {
 	typ := info.TypeCode.Type()
-	return typ == 3 || // Destination Unreachable
-		typ == 5 || // Redirect
-		typ == 11 || // Time Exceeded
-		typ == 12 // Parameter Problem
+	// ICMPv4 error types
+	if typ == layers.ICMPv4TypeDestinationUnreachable ||
+		typ == layers.ICMPv4TypeRedirect ||
+		typ == layers.ICMPv4TypeTimeExceeded ||
+		typ == layers.ICMPv4TypeParameterProblem {
+		return true
+	}
+	// ICMPv6 error types (type 3 already matched above as v4 DestUnreachable)
+	if typ == layers.ICMPv6TypeDestinationUnreachable ||
+		typ == layers.ICMPv6TypePacketTooBig ||
+		typ == layers.ICMPv6TypeParameterProblem {
+		return true
+	}
+	return false
 }
 
 // parseOriginalPacket extracts info about the original packet from ICMP payload
 func (info ICMPInfo) parseOriginalPacket() string {
-	if info.PayloadLen < MaxICMPPayloadLength {
+	if info.PayloadLen == 0 {
 		return ""
 	}
 
-	// TODO: handle IPv6
-	if version := (info.PayloadData[0] >> 4) & 0xF; version != 4 {
+	version := (info.PayloadData[0] >> 4) & 0xF
+
+	var protocol uint8
+	var srcIP, dstIP net.IP
+	var transportData []byte
+
+	switch version {
+	case 4:
+		// 20-byte IPv4 header + 8-byte transport minimum
+		if info.PayloadLen < 28 {
+			return ""
+		}
+		protocol = info.PayloadData[9]
+		srcIP = net.IP(info.PayloadData[12:16])
+		dstIP = net.IP(info.PayloadData[16:20])
+		transportData = info.PayloadData[20:]
+	case 6:
+		// 40-byte IPv6 header + 8-byte transport minimum
+		if info.PayloadLen < 48 {
+			return ""
+		}
+		// Next Header field in IPv6 header
+		protocol = info.PayloadData[6]
+		srcIP = net.IP(info.PayloadData[8:24])
+		dstIP = net.IP(info.PayloadData[24:40])
+		transportData = info.PayloadData[40:]
+	default:
 		return ""
 	}
 
-	protocol := info.PayloadData[9]
-	srcIP := net.IP(info.PayloadData[12:16])
-	dstIP := net.IP(info.PayloadData[16:20])
-
-	transportData := info.PayloadData[20:]
-
 	switch nftypes.Protocol(protocol) {
 	case nftypes.TCP:
 		srcPort := uint16(transportData[0])<<8 | uint16(transportData[1])
@@ -247,9 +280,10 @@ func (t *ICMPTracker) track(
 	t.sendEvent(nftypes.TypeStart, conn, ruleId)
 }
 
-// IsValidInbound checks if an inbound ICMP Echo Reply matches a tracked request
+// IsValidInbound checks if an inbound ICMP Echo Reply matches a tracked request.
+// Accepts both ICMPv4 (type 0) and ICMPv6 (type 129) echo replies.
 func (t *ICMPTracker) IsValidInbound(srcIP netip.Addr, dstIP netip.Addr, id uint16, icmpType uint8, size int) bool {
-	if icmpType != uint8(layers.ICMPv4TypeEchoReply) {
+	if icmpType != uint8(layers.ICMPv4TypeEchoReply) && icmpType != uint8(layers.ICMPv6TypeEchoReply) {
 		return false
 	}
 
@@ -301,6 +335,13 @@ func (t *ICMPTracker) cleanup() {
 	}
 }
 
+func icmpProtocolForAddr(ip netip.Addr) nftypes.Protocol {
+	if ip.Is6() {
+		return nftypes.ICMPv6
+	}
+	return nftypes.ICMP
+}
+
 // Close stops the cleanup routine and releases resources
 func (t *ICMPTracker) Close() {
 	t.tickerCancel()
@@ -316,7 +357,7 @@ func (t *ICMPTracker) sendEvent(typ nftypes.Type, conn *ICMPConnTrack, ruleID []
 		Type:      typ,
 		RuleID:    ruleID,
 		Direction: conn.Direction,
-		Protocol:  nftypes.ICMP, // TODO: adjust for IPv6/icmpv6
+		Protocol:  icmpProtocolForAddr(conn.SourceIP),
 		SourceIP:  conn.SourceIP,
 		DestIP:    conn.DestIP,
 		ICMPType:  conn.ICMPType,
@@ -334,7 +375,7 @@ func (t *ICMPTracker) sendStartEvent(direction nftypes.Direction, srcIP netip.Ad
 		Type:      nftypes.TypeStart,
 		RuleID:    ruleID,
 		Direction: direction,
-		Protocol:  nftypes.ICMP,
+		Protocol:  icmpProtocolForAddr(srcIP),
 		SourceIP:  srcIP,
 		DestIP:    dstIP,
 		ICMPType:  typ,

client/firewall/uspfilter/filter.go

@@ -35,8 +35,10 @@ import (
 const (
 	layerTypeAll = 255
 
-	// ipTCPHeaderMinSize represents minimum IP (20) + TCP (20) header size for MSS calculation
-	ipTCPHeaderMinSize = 40
+	// ipv4TCPHeaderMinSize represents minimum IPv4 (20) + TCP (20) header size for MSS calculation
+	ipv4TCPHeaderMinSize = 40
+	// ipv6TCPHeaderMinSize represents minimum IPv6 (40) + TCP (20) header size for MSS calculation
+	ipv6TCPHeaderMinSize = 60
 )
 
 // serviceKey represents a protocol/port combination for netstack service registry
@@ -137,9 +139,21 @@ type Manager struct {
 	netstackServices     map[serviceKey]struct{}
 	netstackServiceMutex sync.RWMutex
 
-	mtu             uint16
-	mssClampValue   uint16
-	mssClampEnabled bool
+	mtu               uint16
+	mssClampValueIPv4 uint16
+	mssClampValueIPv6 uint16
+	mssClampEnabled   bool
+
+	// Only one hook per protocol is supported. Outbound direction only.
+	udpHookOut atomic.Pointer[packetHook]
+	tcpHookOut atomic.Pointer[packetHook]
+}
+
+// packetHook stores a registered hook for a specific IP:port.
+type packetHook struct {
+	ip   netip.Addr
+	port uint16
+	fn   func([]byte) bool
 }
 
 // decoder for packages
@@ -152,11 +166,28 @@ type decoder struct {
 	icmp4   layers.ICMPv4
 	icmp6   layers.ICMPv6
 	decoded []gopacket.LayerType
-	parser  *gopacket.DecodingLayerParser
+	parser4 *gopacket.DecodingLayerParser
+	parser6 *gopacket.DecodingLayerParser
 
 	dnatOrigPort uint16
 }
 
+// decodePacket decodes packet data using the appropriate parser based on IP version.
+func (d *decoder) decodePacket(data []byte) error {
+	if len(data) == 0 {
+		return errors.New("empty packet")
+	}
+	version := data[0] >> 4
+	switch version {
+	case 4:
+		return d.parser4.DecodeLayers(data, &d.decoded)
+	case 6:
+		return d.parser6.DecodeLayers(data, &d.decoded)
+	default:
+		return fmt.Errorf("unknown IP version %d", version)
+	}
+}
+
 // Create userspace firewall manager constructor
 func Create(iface common.IFaceMapper, disableServerRoutes bool, flowLogger nftypes.FlowLogger, mtu uint16) (*Manager, error) {
 	return create(iface, nil, disableServerRoutes, flowLogger, mtu)
@@ -214,11 +245,17 @@ func create(iface common.IFaceMapper, nativeFirewall firewall.Manager, disableSe
 				d := &decoder{
 					decoded: []gopacket.LayerType{},
 				}
-				d.parser = gopacket.NewDecodingLayerParser(
+				d.parser4 = gopacket.NewDecodingLayerParser(
 					layers.LayerTypeIPv4,
 					&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 				)
-				d.parser.IgnoreUnsupported = true
+				d.parser4.IgnoreUnsupported = true
+
+				d.parser6 = gopacket.NewDecodingLayerParser(
+					layers.LayerTypeIPv6,
+					&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+				)
+				d.parser6.IgnoreUnsupported = true
 				return d
 			},
 		},
@@ -244,7 +281,8 @@ func create(iface common.IFaceMapper, nativeFirewall firewall.Manager, disableSe
 
 	if !disableMSSClamping {
 		m.mssClampEnabled = true
-		m.mssClampValue = mtu - ipTCPHeaderMinSize
+		m.mssClampValueIPv4 = mtu - ipv4TCPHeaderMinSize
+		m.mssClampValueIPv6 = mtu - ipv6TCPHeaderMinSize
 	}
 	if err := m.localipmanager.UpdateLocalIPs(iface); err != nil {
 		return nil, fmt.Errorf("update local IPs: %w", err)
@@ -271,9 +309,14 @@ func (m *Manager) blockInvalidRouted(iface common.IFaceMapper) (firewall.Rule, e
 	wgPrefix := iface.Address().Network
 	log.Debugf("blocking invalid routed traffic for %s", wgPrefix)
 
+	sources := []netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)}
+	if v6 := iface.Address().IPv6Net; v6.IsValid() {
+		sources = append(sources, netip.PrefixFrom(netip.IPv6Unspecified(), 0))
+	}
+
 	rule, err := m.addRouteFiltering(
 		nil,
-		[]netip.Prefix{netip.PrefixFrom(netip.IPv4Unspecified(), 0)},
+		sources,
 		firewall.Network{Prefix: wgPrefix},
 		firewall.ProtocolALL,
 		nil,
@@ -281,7 +324,22 @@ func (m *Manager) blockInvalidRouted(iface common.IFaceMapper) (firewall.Rule, e
 		firewall.ActionDrop,
 	)
 	if err != nil {
-		return nil, fmt.Errorf("block wg nte : %w", err)
+		return nil, fmt.Errorf("block wg v4 net: %w", err)
+	}
+
+	if v6Net := iface.Address().IPv6Net; v6Net.IsValid() {
+		log.Debugf("blocking invalid routed traffic for %s", v6Net)
+		if _, err := m.addRouteFiltering(
+			nil,
+			sources,
+			firewall.Network{Prefix: v6Net},
+			firewall.ProtocolALL,
+			nil,
+			nil,
+			firewall.ActionDrop,
+		); err != nil {
+			return nil, fmt.Errorf("block wg v6 net: %w", err)
+		}
 	}
 
 	// TODO: Block networks that we're a client of
@@ -498,7 +556,7 @@ func (m *Manager) addRouteFiltering(
 		mgmtId:     id,
 		sources:    sources,
 		dstSet:     destination.Set,
-		protoLayer: protoToLayer(proto, layers.LayerTypeIPv4),
+		protoLayer: protoToLayer(proto, ipLayerFromPrefix(destination.Prefix)),
 		srcPort:    sPort,
 		dstPort:    dPort,
 		action:     action,
@@ -594,6 +652,8 @@ func (m *Manager) resetState() {
 	maps.Clear(m.incomingRules)
 	maps.Clear(m.routeRulesMap)
 	m.routeRules = m.routeRules[:0]
+	m.udpHookOut.Store(nil)
+	m.tcpHookOut.Store(nil)
 
 	if m.udpTracker != nil {
 		m.udpTracker.Close()
@@ -650,11 +710,7 @@ func (m *Manager) UpdateSet(set firewall.Set, prefixes []netip.Prefix) error {
 	}
 
 	destinations := matches[0].destinations
-	for _, prefix := range prefixes {
-		if prefix.Addr().Is4() {
-			destinations = append(destinations, prefix)
-		}
-	}
+	destinations = append(destinations, prefixes...)
 
 	slices.SortFunc(destinations, func(a, b netip.Prefix) int {
 		cmp := a.Addr().Compare(b.Addr())
@@ -693,7 +749,7 @@ func (m *Manager) filterOutbound(packetData []byte, size int) bool {
 	d := m.decoders.Get().(*decoder)
 	defer m.decoders.Put(d)
 
-	if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+	if err := d.decodePacket(packetData); err != nil {
 		return false
 	}
 
@@ -713,6 +769,9 @@ func (m *Manager) filterOutbound(packetData []byte, size int) bool {
 			return true
 		}
 	case layers.LayerTypeTCP:
+		if m.tcpHooksDrop(uint16(d.tcp.DstPort), dstIP, packetData) {
+			return true
+		}
 		// Clamp MSS on all TCP SYN packets, including those from local IPs.
 		// SNATed routed traffic may appear as local IP but still requires clamping.
 		if m.mssClampEnabled {
@@ -774,12 +833,28 @@ func (m *Manager) clampTCPMSS(packetData []byte, d *decoder) bool {
 		return false
 	}
 
+	var mssClampValue uint16
+	var ipHeaderSize int
+	switch d.decoded[0] {
+	case layers.LayerTypeIPv4:
+		mssClampValue = m.mssClampValueIPv4
+		ipHeaderSize = int(d.ip4.IHL) * 4
+		if ipHeaderSize < 20 {
+			return false
+		}
+	case layers.LayerTypeIPv6:
+		mssClampValue = m.mssClampValueIPv6
+		ipHeaderSize = 40
+	default:
+		return false
+	}
+
 	mssOptionIndex := -1
 	var currentMSS uint16
 	for i, opt := range d.tcp.Options {
 		if opt.OptionType == layers.TCPOptionKindMSS && len(opt.OptionData) == 2 {
 			currentMSS = binary.BigEndian.Uint16(opt.OptionData)
-			if currentMSS > m.mssClampValue {
+			if currentMSS > mssClampValue {
 				mssOptionIndex = i
 				break
 			}
@@ -790,20 +865,15 @@ func (m *Manager) clampTCPMSS(packetData []byte, d *decoder) bool {
 		return false
 	}
 
-	ipHeaderSize := int(d.ip4.IHL) * 4
-	if ipHeaderSize < 20 {
+	if !m.updateMSSOption(packetData, d, mssOptionIndex, mssClampValue, ipHeaderSize) {
 		return false
 	}
 
-	if !m.updateMSSOption(packetData, d, mssOptionIndex, ipHeaderSize) {
-		return false
-	}
-
-	m.logger.Trace2("Clamped TCP MSS from %d to %d", currentMSS, m.mssClampValue)
+	m.logger.Trace2("Clamped TCP MSS from %d to %d", currentMSS, mssClampValue)
 	return true
 }
 
-func (m *Manager) updateMSSOption(packetData []byte, d *decoder, mssOptionIndex, ipHeaderSize int) bool {
+func (m *Manager) updateMSSOption(packetData []byte, d *decoder, mssOptionIndex int, mssClampValue uint16, ipHeaderSize int) bool {
 	tcpHeaderStart := ipHeaderSize
 	tcpOptionsStart := tcpHeaderStart + 20
 
@@ -818,7 +888,7 @@ func (m *Manager) updateMSSOption(packetData []byte, d *decoder, mssOptionIndex,
 	}
 
 	mssValueOffset := optOffset + 2
-	binary.BigEndian.PutUint16(packetData[mssValueOffset:mssValueOffset+2], m.mssClampValue)
+	binary.BigEndian.PutUint16(packetData[mssValueOffset:mssValueOffset+2], mssClampValue)
 
 	m.recalculateTCPChecksum(packetData, d, tcpHeaderStart)
 	return true
@@ -828,18 +898,32 @@ func (m *Manager) recalculateTCPChecksum(packetData []byte, d *decoder, tcpHeade
 	tcpLayer := packetData[tcpHeaderStart:]
 	tcpLength := len(packetData) - tcpHeaderStart
 
+	// Zero out existing checksum
 	tcpLayer[16] = 0
 	tcpLayer[17] = 0
 
+	// Build pseudo-header checksum based on IP version
 	var pseudoSum uint32
-	pseudoSum += uint32(d.ip4.SrcIP[0])<<8 | uint32(d.ip4.SrcIP[1])
-	pseudoSum += uint32(d.ip4.SrcIP[2])<<8 | uint32(d.ip4.SrcIP[3])
-	pseudoSum += uint32(d.ip4.DstIP[0])<<8 | uint32(d.ip4.DstIP[1])
-	pseudoSum += uint32(d.ip4.DstIP[2])<<8 | uint32(d.ip4.DstIP[3])
-	pseudoSum += uint32(d.ip4.Protocol)
-	pseudoSum += uint32(tcpLength)
+	switch d.decoded[0] {
+	case layers.LayerTypeIPv4:
+		pseudoSum += uint32(d.ip4.SrcIP[0])<<8 | uint32(d.ip4.SrcIP[1])
+		pseudoSum += uint32(d.ip4.SrcIP[2])<<8 | uint32(d.ip4.SrcIP[3])
+		pseudoSum += uint32(d.ip4.DstIP[0])<<8 | uint32(d.ip4.DstIP[1])
+		pseudoSum += uint32(d.ip4.DstIP[2])<<8 | uint32(d.ip4.DstIP[3])
+		pseudoSum += uint32(d.ip4.Protocol)
+		pseudoSum += uint32(tcpLength)
+	case layers.LayerTypeIPv6:
+		for i := 0; i < 16; i += 2 {
+			pseudoSum += uint32(d.ip6.SrcIP[i])<<8 | uint32(d.ip6.SrcIP[i+1])
+		}
+		for i := 0; i < 16; i += 2 {
+			pseudoSum += uint32(d.ip6.DstIP[i])<<8 | uint32(d.ip6.DstIP[i+1])
+		}
+		pseudoSum += uint32(tcpLength)
+		pseudoSum += uint32(layers.IPProtocolTCP)
+	}
 
-	var sum = pseudoSum
+	sum := pseudoSum
 	for i := 0; i < tcpLength-1; i += 2 {
 		sum += uint32(tcpLayer[i])<<8 | uint32(tcpLayer[i+1])
 	}
@@ -877,6 +961,9 @@ func (m *Manager) trackOutbound(d *decoder, srcIP, dstIP netip.Addr, packetData
 		}
 	case layers.LayerTypeICMPv4:
 		m.icmpTracker.TrackOutbound(srcIP, dstIP, d.icmp4.Id, d.icmp4.TypeCode, d.icmp4.Payload, size)
+	case layers.LayerTypeICMPv6:
+		id, tc := icmpv6EchoFields(d)
+		m.icmpTracker.TrackOutbound(srcIP, dstIP, id, tc, d.icmp6.Payload, size)
 	}
 }
 
@@ -890,43 +977,29 @@ func (m *Manager) trackInbound(d *decoder, srcIP, dstIP netip.Addr, ruleID []byt
 		m.tcpTracker.TrackInbound(srcIP, dstIP, uint16(d.tcp.SrcPort), uint16(d.tcp.DstPort), flags, ruleID, size, d.dnatOrigPort)
 	case layers.LayerTypeICMPv4:
 		m.icmpTracker.TrackInbound(srcIP, dstIP, d.icmp4.Id, d.icmp4.TypeCode, ruleID, d.icmp4.Payload, size)
+	case layers.LayerTypeICMPv6:
+		id, tc := icmpv6EchoFields(d)
+		m.icmpTracker.TrackInbound(srcIP, dstIP, id, tc, ruleID, d.icmp6.Payload, size)
 	}
 
 	d.dnatOrigPort = 0
 }
 
-// udpHooksDrop checks if any UDP hooks should drop the packet
 func (m *Manager) udpHooksDrop(dport uint16, dstIP netip.Addr, packetData []byte) bool {
-	m.mutex.RLock()
-	defer m.mutex.RUnlock()
+	return hookMatches(m.udpHookOut.Load(), dstIP, dport, packetData)
+}
 
-	// Check specific destination IP first
-	if rules, exists := m.outgoingRules[dstIP]; exists {
-		for _, rule := range rules {
-			if rule.udpHook != nil && portsMatch(rule.dPort, dport) {
-				return rule.udpHook(packetData)
-			}
-		}
+func (m *Manager) tcpHooksDrop(dport uint16, dstIP netip.Addr, packetData []byte) bool {
+	return hookMatches(m.tcpHookOut.Load(), dstIP, dport, packetData)
+}
+
+func hookMatches(h *packetHook, dstIP netip.Addr, dport uint16, packetData []byte) bool {
+	if h == nil {
+		return false
 	}
-
-	// Check IPv4 unspecified address
-	if rules, exists := m.outgoingRules[netip.IPv4Unspecified()]; exists {
-		for _, rule := range rules {
-			if rule.udpHook != nil && portsMatch(rule.dPort, dport) {
-				return rule.udpHook(packetData)
-			}
-		}
+	if h.ip == dstIP && h.port == dport {
+		return h.fn(packetData)
 	}
-
-	// Check IPv6 unspecified address
-	if rules, exists := m.outgoingRules[netip.IPv6Unspecified()]; exists {
-		for _, rule := range rules {
-			if rule.udpHook != nil && portsMatch(rule.dPort, dport) {
-				return rule.udpHook(packetData)
-			}
-		}
-	}
-
 	return false
 }
 
@@ -949,15 +1022,19 @@ func (m *Manager) filterInbound(packetData []byte, size int) bool {
 
 	// TODO: pass fragments of routed packets to forwarder
 	if fragment {
-		m.logger.Trace4("packet is a fragment: src=%v dst=%v id=%v flags=%v",
-			srcIP, dstIP, d.ip4.Id, d.ip4.Flags)
+		if d.decoded[0] == layers.LayerTypeIPv4 {
+			m.logger.Trace4("packet is a fragment: src=%v dst=%v id=%v flags=%v",
+				srcIP, dstIP, d.ip4.Id, d.ip4.Flags)
+		} else {
+			m.logger.Trace2("packet is an IPv6 fragment: src=%v dst=%v", srcIP, dstIP)
+		}
 		return false
 	}
 
 	// TODO: optimize port DNAT by caching matched rules in conntrack
 	if translated := m.translateInboundPortDNAT(packetData, d, srcIP, dstIP); translated {
 		// Re-decode after port DNAT translation to update port information
-		if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+		if err := d.decodePacket(packetData); err != nil {
 			m.logger.Error1("failed to re-decode packet after port DNAT: %v", err)
 			return true
 		}
@@ -966,7 +1043,7 @@ func (m *Manager) filterInbound(packetData []byte, size int) bool {
 
 	if translated := m.translateInboundReverse(packetData, d); translated {
 		// Re-decode after translation to get original addresses
-		if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+		if err := d.decodePacket(packetData); err != nil {
 			m.logger.Error1("failed to re-decode packet after reverse DNAT: %v", err)
 			return true
 		}
@@ -1098,6 +1175,48 @@ func (m *Manager) handleRoutedTraffic(d *decoder, srcIP, dstIP netip.Addr, packe
 	return true
 }
 
+// icmpv6EchoFields extracts the echo identifier from an ICMPv6 packet and maps
+// the ICMPv6 type code to an ICMPv4TypeCode so the ICMP conntrack can handle
+// both families uniformly. The echo ID is in the first two payload bytes.
+func icmpv6EchoFields(d *decoder) (id uint16, tc layers.ICMPv4TypeCode) {
+	if len(d.icmp6.Payload) >= 2 {
+		id = uint16(d.icmp6.Payload[0])<<8 | uint16(d.icmp6.Payload[1])
+	}
+	// Map ICMPv6 echo types to ICMPv4 equivalents for unified tracking.
+	switch d.icmp6.TypeCode.Type() {
+	case layers.ICMPv6TypeEchoRequest:
+		tc = layers.CreateICMPv4TypeCode(layers.ICMPv4TypeEchoRequest, 0)
+	case layers.ICMPv6TypeEchoReply:
+		tc = layers.CreateICMPv4TypeCode(layers.ICMPv4TypeEchoReply, 0)
+	default:
+		tc = layers.CreateICMPv4TypeCode(d.icmp6.TypeCode.Type(), d.icmp6.TypeCode.Code())
+	}
+	return id, tc
+}
+
+// protoLayerMatches checks if a packet's protocol layer matches a rule's expected
+// protocol layer. ICMPv4 and ICMPv6 are treated as equivalent when matching
+// ICMP rules since management sends a single ICMP rule for both families.
+func protoLayerMatches(ruleLayer, packetLayer gopacket.LayerType) bool {
+	if ruleLayer == packetLayer {
+		return true
+	}
+	if ruleLayer == layers.LayerTypeICMPv4 && packetLayer == layers.LayerTypeICMPv6 {
+		return true
+	}
+	if ruleLayer == layers.LayerTypeICMPv6 && packetLayer == layers.LayerTypeICMPv4 {
+		return true
+	}
+	return false
+}
+
+func ipLayerFromPrefix(p netip.Prefix) gopacket.LayerType {
+	if p.Addr().Is6() {
+		return layers.LayerTypeIPv6
+	}
+	return layers.LayerTypeIPv4
+}
+
 func protoToLayer(proto firewall.Protocol, ipLayer gopacket.LayerType) gopacket.LayerType {
 	switch proto {
 	case firewall.ProtocolTCP:
@@ -1121,8 +1240,10 @@ func getProtocolFromPacket(d *decoder) nftypes.Protocol {
 		return nftypes.TCP
 	case layers.LayerTypeUDP:
 		return nftypes.UDP
-	case layers.LayerTypeICMPv4, layers.LayerTypeICMPv6:
+	case layers.LayerTypeICMPv4:
 		return nftypes.ICMP
+	case layers.LayerTypeICMPv6:
+		return nftypes.ICMPv6
 	default:
 		return nftypes.ProtocolUnknown
 	}
@@ -1143,7 +1264,7 @@ func getPortsFromPacket(d *decoder) (srcPort, dstPort uint16) {
 // It returns true, false if the packet is valid and not a fragment.
 // It returns true, true if the packet is a fragment and valid.
 func (m *Manager) isValidPacket(d *decoder, packetData []byte) (bool, bool) {
-	if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+	if err := d.decodePacket(packetData); err != nil {
 		m.logger.Trace1("couldn't decode packet, err: %s", err)
 		return false, false
 	}
@@ -1156,10 +1277,18 @@ func (m *Manager) isValidPacket(d *decoder, packetData []byte) (bool, bool) {
 	}
 
 	// Fragments are also valid
-	if l == 1 && d.decoded[0] == layers.LayerTypeIPv4 {
-		ip4 := d.ip4
-		if ip4.Flags&layers.IPv4MoreFragments != 0 || ip4.FragOffset != 0 {
-			return true, true
+	if l == 1 {
+		switch d.decoded[0] {
+		case layers.LayerTypeIPv4:
+			if d.ip4.Flags&layers.IPv4MoreFragments != 0 || d.ip4.FragOffset != 0 {
+				return true, true
+			}
+		case layers.LayerTypeIPv6:
+			// IPv6 uses Fragment extension header (NextHeader=44). If gopacket
+			// only decoded the IPv6 layer, the transport is in a fragment.
+			if d.ip6.NextHeader == layers.IPProtocolIPv6Fragment {
+				return true, true
+			}
 		}
 	}
 
@@ -1197,21 +1326,34 @@ func (m *Manager) isValidTrackedConnection(d *decoder, srcIP, dstIP netip.Addr,
 			size,
 		)
 
-		// TODO: ICMPv6
+	case layers.LayerTypeICMPv6:
+		id, _ := icmpv6EchoFields(d)
+		return m.icmpTracker.IsValidInbound(
+			srcIP,
+			dstIP,
+			id,
+			d.icmp6.TypeCode.Type(),
+			size,
+		)
 	}
 
 	return false
 }
 
-// isSpecialICMP returns true if the packet is a special ICMP packet that should be allowed
+// isSpecialICMP returns true if the packet is a special ICMP error packet that should be allowed.
 func (m *Manager) isSpecialICMP(d *decoder) bool {
-	if d.decoded[1] != layers.LayerTypeICMPv4 {
-		return false
+	switch d.decoded[1] {
+	case layers.LayerTypeICMPv4:
+		icmpType := d.icmp4.TypeCode.Type()
+		return icmpType == layers.ICMPv4TypeDestinationUnreachable ||
+			icmpType == layers.ICMPv4TypeTimeExceeded
+	case layers.LayerTypeICMPv6:
+		icmpType := d.icmp6.TypeCode.Type()
+		return icmpType == layers.ICMPv6TypeDestinationUnreachable ||
+			icmpType == layers.ICMPv6TypePacketTooBig ||
+			icmpType == layers.ICMPv6TypeTimeExceeded
 	}
-
-	icmpType := d.icmp4.TypeCode.Type()
-	return icmpType == layers.ICMPv4TypeDestinationUnreachable ||
-		icmpType == layers.ICMPv4TypeTimeExceeded
+	return false
 }
 
 func (m *Manager) peerACLsBlock(srcIP netip.Addr, d *decoder, packetData []byte) ([]byte, bool) {
@@ -1268,7 +1410,7 @@ func validateRule(ip netip.Addr, packetData []byte, rules map[string]PeerRule, d
 			return rule.mgmtId, rule.drop, true
 		}
 
-		if payloadLayer != rule.protoLayer {
+		if !protoLayerMatches(rule.protoLayer, payloadLayer) {
 			continue
 		}
 
@@ -1278,12 +1420,6 @@ func validateRule(ip netip.Addr, packetData []byte, rules map[string]PeerRule, d
 				return rule.mgmtId, rule.drop, true
 			}
 		case layers.LayerTypeUDP:
-			// if rule has UDP hook (and if we are here we match this rule)
-			// we ignore rule.drop and call this hook
-			if rule.udpHook != nil {
-				return rule.mgmtId, rule.udpHook(packetData), true
-			}
-
 			if portsMatch(rule.sPort, uint16(d.udp.SrcPort)) && portsMatch(rule.dPort, uint16(d.udp.DstPort)) {
 				return rule.mgmtId, rule.drop, true
 			}
@@ -1309,8 +1445,7 @@ func (m *Manager) routeACLsPass(srcIP, dstIP netip.Addr, protoLayer gopacket.Lay
 }
 
 func (m *Manager) ruleMatches(rule *RouteRule, srcAddr, dstAddr netip.Addr, protoLayer gopacket.LayerType, srcPort, dstPort uint16) bool {
-	// TODO: handle ipv6 vs ipv4 icmp rules
-	if rule.protoLayer != layerTypeAll && rule.protoLayer != protoLayer {
+	if rule.protoLayer != layerTypeAll && !protoLayerMatches(rule.protoLayer, protoLayer) {
 		return false
 	}
 
@@ -1342,65 +1477,30 @@ func (m *Manager) ruleMatches(rule *RouteRule, srcAddr, dstAddr netip.Addr, prot
 	return sourceMatched
 }
 
-// AddUDPPacketHook calls hook when UDP packet from given direction matched
-//
-// Hook function returns flag which indicates should be the matched package dropped or not
-func (m *Manager) AddUDPPacketHook(in bool, ip netip.Addr, dPort uint16, hook func(packet []byte) bool) string {
-	r := PeerRule{
-		id:         uuid.New().String(),
-		ip:         ip,
-		protoLayer: layers.LayerTypeUDP,
-		dPort:      &firewall.Port{Values: []uint16{dPort}},
-		ipLayer:    layers.LayerTypeIPv6,
-		udpHook:    hook,
+// SetUDPPacketHook sets the outbound UDP packet hook. Pass nil hook to remove.
+func (m *Manager) SetUDPPacketHook(ip netip.Addr, dPort uint16, hook func(packet []byte) bool) {
+	if hook == nil {
+		m.udpHookOut.Store(nil)
+		return
 	}
-
-	if ip.Is4() {
-		r.ipLayer = layers.LayerTypeIPv4
-	}
-
-	m.mutex.Lock()
-	if in {
-		// Incoming UDP hooks are stored in allow rules map
-		if _, ok := m.incomingRules[r.ip]; !ok {
-			m.incomingRules[r.ip] = make(map[string]PeerRule)
-		}
-		m.incomingRules[r.ip][r.id] = r
-	} else {
-		if _, ok := m.outgoingRules[r.ip]; !ok {
-			m.outgoingRules[r.ip] = make(map[string]PeerRule)
-		}
-		m.outgoingRules[r.ip][r.id] = r
-	}
-	m.mutex.Unlock()
-
-	return r.id
+	m.udpHookOut.Store(&packetHook{
+		ip:   ip,
+		port: dPort,
+		fn:   hook,
+	})
 }
 
-// RemovePacketHook removes packet hook by given ID
-func (m *Manager) RemovePacketHook(hookID string) error {
-	m.mutex.Lock()
-	defer m.mutex.Unlock()
-
-	// Check incoming hooks (stored in allow rules)
-	for _, arr := range m.incomingRules {
-		for _, r := range arr {
-			if r.id == hookID {
-				delete(arr, r.id)
-				return nil
-			}
-		}
+// SetTCPPacketHook sets the outbound TCP packet hook. Pass nil hook to remove.
+func (m *Manager) SetTCPPacketHook(ip netip.Addr, dPort uint16, hook func(packet []byte) bool) {
+	if hook == nil {
+		m.tcpHookOut.Store(nil)
+		return
 	}
-	// Check outgoing hooks
-	for _, arr := range m.outgoingRules {
-		for _, r := range arr {
-			if r.id == hookID {
-				delete(arr, r.id)
-				return nil
-			}
-		}
-	}
-	return fmt.Errorf("hook with given id not found")
+	m.tcpHookOut.Store(&packetHook{
+		ip:   ip,
+		port: dPort,
+		fn:   hook,
+	})
 }
 
 // SetLogLevel sets the log level for the firewall manager
@@ -1515,7 +1615,8 @@ func (m *Manager) shouldForward(d *decoder, dstIP netip.Addr) bool {
 	}
 
 	// traffic to our other local interfaces (not NetBird IP) - always forward
-	if dstIP != m.wgIface.Address().IP {
+	addr := m.wgIface.Address()
+	if dstIP != addr.IP && (!addr.IPv6.IsValid() || dstIP != addr.IPv6) {
 		return true
 	}
 

client/firewall/uspfilter/filter_bench_test.go

@@ -1023,7 +1023,8 @@ func BenchmarkMSSClamping(b *testing.B) {
 			}()
 
 			manager.mssClampEnabled = true
-			manager.mssClampValue = 1240
+			manager.mssClampValueIPv4 = 1240
+			manager.mssClampValueIPv6 = 1220
 
 			srcIP := net.ParseIP("100.64.0.2")
 			dstIP := net.ParseIP("8.8.8.8")
@@ -1088,7 +1089,8 @@ func BenchmarkMSSClampingOverhead(b *testing.B) {
 
 			manager.mssClampEnabled = sc.enabled
 			if sc.enabled {
-				manager.mssClampValue = 1240
+				manager.mssClampValueIPv4 = 1240
+				manager.mssClampValueIPv6 = 1220
 			}
 
 			srcIP := net.ParseIP("100.64.0.2")
@@ -1141,7 +1143,8 @@ func BenchmarkMSSClampingMemory(b *testing.B) {
 			}()
 
 			manager.mssClampEnabled = true
-			manager.mssClampValue = 1240
+			manager.mssClampValueIPv4 = 1240
+			manager.mssClampValueIPv6 = 1220
 
 			srcIP := net.ParseIP("100.64.0.2")
 			dstIP := net.ParseIP("8.8.8.8")

client/firewall/uspfilter/filter_filter_test.go

@@ -539,53 +539,236 @@ func TestPeerACLFiltering(t *testing.T) {
 	}
 }
 
+func TestPeerACLFilteringIPv6(t *testing.T) {
+	localIP := netip.MustParseAddr("100.10.0.100")
+	localIPv6 := netip.MustParseAddr("fd00::100")
+	wgNet := netip.MustParsePrefix("100.10.0.0/16")
+	wgNetV6 := netip.MustParsePrefix("fd00::/64")
+
+	ifaceMock := &IFaceMock{
+		SetFilterFunc: func(device.PacketFilter) error { return nil },
+		AddressFunc: func() wgaddr.Address {
+			return wgaddr.Address{
+				IP:      localIP,
+				Network: wgNet,
+				IPv6:    localIPv6,
+				IPv6Net: wgNetV6,
+			}
+		},
+	}
+
+	manager, err := Create(ifaceMock, false, flowLogger, iface.DefaultMTU)
+	require.NoError(t, err)
+	t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
+
+	err = manager.UpdateLocalIPs()
+	require.NoError(t, err)
+
+	testCases := []struct {
+		name            string
+		srcIP           string
+		dstIP           string
+		proto           fw.Protocol
+		srcPort         uint16
+		dstPort         uint16
+		ruleIP          string
+		ruleProto       fw.Protocol
+		ruleDstPort     *fw.Port
+		ruleAction      fw.Action
+		shouldBeBlocked bool
+	}{
+		{
+			name:            "IPv6: allow TCP from peer",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolTCP,
+			srcPort:         12345,
+			dstPort:         443,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolTCP,
+			ruleDstPort:     &fw.Port{Values: []uint16{443}},
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: false,
+		},
+		{
+			name:            "IPv6: allow UDP from peer",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolUDP,
+			srcPort:         12345,
+			dstPort:         53,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolUDP,
+			ruleDstPort:     &fw.Port{Values: []uint16{53}},
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: false,
+		},
+		{
+			name:            "IPv6: allow ICMPv6 from peer",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolICMP,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolICMP,
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: false,
+		},
+		{
+			name:            "IPv6: block TCP without rule",
+			srcIP:           "fd00::2",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolTCP,
+			srcPort:         12345,
+			dstPort:         443,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolTCP,
+			ruleDstPort:     &fw.Port{Values: []uint16{443}},
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: true,
+		},
+		{
+			name:            "IPv6: drop rule",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolTCP,
+			srcPort:         12345,
+			dstPort:         22,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolTCP,
+			ruleDstPort:     &fw.Port{Values: []uint16{22}},
+			ruleAction:      fw.ActionDrop,
+			shouldBeBlocked: true,
+		},
+		{
+			name:            "IPv6: allow all protocols",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolUDP,
+			srcPort:         12345,
+			dstPort:         9999,
+			ruleIP:          "fd00::1",
+			ruleProto:       fw.ProtocolALL,
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: false,
+		},
+		{
+			name:            "IPv6: v4 wildcard ICMP rule matches ICMPv6 via protoLayerMatches",
+			srcIP:           "fd00::1",
+			dstIP:           "fd00::100",
+			proto:           fw.ProtocolICMP,
+			ruleIP:          "0.0.0.0",
+			ruleProto:       fw.ProtocolICMP,
+			ruleAction:      fw.ActionAccept,
+			shouldBeBlocked: false,
+		},
+	}
+
+	t.Run("IPv6 implicit DROP (no rules)", func(t *testing.T) {
+		packet := createTestPacket(t, "fd00::1", "fd00::100", fw.ProtocolTCP, 12345, 443)
+		isDropped := manager.FilterInbound(packet, 0)
+		require.True(t, isDropped, "IPv6 packet should be dropped when no rules exist")
+	})
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			if tc.ruleAction == fw.ActionDrop {
+				rules, err := manager.AddPeerFiltering(nil, net.ParseIP(tc.ruleIP), fw.ProtocolALL, nil, nil, fw.ActionAccept, "")
+				require.NoError(t, err)
+				t.Cleanup(func() {
+					for _, rule := range rules {
+						require.NoError(t, manager.DeletePeerRule(rule))
+					}
+				})
+			}
+
+			rules, err := manager.AddPeerFiltering(nil, net.ParseIP(tc.ruleIP), tc.ruleProto, nil, tc.ruleDstPort, tc.ruleAction, "")
+			require.NoError(t, err)
+			require.NotEmpty(t, rules)
+			t.Cleanup(func() {
+				for _, rule := range rules {
+					require.NoError(t, manager.DeletePeerRule(rule))
+				}
+			})
+
+			packet := createTestPacket(t, tc.srcIP, tc.dstIP, tc.proto, tc.srcPort, tc.dstPort)
+			isDropped := manager.FilterInbound(packet, 0)
+			require.Equal(t, tc.shouldBeBlocked, isDropped, "packet filter result mismatch")
+		})
+	}
+}
+
 func createTestPacket(t *testing.T, srcIP, dstIP string, proto fw.Protocol, srcPort, dstPort uint16) []byte {
 	t.Helper()
 
+	src := net.ParseIP(srcIP)
+	dst := net.ParseIP(dstIP)
+
 	buf := gopacket.NewSerializeBuffer()
 	opts := gopacket.SerializeOptions{
 		ComputeChecksums: true,
 		FixLengths:       true,
 	}
 
-	ipLayer := &layers.IPv4{
-		Version: 4,
-		TTL:     64,
-		SrcIP:   net.ParseIP(srcIP),
-		DstIP:   net.ParseIP(dstIP),
-	}
+	// Detect address family
+	isV6 := src.To4() == nil
 
 	var err error
-	switch proto {
-	case fw.ProtocolTCP:
-		ipLayer.Protocol = layers.IPProtocolTCP
-		tcp := &layers.TCP{
-			SrcPort: layers.TCPPort(srcPort),
-			DstPort: layers.TCPPort(dstPort),
-		}
-		err = tcp.SetNetworkLayerForChecksum(ipLayer)
-		require.NoError(t, err)
-		err = gopacket.SerializeLayers(buf, opts, ipLayer, tcp)
 
-	case fw.ProtocolUDP:
-		ipLayer.Protocol = layers.IPProtocolUDP
-		udp := &layers.UDP{
-			SrcPort: layers.UDPPort(srcPort),
-			DstPort: layers.UDPPort(dstPort),
+	if isV6 {
+		ip6 := &layers.IPv6{
+			Version:  6,
+			HopLimit: 64,
+			SrcIP:    src,
+			DstIP:    dst,
 		}
-		err = udp.SetNetworkLayerForChecksum(ipLayer)
-		require.NoError(t, err)
-		err = gopacket.SerializeLayers(buf, opts, ipLayer, udp)
 
-	case fw.ProtocolICMP:
-		ipLayer.Protocol = layers.IPProtocolICMPv4
-		icmp := &layers.ICMPv4{
-			TypeCode: layers.CreateICMPv4TypeCode(layers.ICMPv4TypeEchoRequest, 0),
+		switch proto {
+		case fw.ProtocolTCP:
+			ip6.NextHeader = layers.IPProtocolTCP
+			tcp := &layers.TCP{SrcPort: layers.TCPPort(srcPort), DstPort: layers.TCPPort(dstPort)}
+			_ = tcp.SetNetworkLayerForChecksum(ip6)
+			err = gopacket.SerializeLayers(buf, opts, ip6, tcp)
+		case fw.ProtocolUDP:
+			ip6.NextHeader = layers.IPProtocolUDP
+			udp := &layers.UDP{SrcPort: layers.UDPPort(srcPort), DstPort: layers.UDPPort(dstPort)}
+			_ = udp.SetNetworkLayerForChecksum(ip6)
+			err = gopacket.SerializeLayers(buf, opts, ip6, udp)
+		case fw.ProtocolICMP:
+			ip6.NextHeader = layers.IPProtocolICMPv6
+			icmp := &layers.ICMPv6{
+				TypeCode: layers.CreateICMPv6TypeCode(layers.ICMPv6TypeEchoRequest, 0),
+			}
+			_ = icmp.SetNetworkLayerForChecksum(ip6)
+			err = gopacket.SerializeLayers(buf, opts, ip6, icmp)
+		default:
+			err = gopacket.SerializeLayers(buf, opts, ip6)
+		}
+	} else {
+		ip4 := &layers.IPv4{
+			Version: 4,
+			TTL:     64,
+			SrcIP:   src,
+			DstIP:   dst,
 		}
-		err = gopacket.SerializeLayers(buf, opts, ipLayer, icmp)
 
-	default:
-		err = gopacket.SerializeLayers(buf, opts, ipLayer)
+		switch proto {
+		case fw.ProtocolTCP:
+			ip4.Protocol = layers.IPProtocolTCP
+			tcp := &layers.TCP{SrcPort: layers.TCPPort(srcPort), DstPort: layers.TCPPort(dstPort)}
+			_ = tcp.SetNetworkLayerForChecksum(ip4)
+			err = gopacket.SerializeLayers(buf, opts, ip4, tcp)
+		case fw.ProtocolUDP:
+			ip4.Protocol = layers.IPProtocolUDP
+			udp := &layers.UDP{SrcPort: layers.UDPPort(srcPort), DstPort: layers.UDPPort(dstPort)}
+			_ = udp.SetNetworkLayerForChecksum(ip4)
+			err = gopacket.SerializeLayers(buf, opts, ip4, udp)
+		case fw.ProtocolICMP:
+			ip4.Protocol = layers.IPProtocolICMPv4
+			icmp := &layers.ICMPv4{TypeCode: layers.CreateICMPv4TypeCode(layers.ICMPv4TypeEchoRequest, 0)}
+			err = gopacket.SerializeLayers(buf, opts, ip4, icmp)
+		default:
+			err = gopacket.SerializeLayers(buf, opts, ip4)
+		}
 	}
 
 	require.NoError(t, err)
@@ -1498,3 +1681,103 @@ func TestRouteACLSet(t *testing.T) {
 	_, isAllowed = manager.routeACLsPass(srcIP, dstIP, protoToLayer(fw.ProtocolTCP, layers.LayerTypeIPv4), 12345, 80)
 	require.True(t, isAllowed, "After set update, traffic to the added network should be allowed")
 }
+
+// TestRouteACLFilteringIPv6 tests IPv6 route ACL matching directly via routeACLsPass.
+// Note: full FilterInbound for routed IPv6 traffic drops at the forwarder stage (IPv4-only)
+// but the ACL decision itself is correct.
+func TestRouteACLFilteringIPv6(t *testing.T) {
+	manager := setupRoutedManager(t, "10.10.0.100/16")
+
+	v6Dst := netip.MustParsePrefix("fd00:dead:beef::/48")
+	_, err := manager.AddRouteFiltering(
+		nil,
+		[]netip.Prefix{netip.MustParsePrefix("fd00::/16")},
+		fw.Network{Prefix: v6Dst},
+		fw.ProtocolTCP,
+		nil,
+		&fw.Port{Values: []uint16{80}},
+		fw.ActionAccept,
+	)
+	require.NoError(t, err)
+
+	_, err = manager.AddRouteFiltering(
+		nil,
+		[]netip.Prefix{netip.MustParsePrefix("fd00::/16")},
+		fw.Network{Prefix: netip.MustParsePrefix("fd00:dead:beef:1::/64")},
+		fw.ProtocolALL,
+		nil,
+		nil,
+		fw.ActionDrop,
+	)
+	require.NoError(t, err)
+
+	tests := []struct {
+		name    string
+		srcIP   netip.Addr
+		dstIP   netip.Addr
+		proto   gopacket.LayerType
+		srcPort uint16
+		dstPort uint16
+		allowed bool
+	}{
+		{
+			name:    "IPv6 TCP to allowed dest",
+			srcIP:   netip.MustParseAddr("fd00::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef::80"),
+			proto:   layers.LayerTypeTCP,
+			srcPort: 12345,
+			dstPort: 80,
+			allowed: true,
+		},
+		{
+			name:    "IPv6 TCP wrong port",
+			srcIP:   netip.MustParseAddr("fd00::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef::80"),
+			proto:   layers.LayerTypeTCP,
+			srcPort: 12345,
+			dstPort: 443,
+			allowed: false,
+		},
+		{
+			name:    "IPv6 UDP not matched by TCP rule",
+			srcIP:   netip.MustParseAddr("fd00::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef::80"),
+			proto:   layers.LayerTypeUDP,
+			srcPort: 12345,
+			dstPort: 80,
+			allowed: false,
+		},
+		{
+			name:    "IPv6 ICMPv6 matches ICMP rule via protoLayerMatches",
+			srcIP:   netip.MustParseAddr("fd00::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef::80"),
+			proto:   layers.LayerTypeICMPv6,
+			allowed: false,
+		},
+		{
+			name:    "IPv6 to denied subnet",
+			srcIP:   netip.MustParseAddr("fd00::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef:1::1"),
+			proto:   layers.LayerTypeTCP,
+			srcPort: 12345,
+			dstPort: 80,
+			allowed: false,
+		},
+		{
+			name:    "IPv6 source outside allowed range",
+			srcIP:   netip.MustParseAddr("fe80::1"),
+			dstIP:   netip.MustParseAddr("fd00:dead:beef::80"),
+			proto:   layers.LayerTypeTCP,
+			srcPort: 12345,
+			dstPort: 80,
+			allowed: false,
+		},
+	}
+
+	for _, tc := range tests {
+		t.Run(tc.name, func(t *testing.T) {
+			_, pass := manager.routeACLsPass(tc.srcIP, tc.dstIP, tc.proto, tc.srcPort, tc.dstPort)
+			require.Equal(t, tc.allowed, pass, "route ACL result mismatch")
+		})
+	}
+}

client/firewall/uspfilter/filter_test.go

@@ -12,6 +12,7 @@ import (
 	"github.com/google/gopacket"
 	"github.com/google/gopacket/layers"
 	"github.com/sirupsen/logrus"
+	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 	wgdevice "golang.zx2c4.com/wireguard/device"
 
@@ -186,81 +187,52 @@ func TestManagerDeleteRule(t *testing.T) {
 	}
 }
 
-func TestAddUDPPacketHook(t *testing.T) {
-	tests := []struct {
-		name       string
-		in         bool
-		expDir     fw.RuleDirection
-		ip         netip.Addr
-		dPort      uint16
-		hook       func([]byte) bool
-		expectedID string
-	}{
-		{
-			name:   "Test Outgoing UDP Packet Hook",
-			in:     false,
-			expDir: fw.RuleDirectionOUT,
-			ip:     netip.MustParseAddr("10.168.0.1"),
-			dPort:  8000,
-			hook:   func([]byte) bool { return true },
-		},
-		{
-			name:   "Test Incoming UDP Packet Hook",
-			in:     true,
-			expDir: fw.RuleDirectionIN,
-			ip:     netip.MustParseAddr("::1"),
-			dPort:  9000,
-			hook:   func([]byte) bool { return false },
-		},
-	}
+func TestSetUDPPacketHook(t *testing.T) {
+	manager, err := Create(&IFaceMock{
+		SetFilterFunc: func(device.PacketFilter) error { return nil },
+	}, false, flowLogger, nbiface.DefaultMTU)
+	require.NoError(t, err)
+	t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
 
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			manager, err := Create(&IFaceMock{
-				SetFilterFunc: func(device.PacketFilter) error { return nil },
-			}, false, flowLogger, nbiface.DefaultMTU)
-			require.NoError(t, err)
+	var called bool
+	manager.SetUDPPacketHook(netip.MustParseAddr("10.168.0.1"), 8000, func([]byte) bool {
+		called = true
+		return true
+	})
 
-			manager.AddUDPPacketHook(tt.in, tt.ip, tt.dPort, tt.hook)
+	h := manager.udpHookOut.Load()
+	require.NotNil(t, h)
+	assert.Equal(t, netip.MustParseAddr("10.168.0.1"), h.ip)
+	assert.Equal(t, uint16(8000), h.port)
+	assert.True(t, h.fn(nil))
+	assert.True(t, called)
 
-			var addedRule PeerRule
-			if tt.in {
-				// Incoming UDP hooks are stored in allow rules map
-				if len(manager.incomingRules[tt.ip]) != 1 {
-					t.Errorf("expected 1 incoming rule, got %d", len(manager.incomingRules[tt.ip]))
-					return
-				}
-				for _, rule := range manager.incomingRules[tt.ip] {
-					addedRule = rule
-				}
-			} else {
-				if len(manager.outgoingRules[tt.ip]) != 1 {
-					t.Errorf("expected 1 outgoing rule, got %d", len(manager.outgoingRules[tt.ip]))
-					return
-				}
-				for _, rule := range manager.outgoingRules[tt.ip] {
-					addedRule = rule
-				}
-			}
+	manager.SetUDPPacketHook(netip.MustParseAddr("10.168.0.1"), 8000, nil)
+	assert.Nil(t, manager.udpHookOut.Load())
+}
 
-			if tt.ip.Compare(addedRule.ip) != 0 {
-				t.Errorf("expected ip %s, got %s", tt.ip, addedRule.ip)
-				return
-			}
-			if tt.dPort != addedRule.dPort.Values[0] {
-				t.Errorf("expected dPort %d, got %d", tt.dPort, addedRule.dPort.Values[0])
-				return
-			}
-			if layers.LayerTypeUDP != addedRule.protoLayer {
-				t.Errorf("expected protoLayer %s, got %s", layers.LayerTypeUDP, addedRule.protoLayer)
-				return
-			}
-			if addedRule.udpHook == nil {
-				t.Errorf("expected udpHook to be set")
-				return
-			}
-		})
-	}
+func TestSetTCPPacketHook(t *testing.T) {
+	manager, err := Create(&IFaceMock{
+		SetFilterFunc: func(device.PacketFilter) error { return nil },
+	}, false, flowLogger, nbiface.DefaultMTU)
+	require.NoError(t, err)
+	t.Cleanup(func() { require.NoError(t, manager.Close(nil)) })
+
+	var called bool
+	manager.SetTCPPacketHook(netip.MustParseAddr("10.168.0.1"), 53, func([]byte) bool {
+		called = true
+		return true
+	})
+
+	h := manager.tcpHookOut.Load()
+	require.NotNil(t, h)
+	assert.Equal(t, netip.MustParseAddr("10.168.0.1"), h.ip)
+	assert.Equal(t, uint16(53), h.port)
+	assert.True(t, h.fn(nil))
+	assert.True(t, called)
+
+	manager.SetTCPPacketHook(netip.MustParseAddr("10.168.0.1"), 53, nil)
+	assert.Nil(t, manager.tcpHookOut.Load())
 }
 
 // TestPeerRuleLifecycleDenyRules verifies that deny rules are correctly added
@@ -530,39 +502,12 @@ func TestRemovePacketHook(t *testing.T) {
 		require.NoError(t, manager.Close(nil))
 	}()
 
-	// Add a UDP packet hook
-	hookFunc := func(data []byte) bool { return true }
-	hookID := manager.AddUDPPacketHook(false, netip.MustParseAddr("192.168.0.1"), 8080, hookFunc)
+	manager.SetUDPPacketHook(netip.MustParseAddr("192.168.0.1"), 8080, func([]byte) bool { return true })
 
-	// Assert the hook is added by finding it in the manager's outgoing rules
-	found := false
-	for _, arr := range manager.outgoingRules {
-		for _, rule := range arr {
-			if rule.id == hookID {
-				found = true
-				break
-			}
-		}
-	}
+	require.NotNil(t, manager.udpHookOut.Load(), "hook should be registered")
 
-	if !found {
-		t.Fatalf("The hook was not added properly.")
-	}
-
-	// Now remove the packet hook
-	err = manager.RemovePacketHook(hookID)
-	if err != nil {
-		t.Fatalf("Failed to remove hook: %s", err)
-	}
-
-	// Assert the hook is removed by checking it in the manager's outgoing rules
-	for _, arr := range manager.outgoingRules {
-		for _, rule := range arr {
-			if rule.id == hookID {
-				t.Fatalf("The hook was not removed properly.")
-			}
-		}
-	}
+	manager.SetUDPPacketHook(netip.MustParseAddr("192.168.0.1"), 8080, nil)
+	assert.Nil(t, manager.udpHookOut.Load(), "hook should be removed")
 }
 
 func TestProcessOutgoingHooks(t *testing.T) {
@@ -582,18 +527,22 @@ func TestProcessOutgoingHooks(t *testing.T) {
 			d := &decoder{
 				decoded: []gopacket.LayerType{},
 			}
-			d.parser = gopacket.NewDecodingLayerParser(
+			d.parser4 = gopacket.NewDecodingLayerParser(
 				layers.LayerTypeIPv4,
 				&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 			)
-			d.parser.IgnoreUnsupported = true
+			d.parser4.IgnoreUnsupported = true
+			d.parser6 = gopacket.NewDecodingLayerParser(
+				layers.LayerTypeIPv6,
+				&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+			)
+			d.parser6.IgnoreUnsupported = true
 			return d
 		},
 	}
 
 	hookCalled := false
-	hookID := manager.AddUDPPacketHook(
-		false,
+	manager.SetUDPPacketHook(
 		netip.MustParseAddr("100.10.0.100"),
 		53,
 		func([]byte) bool {
@@ -601,7 +550,6 @@ func TestProcessOutgoingHooks(t *testing.T) {
 			return true
 		},
 	)
-	require.NotEmpty(t, hookID)
 
 	// Create test UDP packet
 	ipv4 := &layers.IPv4{
@@ -687,11 +635,16 @@ func TestStatefulFirewall_UDPTracking(t *testing.T) {
 			d := &decoder{
 				decoded: []gopacket.LayerType{},
 			}
-			d.parser = gopacket.NewDecodingLayerParser(
+			d.parser4 = gopacket.NewDecodingLayerParser(
 				layers.LayerTypeIPv4,
 				&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 			)
-			d.parser.IgnoreUnsupported = true
+			d.parser4.IgnoreUnsupported = true
+			d.parser6 = gopacket.NewDecodingLayerParser(
+				layers.LayerTypeIPv6,
+				&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+			)
+			d.parser6.IgnoreUnsupported = true
 			return d
 		},
 	}
@@ -1097,8 +1050,8 @@ func TestMSSClamping(t *testing.T) {
 	}()
 
 	require.True(t, manager.mssClampEnabled, "MSS clamping should be enabled by default")
-	expectedMSSValue := uint16(1280 - ipTCPHeaderMinSize)
-	require.Equal(t, expectedMSSValue, manager.mssClampValue, "MSS clamp value should be MTU - 40")
+	require.Equal(t, uint16(1280-ipv4TCPHeaderMinSize), manager.mssClampValueIPv4, "IPv4 MSS clamp value should be MTU - 40")
+	require.Equal(t, uint16(1280-ipv6TCPHeaderMinSize), manager.mssClampValueIPv6, "IPv6 MSS clamp value should be MTU - 60")
 
 	err = manager.UpdateLocalIPs()
 	require.NoError(t, err)
@@ -1116,7 +1069,7 @@ func TestMSSClamping(t *testing.T) {
 		require.Len(t, d.tcp.Options, 1, "Should have MSS option")
 		require.Equal(t, uint8(layers.TCPOptionKindMSS), uint8(d.tcp.Options[0].OptionType))
 		actualMSS := binary.BigEndian.Uint16(d.tcp.Options[0].OptionData)
-		require.Equal(t, expectedMSSValue, actualMSS, "MSS should be clamped to MTU - 40")
+		require.Equal(t, manager.mssClampValueIPv4, actualMSS, "MSS should be clamped to MTU - 40")
 	})
 
 	t.Run("SYN packet with low MSS unchanged", func(t *testing.T) {
@@ -1140,7 +1093,7 @@ func TestMSSClamping(t *testing.T) {
 		d := parsePacket(t, packet)
 		require.Len(t, d.tcp.Options, 1, "Should have MSS option")
 		actualMSS := binary.BigEndian.Uint16(d.tcp.Options[0].OptionData)
-		require.Equal(t, expectedMSSValue, actualMSS, "MSS in SYN-ACK should be clamped")
+		require.Equal(t, manager.mssClampValueIPv4, actualMSS, "MSS in SYN-ACK should be clamped")
 	})
 
 	t.Run("Non-SYN packet unchanged", func(t *testing.T) {
@@ -1312,13 +1265,18 @@ func TestShouldForward(t *testing.T) {
 		d := &decoder{
 			decoded: []gopacket.LayerType{},
 		}
-		d.parser = gopacket.NewDecodingLayerParser(
+		d.parser4 = gopacket.NewDecodingLayerParser(
 			layers.LayerTypeIPv4,
 			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 		)
-		d.parser.IgnoreUnsupported = true
+		d.parser4.IgnoreUnsupported = true
+		d.parser6 = gopacket.NewDecodingLayerParser(
+			layers.LayerTypeIPv6,
+			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+		)
+		d.parser6.IgnoreUnsupported = true
 
-		err = d.parser.DecodeLayers(buf.Bytes(), &d.decoded)
+		err = d.decodePacket(buf.Bytes())
 		require.NoError(t, err)
 
 		return d
@@ -1378,6 +1336,44 @@ func TestShouldForward(t *testing.T) {
 		},
 	}
 
+	// Add IPv6 to the interface and test dual-stack cases
+	wgIPv6 := netip.MustParseAddr("fd00::1")
+	otherIPv6 := netip.MustParseAddr("fd00::2")
+	ifaceMock.AddressFunc = func() wgaddr.Address {
+		return wgaddr.Address{
+			IP:      wgIP,
+			Network: netip.PrefixFrom(wgIP, 24),
+			IPv6:    wgIPv6,
+			IPv6Net: netip.PrefixFrom(wgIPv6, 64),
+		}
+	}
+
+	// Re-create manager to pick up the new address with IPv6
+	require.NoError(t, manager.Close(nil))
+	manager, err = Create(ifaceMock, false, flowLogger, nbiface.DefaultMTU)
+	require.NoError(t, err)
+
+	v6Cases := []struct {
+		name        string
+		dstIP       netip.Addr
+		expected    bool
+		description string
+	}{
+		{"v6 traffic to other address", otherIPv6, true, "should forward v6 traffic not destined to our v6 address"},
+		{"v6 traffic to our v6 IP", wgIPv6, false, "should not forward traffic destined to our v6 address"},
+		{"v4 traffic to other with v6 configured", otherIP, true, "should forward v4 traffic when v6 configured"},
+		{"v4 traffic to our v4 IP with v6 configured", wgIP, false, "should not forward traffic to our v4 address"},
+	}
+	for _, tt := range v6Cases {
+		t.Run(tt.name, func(t *testing.T) {
+			manager.localForwarding = true
+			manager.netstack = false
+			decoder := createTCPDecoder(8080)
+			result := manager.shouldForward(decoder, tt.dstIP)
+			require.Equal(t, tt.expected, result, tt.description)
+		})
+	}
+
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
 			// Configure manager

client/firewall/uspfilter/forwarder/endpoint.go

@@ -1,7 +1,8 @@
 package forwarder
 
 import (
-	"fmt"
+	"net"
+	"strconv"
 	"sync/atomic"
 
 	wgdevice "golang.zx2c4.com/wireguard/device"
@@ -47,17 +48,23 @@ func (e *endpoint) LinkAddress() tcpip.LinkAddress {
 func (e *endpoint) WritePackets(pkts stack.PacketBufferList) (int, tcpip.Error) {
 	var written int
 	for _, pkt := range pkts.AsSlice() {
-		netHeader := header.IPv4(pkt.NetworkHeader().View().AsSlice())
-
 		data := stack.PayloadSince(pkt.NetworkHeader())
 		if data == nil {
 			continue
 		}
 
-		// Send the packet through WireGuard
-		address := netHeader.DestinationAddress()
-		err := e.device.CreateOutboundPacket(data.AsSlice(), address.AsSlice())
-		if err != nil {
+		raw := pkt.NetworkHeader().View().AsSlice()
+		if len(raw) == 0 {
+			continue
+		}
+		var address tcpip.Address
+		if raw[0]>>4 == 6 {
+			address = header.IPv6(raw).DestinationAddress()
+		} else {
+			address = header.IPv4(raw).DestinationAddress()
+		}
+
+		if err := e.device.CreateOutboundPacket(data.AsSlice(), address.AsSlice()); err != nil {
 			e.logger.Error1("CreateOutboundPacket: %v", err)
 			continue
 		}
@@ -103,5 +110,7 @@ type epID stack.TransportEndpointID
 
 func (i epID) String() string {
 	// src and remote is swapped
-	return fmt.Sprintf("%s:%d → %s:%d", i.RemoteAddress, i.RemotePort, i.LocalAddress, i.LocalPort)
+	return net.JoinHostPort(i.RemoteAddress.String(), strconv.Itoa(int(i.RemotePort))) +
+		" → " +
+		net.JoinHostPort(i.LocalAddress.String(), strconv.Itoa(int(i.LocalPort)))
 }

client/firewall/uspfilter/forwarder/forwarder.go

@@ -14,6 +14,7 @@ import (
 	"gvisor.dev/gvisor/pkg/tcpip"
 	"gvisor.dev/gvisor/pkg/tcpip/header"
 	"gvisor.dev/gvisor/pkg/tcpip/network/ipv4"
+	"gvisor.dev/gvisor/pkg/tcpip/network/ipv6"
 	"gvisor.dev/gvisor/pkg/tcpip/stack"
 	"gvisor.dev/gvisor/pkg/tcpip/transport/icmp"
 	"gvisor.dev/gvisor/pkg/tcpip/transport/tcp"
@@ -36,25 +37,31 @@ type Forwarder struct {
 	logger     *nblog.Logger
 	flowLogger nftypes.FlowLogger
 	// ruleIdMap is used to store the rule ID for a given connection
-	ruleIdMap        sync.Map
-	stack            *stack.Stack
-	endpoint         *endpoint
-	udpForwarder     *udpForwarder
-	ctx              context.Context
-	cancel           context.CancelFunc
-	ip               tcpip.Address
-	netstack         bool
-	hasRawICMPAccess bool
-	pingSemaphore    chan struct{}
+	ruleIdMap          sync.Map
+	stack              *stack.Stack
+	endpoint           *endpoint
+	udpForwarder       *udpForwarder
+	ctx                context.Context
+	cancel             context.CancelFunc
+	ip                 tcpip.Address
+	ipv6               tcpip.Address
+	netstack           bool
+	hasRawICMPAccess   bool
+	hasRawICMPv6Access bool
+	pingSemaphore      chan struct{}
 }
 
 func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.FlowLogger, netstack bool, mtu uint16) (*Forwarder, error) {
 	s := stack.New(stack.Options{
-		NetworkProtocols: []stack.NetworkProtocolFactory{ipv4.NewProtocol},
+		NetworkProtocols: []stack.NetworkProtocolFactory{
+			ipv4.NewProtocol,
+			ipv6.NewProtocol,
+		},
 		TransportProtocols: []stack.TransportProtocolFactory{
 			tcp.NewProtocol,
 			udp.NewProtocol,
 			icmp.NewProtocol4,
+			icmp.NewProtocol6,
 		},
 		HandleLocal: false,
 	})
@@ -73,7 +80,7 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 	protoAddr := tcpip.ProtocolAddress{
 		Protocol: ipv4.ProtocolNumber,
 		AddressWithPrefix: tcpip.AddressWithPrefix{
-			Address:   tcpip.AddrFromSlice(iface.Address().IP.AsSlice()),
+			Address:   tcpip.AddrFrom4(iface.Address().IP.As4()),
 			PrefixLen: iface.Address().Network.Bits(),
 		},
 	}
@@ -82,6 +89,19 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 		return nil, fmt.Errorf("failed to add protocol address: %s", err)
 	}
 
+	if v6 := iface.Address().IPv6; v6.IsValid() {
+		v6Addr := tcpip.ProtocolAddress{
+			Protocol: ipv6.ProtocolNumber,
+			AddressWithPrefix: tcpip.AddressWithPrefix{
+				Address:   tcpip.AddrFrom16(v6.As16()),
+				PrefixLen: iface.Address().IPv6Net.Bits(),
+			},
+		}
+		if err := s.AddProtocolAddress(nicID, v6Addr, stack.AddressProperties{}); err != nil {
+			return nil, fmt.Errorf("add IPv6 protocol address: %s", err)
+		}
+	}
+
 	defaultSubnet, err := tcpip.NewSubnet(
 		tcpip.AddrFrom4([4]byte{0, 0, 0, 0}),
 		tcpip.MaskFromBytes([]byte{0, 0, 0, 0}),
@@ -90,6 +110,14 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 		return nil, fmt.Errorf("creating default subnet: %w", err)
 	}
 
+	defaultSubnetV6, err := tcpip.NewSubnet(
+		tcpip.AddrFrom16([16]byte{}),
+		tcpip.MaskFromBytes(make([]byte, 16)),
+	)
+	if err != nil {
+		return nil, fmt.Errorf("creating default v6 subnet: %w", err)
+	}
+
 	if err := s.SetPromiscuousMode(nicID, true); err != nil {
 		return nil, fmt.Errorf("set promiscuous mode: %s", err)
 	}
@@ -98,10 +126,8 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 	}
 
 	s.SetRouteTable([]tcpip.Route{
-		{
-			Destination: defaultSubnet,
-			NIC:         nicID,
-		},
+		{Destination: defaultSubnet, NIC: nicID},
+		{Destination: defaultSubnetV6, NIC: nicID},
 	})
 
 	ctx, cancel := context.WithCancel(context.Background())
@@ -114,7 +140,8 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 		ctx:           ctx,
 		cancel:        cancel,
 		netstack:      netstack,
-		ip:            tcpip.AddrFromSlice(iface.Address().IP.AsSlice()),
+		ip:            tcpip.AddrFrom4(iface.Address().IP.As4()),
+		ipv6:          addrFromNetipAddr(iface.Address().IPv6),
 		pingSemaphore: make(chan struct{}, 3),
 	}
 
@@ -131,7 +158,10 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 	udpForwarder := udp.NewForwarder(s, f.handleUDP)
 	s.SetTransportProtocolHandler(udp.ProtocolNumber, udpForwarder.HandlePacket)
 
-	s.SetTransportProtocolHandler(icmp.ProtocolNumber4, f.handleICMP)
+	// ICMP is handled directly in InjectIncomingPacket, bypassing gVisor's
+	// network layer. This avoids duplicate echo replies (v4) and the v6
+	// auto-reply bug where gVisor responds at the network layer before
+	// our transport handler fires.
 
 	f.checkICMPCapability()
 
@@ -140,8 +170,30 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 }
 
 func (f *Forwarder) InjectIncomingPacket(payload []byte) error {
-	if len(payload) < header.IPv4MinimumSize {
-		return fmt.Errorf("packet too small: %d bytes", len(payload))
+	if len(payload) == 0 {
+		return fmt.Errorf("empty packet")
+	}
+
+	var protoNum tcpip.NetworkProtocolNumber
+	switch payload[0] >> 4 {
+	case 4:
+		if len(payload) < header.IPv4MinimumSize {
+			return fmt.Errorf("IPv4 packet too small: %d bytes", len(payload))
+		}
+		if f.handleICMPDirect(payload) {
+			return nil
+		}
+		protoNum = ipv4.ProtocolNumber
+	case 6:
+		if len(payload) < header.IPv6MinimumSize {
+			return fmt.Errorf("IPv6 packet too small: %d bytes", len(payload))
+		}
+		if f.handleICMPDirect(payload) {
+			return nil
+		}
+		protoNum = ipv6.ProtocolNumber
+	default:
+		return fmt.Errorf("unknown IP version: %d", payload[0]>>4)
 	}
 
 	pkt := stack.NewPacketBuffer(stack.PacketBufferOptions{
@@ -150,11 +202,95 @@ func (f *Forwarder) InjectIncomingPacket(payload []byte) error {
 	defer pkt.DecRef()
 
 	if f.endpoint.dispatcher != nil {
-		f.endpoint.dispatcher.DeliverNetworkPacket(ipv4.ProtocolNumber, pkt)
+		f.endpoint.dispatcher.DeliverNetworkPacket(protoNum, pkt)
 	}
 	return nil
 }
 
+// handleICMPDirect intercepts ICMP packets from raw IP payloads before they
+// enter gVisor. It synthesizes the TransportEndpointID and PacketBuffer that
+// the existing handlers expect, then dispatches to handleICMP/handleICMPv6.
+// This bypasses gVisor's network layer which causes duplicate v4 echo replies
+// and auto-replies to all v6 echo requests in promiscuous mode.
+//
+// Unlike gVisor's network layer, this does not validate ICMP checksums or
+// reassemble IP fragments. Fragmented ICMP packets fall through to gVisor.
+func parseICMPv4(payload []byte) (ipHdrLen int, src, dst tcpip.Address, ok bool) {
+	ip := header.IPv4(payload)
+	if ip.Protocol() != uint8(header.ICMPv4ProtocolNumber) {
+		return 0, src, dst, false
+	}
+	if ip.FragmentOffset() != 0 || ip.Flags()&header.IPv4FlagMoreFragments != 0 {
+		return 0, src, dst, false
+	}
+	ipHdrLen = int(ip.HeaderLength())
+	if len(payload)-ipHdrLen < header.ICMPv4MinimumSize {
+		return 0, src, dst, false
+	}
+	return ipHdrLen, ip.SourceAddress(), ip.DestinationAddress(), true
+}
+
+func parseICMPv6(payload []byte) (ipHdrLen int, src, dst tcpip.Address, ok bool) {
+	ip := header.IPv6(payload)
+	if ip.NextHeader() != uint8(header.ICMPv6ProtocolNumber) {
+		return 0, src, dst, false
+	}
+	ipHdrLen = header.IPv6MinimumSize
+	if len(payload)-ipHdrLen < header.ICMPv6MinimumSize {
+		return 0, src, dst, false
+	}
+	return ipHdrLen, ip.SourceAddress(), ip.DestinationAddress(), true
+}
+
+func (f *Forwarder) handleICMPDirect(payload []byte) bool {
+	var (
+		ipHdrLen int
+		srcAddr  tcpip.Address
+		dstAddr  tcpip.Address
+		ok       bool
+	)
+	switch payload[0] >> 4 {
+	case 4:
+		ipHdrLen, srcAddr, dstAddr, ok = parseICMPv4(payload)
+	case 6:
+		ipHdrLen, srcAddr, dstAddr, ok = parseICMPv6(payload)
+	}
+	if !ok {
+		return false
+	}
+
+	// Let gVisor handle ICMP destined for our own addresses natively.
+	// Its network-layer auto-reply is correct and efficient for local traffic.
+	if f.ip.Equal(dstAddr) || f.ipv6.Equal(dstAddr) {
+		return false
+	}
+
+	id := stack.TransportEndpointID{
+		LocalAddress:  dstAddr,
+		RemoteAddress: srcAddr,
+	}
+
+	// Build a PacketBuffer with headers consumed the same way gVisor would.
+	pkt := stack.NewPacketBuffer(stack.PacketBufferOptions{
+		Payload: buffer.MakeWithData(payload),
+	})
+	defer pkt.DecRef()
+
+	if _, ok := pkt.NetworkHeader().Consume(ipHdrLen); !ok {
+		return false
+	}
+
+	icmpPayload := payload[ipHdrLen:]
+	if _, ok := pkt.TransportHeader().Consume(len(icmpPayload)); !ok {
+		return false
+	}
+
+	if payload[0]>>4 == 6 {
+		return f.handleICMPv6(id, pkt)
+	}
+	return f.handleICMP(id, pkt)
+}
+
 // Stop gracefully shuts down the forwarder
 func (f *Forwarder) Stop() {
 	f.cancel()
@@ -167,11 +303,14 @@ func (f *Forwarder) Stop() {
 	f.stack.Wait()
 }
 
-func (f *Forwarder) determineDialAddr(addr tcpip.Address) net.IP {
+func (f *Forwarder) determineDialAddr(addr tcpip.Address) netip.Addr {
 	if f.netstack && f.ip.Equal(addr) {
-		return net.IPv4(127, 0, 0, 1)
+		return netip.AddrFrom4([4]byte{127, 0, 0, 1})
 	}
-	return addr.AsSlice()
+	if f.netstack && f.ipv6.Equal(addr) {
+		return netip.IPv6Loopback()
+	}
+	return addrToNetipAddr(addr)
 }
 
 func (f *Forwarder) RegisterRuleID(srcIP, dstIP netip.Addr, srcPort, dstPort uint16, ruleID []byte) {
@@ -205,23 +344,50 @@ func buildKey(srcIP, dstIP netip.Addr, srcPort, dstPort uint16) conntrack.ConnKe
 	}
 }
 
+// addrFromNetipAddr converts a netip.Addr to a gvisor tcpip.Address without allocating.
+func addrFromNetipAddr(addr netip.Addr) tcpip.Address {
+	if !addr.IsValid() {
+		return tcpip.Address{}
+	}
+	if addr.Is4() {
+		return tcpip.AddrFrom4(addr.As4())
+	}
+	return tcpip.AddrFrom16(addr.As16())
+}
+
+// addrToNetipAddr converts a gvisor tcpip.Address to netip.Addr without allocating.
+func addrToNetipAddr(addr tcpip.Address) netip.Addr {
+	switch addr.Len() {
+	case 4:
+		return netip.AddrFrom4(addr.As4())
+	case 16:
+		return netip.AddrFrom16(addr.As16())
+	default:
+		return netip.Addr{}
+	}
+}
+
 // checkICMPCapability tests whether we have raw ICMP socket access at startup.
 func (f *Forwarder) checkICMPCapability() {
+	f.hasRawICMPAccess = probeRawICMP("ip4:icmp", "0.0.0.0", f.logger)
+	f.hasRawICMPv6Access = probeRawICMP("ip6:ipv6-icmp", "::", f.logger)
+}
+
+func probeRawICMP(network, addr string, logger *nblog.Logger) bool {
 	ctx, cancel := context.WithTimeout(context.Background(), 100*time.Millisecond)
 	defer cancel()
 
 	lc := net.ListenConfig{}
-	conn, err := lc.ListenPacket(ctx, "ip4:icmp", "0.0.0.0")
+	conn, err := lc.ListenPacket(ctx, network, addr)
 	if err != nil {
-		f.hasRawICMPAccess = false
-		f.logger.Debug("forwarder: No raw ICMP socket access, will use ping binary fallback")
-		return
+		logger.Debug1("forwarder: no raw %s socket access, will use ping binary fallback", network)
+		return false
 	}
 
 	if err := conn.Close(); err != nil {
-		f.logger.Debug1("forwarder: Failed to close ICMP capability test socket: %v", err)
+		logger.Debug2("forwarder: failed to close %s capability test socket: %v", network, err)
 	}
 
-	f.hasRawICMPAccess = true
-	f.logger.Debug("forwarder: Raw ICMP socket access available")
+	logger.Debug1("forwarder: raw %s socket access available", network)
+	return true
 }

client/firewall/uspfilter/forwarder/icmp.go

@@ -35,7 +35,7 @@ func (f *Forwarder) handleICMP(id stack.TransportEndpointID, pkt *stack.PacketBu
 	}
 
 	icmpData := stack.PayloadSince(pkt.TransportHeader()).AsSlice()
-	conn, err := f.forwardICMPPacket(id, icmpData, uint8(icmpHdr.Type()), uint8(icmpHdr.Code()), 100*time.Millisecond)
+	conn, err := f.forwardICMPPacket(id, icmpData, uint8(icmpHdr.Type()), uint8(icmpHdr.Code()), false, 100*time.Millisecond)
 	if err != nil {
 		f.logger.Error2("forwarder: Failed to forward ICMP packet for %v: %v", epID(id), err)
 		return true
@@ -58,7 +58,7 @@ func (f *Forwarder) handleICMPEcho(flowID uuid.UUID, id stack.TransportEndpointI
 			defer func() { <-f.pingSemaphore }()
 
 			if f.hasRawICMPAccess {
-				f.handleICMPViaSocket(flowID, id, icmpType, icmpCode, icmpData, rxBytes)
+				f.handleICMPViaSocket(flowID, id, icmpType, icmpCode, icmpData, rxBytes, false)
 			} else {
 				f.handleICMPViaPing(flowID, id, icmpType, icmpCode, icmpData, rxBytes)
 			}
@@ -72,18 +72,23 @@ func (f *Forwarder) handleICMPEcho(flowID uuid.UUID, id stack.TransportEndpointI
 
 // forwardICMPPacket creates a raw ICMP socket and sends the packet, returning the connection.
 // The caller is responsible for closing the returned connection.
-func (f *Forwarder) forwardICMPPacket(id stack.TransportEndpointID, payload []byte, icmpType, icmpCode uint8, timeout time.Duration) (net.PacketConn, error) {
+func (f *Forwarder) forwardICMPPacket(id stack.TransportEndpointID, payload []byte, icmpType, icmpCode uint8, v6 bool, timeout time.Duration) (net.PacketConn, error) {
 	ctx, cancel := context.WithTimeout(f.ctx, timeout)
 	defer cancel()
 
+	network, listenAddr := "ip4:icmp", "0.0.0.0"
+	if v6 {
+		network, listenAddr = "ip6:ipv6-icmp", "::"
+	}
+
 	lc := net.ListenConfig{}
-	conn, err := lc.ListenPacket(ctx, "ip4:icmp", "0.0.0.0")
+	conn, err := lc.ListenPacket(ctx, network, listenAddr)
 	if err != nil {
 		return nil, fmt.Errorf("create ICMP socket: %w", err)
 	}
 
 	dstIP := f.determineDialAddr(id.LocalAddress)
-	dst := &net.IPAddr{IP: dstIP}
+	dst := &net.IPAddr{IP: dstIP.AsSlice()}
 
 	if _, err = conn.WriteTo(payload, dst); err != nil {
 		if closeErr := conn.Close(); closeErr != nil {
@@ -98,11 +103,11 @@ func (f *Forwarder) forwardICMPPacket(id stack.TransportEndpointID, payload []by
 	return conn, nil
 }
 
-// handleICMPViaSocket handles ICMP echo requests using raw sockets.
-func (f *Forwarder) handleICMPViaSocket(flowID uuid.UUID, id stack.TransportEndpointID, icmpType, icmpCode uint8, icmpData []byte, rxBytes int) {
+// handleICMPViaSocket handles ICMP echo requests using raw sockets for both v4 and v6.
+func (f *Forwarder) handleICMPViaSocket(flowID uuid.UUID, id stack.TransportEndpointID, icmpType, icmpCode uint8, icmpData []byte, rxBytes int, v6 bool) {
 	sendTime := time.Now()
 
-	conn, err := f.forwardICMPPacket(id, icmpData, icmpType, icmpCode, 5*time.Second)
+	conn, err := f.forwardICMPPacket(id, icmpData, icmpType, icmpCode, v6, 5*time.Second)
 	if err != nil {
 		f.logger.Error2("forwarder: Failed to send ICMP packet for %v: %v", epID(id), err)
 		return
@@ -113,16 +118,20 @@ func (f *Forwarder) handleICMPViaSocket(flowID uuid.UUID, id stack.TransportEndp
 		}
 	}()
 
-	txBytes := f.handleEchoResponse(conn, id)
+	txBytes := f.handleEchoResponse(conn, id, v6)
 	rtt := time.Since(sendTime).Round(10 * time.Microsecond)
 
-	f.logger.Trace4("forwarder: Forwarded ICMP echo reply %v type %v code %v (rtt=%v, raw socket)",
-		epID(id), icmpType, icmpCode, rtt)
+	proto := "ICMP"
+	if v6 {
+		proto = "ICMPv6"
+	}
+	f.logger.Trace5("forwarder: Forwarded %s echo reply %v type %v code %v (rtt=%v, raw socket)",
+		proto, epID(id), icmpType, icmpCode, rtt)
 
 	f.sendICMPEvent(nftypes.TypeEnd, flowID, id, icmpType, icmpCode, uint64(rxBytes), uint64(txBytes))
 }
 
-func (f *Forwarder) handleEchoResponse(conn net.PacketConn, id stack.TransportEndpointID) int {
+func (f *Forwarder) handleEchoResponse(conn net.PacketConn, id stack.TransportEndpointID, v6 bool) int {
 	if err := conn.SetReadDeadline(time.Now().Add(5 * time.Second)); err != nil {
 		f.logger.Error1("forwarder: Failed to set read deadline for ICMP response: %v", err)
 		return 0
@@ -137,6 +146,19 @@ func (f *Forwarder) handleEchoResponse(conn net.PacketConn, id stack.TransportEn
 		return 0
 	}
 
+	if v6 {
+		// Recompute checksum: the raw socket response has a checksum computed
+		// over the real endpoint addresses, but we inject with overlay addresses.
+		icmpHdr := header.ICMPv6(response[:n])
+		icmpHdr.SetChecksum(0)
+		icmpHdr.SetChecksum(header.ICMPv6Checksum(header.ICMPv6ChecksumParams{
+			Header: icmpHdr,
+			Src:    id.LocalAddress,
+			Dst:    id.RemoteAddress,
+		}))
+		return f.injectICMPv6Reply(id, response[:n])
+	}
+
 	return f.injectICMPReply(id, response[:n])
 }
 
@@ -150,19 +172,23 @@ func (f *Forwarder) sendICMPEvent(typ nftypes.Type, flowID uuid.UUID, id stack.T
 		txPackets = 1
 	}
 
-	srcIp := netip.AddrFrom4(id.RemoteAddress.As4())
-	dstIp := netip.AddrFrom4(id.LocalAddress.As4())
+	srcIp := addrToNetipAddr(id.RemoteAddress)
+	dstIp := addrToNetipAddr(id.LocalAddress)
+
+	proto := nftypes.ICMP
+	if srcIp.Is6() {
+		proto = nftypes.ICMPv6
+	}
 
 	fields := nftypes.EventFields{
 		FlowID:    flowID,
 		Type:      typ,
 		Direction: nftypes.Ingress,
-		Protocol:  nftypes.ICMP,
-		// TODO: handle ipv6
-		SourceIP: srcIp,
-		DestIP:   dstIp,
-		ICMPType: icmpType,
-		ICMPCode: icmpCode,
+		Protocol:  proto,
+		SourceIP:  srcIp,
+		DestIP:    dstIp,
+		ICMPType:  icmpType,
+		ICMPCode:  icmpCode,
 
 		RxBytes:   rxBytes,
 		TxBytes:   txBytes,
@@ -209,26 +235,164 @@ func (f *Forwarder) handleICMPViaPing(flowID uuid.UUID, id stack.TransportEndpoi
 	f.sendICMPEvent(nftypes.TypeEnd, flowID, id, icmpType, icmpCode, uint64(rxBytes), uint64(txBytes))
 }
 
+// handleICMPv6 handles ICMPv6 packets from the network stack.
+func (f *Forwarder) handleICMPv6(id stack.TransportEndpointID, pkt *stack.PacketBuffer) bool {
+	icmpHdr := header.ICMPv6(pkt.TransportHeader().View().AsSlice())
+
+	flowID := uuid.New()
+	f.sendICMPEvent(nftypes.TypeStart, flowID, id, uint8(icmpHdr.Type()), uint8(icmpHdr.Code()), 0, 0)
+
+	if icmpHdr.Type() == header.ICMPv6EchoRequest {
+		return f.handleICMPv6Echo(flowID, id, pkt, uint8(icmpHdr.Type()), uint8(icmpHdr.Code()))
+	}
+
+	// For non-echo types (Destination Unreachable, Packet Too Big, etc), forward without waiting
+	if !f.hasRawICMPv6Access {
+		f.logger.Debug2("forwarder: Cannot handle ICMPv6 type %v without raw socket access for %v", icmpHdr.Type(), epID(id))
+		return false
+	}
+
+	icmpData := stack.PayloadSince(pkt.TransportHeader()).AsSlice()
+	conn, err := f.forwardICMPPacket(id, icmpData, uint8(icmpHdr.Type()), uint8(icmpHdr.Code()), true, 100*time.Millisecond)
+	if err != nil {
+		f.logger.Error2("forwarder: Failed to forward ICMPv6 packet for %v: %v", epID(id), err)
+		return true
+	}
+	if err := conn.Close(); err != nil {
+		f.logger.Debug1("forwarder: Failed to close ICMPv6 socket: %v", err)
+	}
+
+	return true
+}
+
+// handleICMPv6Echo handles ICMPv6 echo requests via raw socket or ping binary fallback.
+func (f *Forwarder) handleICMPv6Echo(flowID uuid.UUID, id stack.TransportEndpointID, pkt *stack.PacketBuffer, icmpType, icmpCode uint8) bool {
+	select {
+	case f.pingSemaphore <- struct{}{}:
+		icmpData := stack.PayloadSince(pkt.TransportHeader()).ToSlice()
+		rxBytes := pkt.Size()
+
+		go func() {
+			defer func() { <-f.pingSemaphore }()
+
+			if f.hasRawICMPv6Access {
+				f.handleICMPViaSocket(flowID, id, icmpType, icmpCode, icmpData, rxBytes, true)
+			} else {
+				f.handleICMPv6ViaPing(flowID, id, icmpType, icmpCode, icmpData, rxBytes)
+			}
+		}()
+	default:
+		f.logger.Debug3("forwarder: ICMPv6 rate limit exceeded for %v type %v code %v", epID(id), icmpType, icmpCode)
+	}
+	return true
+}
+
+// handleICMPv6ViaPing uses the system ping6 binary for ICMPv6 echo.
+func (f *Forwarder) handleICMPv6ViaPing(flowID uuid.UUID, id stack.TransportEndpointID, icmpType, icmpCode uint8, icmpData []byte, rxBytes int) {
+	ctx, cancel := context.WithTimeout(f.ctx, 5*time.Second)
+	defer cancel()
+
+	dstIP := f.determineDialAddr(id.LocalAddress)
+	cmd := buildPingCommand(ctx, dstIP, 5*time.Second)
+
+	pingStart := time.Now()
+	if err := cmd.Run(); err != nil {
+		f.logger.Warn4("forwarder: Ping6 failed for %v type %v code %v: %v", epID(id), icmpType, icmpCode, err)
+		return
+	}
+	rtt := time.Since(pingStart).Round(10 * time.Microsecond)
+
+	f.logger.Trace3("forwarder: Forwarded ICMPv6 echo request %v type %v code %v",
+		epID(id), icmpType, icmpCode)
+
+	txBytes := f.synthesizeICMPv6EchoReply(id, icmpData)
+
+	f.logger.Trace4("forwarder: Forwarded ICMPv6 echo reply %v type %v code %v (rtt=%v, ping binary)",
+		epID(id), icmpType, icmpCode, rtt)
+
+	f.sendICMPEvent(nftypes.TypeEnd, flowID, id, icmpType, icmpCode, uint64(rxBytes), uint64(txBytes))
+}
+
+// synthesizeICMPv6EchoReply creates an ICMPv6 echo reply and injects it back.
+func (f *Forwarder) synthesizeICMPv6EchoReply(id stack.TransportEndpointID, icmpData []byte) int {
+	replyICMP := make([]byte, len(icmpData))
+	copy(replyICMP, icmpData)
+
+	replyHdr := header.ICMPv6(replyICMP)
+	replyHdr.SetType(header.ICMPv6EchoReply)
+	replyHdr.SetChecksum(0)
+	// ICMPv6Checksum computes the pseudo-header internally from Src/Dst.
+	// Header contains the full ICMP message, so PayloadCsum/PayloadLen are zero.
+	replyHdr.SetChecksum(header.ICMPv6Checksum(header.ICMPv6ChecksumParams{
+		Header: replyHdr,
+		Src:    id.LocalAddress,
+		Dst:    id.RemoteAddress,
+	}))
+
+	return f.injectICMPv6Reply(id, replyICMP)
+}
+
+// injectICMPv6Reply wraps an ICMPv6 payload in an IPv6 header and sends to the peer.
+func (f *Forwarder) injectICMPv6Reply(id stack.TransportEndpointID, icmpPayload []byte) int {
+	ipHdr := make([]byte, header.IPv6MinimumSize)
+	ip := header.IPv6(ipHdr)
+	ip.Encode(&header.IPv6Fields{
+		PayloadLength:     uint16(len(icmpPayload)),
+		TransportProtocol: header.ICMPv6ProtocolNumber,
+		HopLimit:          64,
+		SrcAddr:           id.LocalAddress,
+		DstAddr:           id.RemoteAddress,
+	})
+
+	fullPacket := make([]byte, 0, len(ipHdr)+len(icmpPayload))
+	fullPacket = append(fullPacket, ipHdr...)
+	fullPacket = append(fullPacket, icmpPayload...)
+
+	if err := f.endpoint.device.CreateOutboundPacket(fullPacket, id.RemoteAddress.AsSlice()); err != nil {
+		f.logger.Error1("forwarder: Failed to send ICMPv6 reply to peer: %v", err)
+		return 0
+	}
+
+	return len(fullPacket)
+}
+
+const (
+	pingBin  = "ping"
+	ping6Bin = "ping6"
+)
+
 // buildPingCommand creates a platform-specific ping command.
-func buildPingCommand(ctx context.Context, target net.IP, timeout time.Duration) *exec.Cmd {
+// Most platforms auto-detect IPv6 from raw addresses. macOS/iOS/OpenBSD require ping6.
+func buildPingCommand(ctx context.Context, target netip.Addr, timeout time.Duration) *exec.Cmd {
 	timeoutSec := int(timeout.Seconds())
 	if timeoutSec < 1 {
 		timeoutSec = 1
 	}
 
+	isV6 := target.Is6()
+	timeoutStr := fmt.Sprintf("%d", timeoutSec)
+
 	switch runtime.GOOS {
 	case "linux", "android":
-		return exec.CommandContext(ctx, "ping", "-c", "1", "-W", fmt.Sprintf("%d", timeoutSec), "-q", target.String())
+		return exec.CommandContext(ctx, pingBin, "-c", "1", "-W", timeoutStr, "-q", target.String())
 	case "darwin", "ios":
-		return exec.CommandContext(ctx, "ping", "-c", "1", "-t", fmt.Sprintf("%d", timeoutSec), "-q", target.String())
+		bin := pingBin
+		if isV6 {
+			bin = ping6Bin
+		}
+		return exec.CommandContext(ctx, bin, "-c", "1", "-t", timeoutStr, "-q", target.String())
 	case "freebsd":
-		return exec.CommandContext(ctx, "ping", "-c", "1", "-t", fmt.Sprintf("%d", timeoutSec), target.String())
+		return exec.CommandContext(ctx, pingBin, "-c", "1", "-t", timeoutStr, target.String())
 	case "openbsd", "netbsd":
-		return exec.CommandContext(ctx, "ping", "-c", "1", "-w", fmt.Sprintf("%d", timeoutSec), target.String())
+		bin := pingBin
+		if isV6 {
+			bin = ping6Bin
+		}
+		return exec.CommandContext(ctx, bin, "-c", "1", "-w", timeoutStr, target.String())
 	case "windows":
-		return exec.CommandContext(ctx, "ping", "-n", "1", "-w", fmt.Sprintf("%d", timeoutSec*1000), target.String())
+		return exec.CommandContext(ctx, pingBin, "-n", "1", "-w", fmt.Sprintf("%d", timeoutSec*1000), target.String())
 	default:
-		return exec.CommandContext(ctx, "ping", "-c", "1", target.String())
+		return exec.CommandContext(ctx, pingBin, "-c", "1", target.String())
 	}
 }
 

client/firewall/uspfilter/forwarder/tcp.go

@@ -2,10 +2,9 @@ package forwarder
 
 import (
 	"context"
-	"fmt"
 	"io"
 	"net"
-	"net/netip"
+	"strconv"
 	"sync"
 
 	"github.com/google/uuid"
@@ -33,7 +32,7 @@ func (f *Forwarder) handleTCP(r *tcp.ForwarderRequest) {
 		}
 	}()
 
-	dialAddr := fmt.Sprintf("%s:%d", f.determineDialAddr(id.LocalAddress), id.LocalPort)
+	dialAddr := net.JoinHostPort(f.determineDialAddr(id.LocalAddress).String(), strconv.Itoa(int(id.LocalPort)))
 
 	outConn, err := (&net.Dialer{}).DialContext(f.ctx, "tcp", dialAddr)
 	if err != nil {
@@ -133,15 +132,14 @@ func (f *Forwarder) proxyTCP(id stack.TransportEndpointID, inConn *gonet.TCPConn
 }
 
 func (f *Forwarder) sendTCPEvent(typ nftypes.Type, flowID uuid.UUID, id stack.TransportEndpointID, rxBytes, txBytes, rxPackets, txPackets uint64) {
-	srcIp := netip.AddrFrom4(id.RemoteAddress.As4())
-	dstIp := netip.AddrFrom4(id.LocalAddress.As4())
+	srcIp := addrToNetipAddr(id.RemoteAddress)
+	dstIp := addrToNetipAddr(id.LocalAddress)
 
 	fields := nftypes.EventFields{
-		FlowID:    flowID,
-		Type:      typ,
-		Direction: nftypes.Ingress,
-		Protocol:  nftypes.TCP,
-		// TODO: handle ipv6
+		FlowID:     flowID,
+		Type:       typ,
+		Direction:  nftypes.Ingress,
+		Protocol:   nftypes.TCP,
 		SourceIP:   srcIp,
 		DestIP:     dstIp,
 		SourcePort: id.RemotePort,

client/firewall/uspfilter/forwarder/udp.go

@@ -6,7 +6,7 @@ import (
 	"fmt"
 	"io"
 	"net"
-	"net/netip"
+	"strconv"
 	"sync"
 	"sync/atomic"
 	"time"
@@ -158,7 +158,7 @@ func (f *Forwarder) handleUDP(r *udp.ForwarderRequest) bool {
 		}
 	}()
 
-	dstAddr := fmt.Sprintf("%s:%d", f.determineDialAddr(id.LocalAddress), id.LocalPort)
+	dstAddr := net.JoinHostPort(f.determineDialAddr(id.LocalAddress).String(), strconv.Itoa(int(id.LocalPort)))
 	outConn, err := (&net.Dialer{}).DialContext(f.ctx, "udp", dstAddr)
 	if err != nil {
 		f.logger.Debug2("forwarder: UDP dial error for %v: %v", epID(id), err)
@@ -276,15 +276,14 @@ func (f *Forwarder) proxyUDP(ctx context.Context, pConn *udpPacketConn, id stack
 
 // sendUDPEvent stores flow events for UDP connections
 func (f *Forwarder) sendUDPEvent(typ nftypes.Type, flowID uuid.UUID, id stack.TransportEndpointID, rxBytes, txBytes, rxPackets, txPackets uint64) {
-	srcIp := netip.AddrFrom4(id.RemoteAddress.As4())
-	dstIp := netip.AddrFrom4(id.LocalAddress.As4())
+	srcIp := addrToNetipAddr(id.RemoteAddress)
+	dstIp := addrToNetipAddr(id.LocalAddress)
 
 	fields := nftypes.EventFields{
-		FlowID:    flowID,
-		Type:      typ,
-		Direction: nftypes.Ingress,
-		Protocol:  nftypes.UDP,
-		// TODO: handle ipv6
+		FlowID:     flowID,
+		Type:       typ,
+		Direction:  nftypes.Ingress,
+		Protocol:   nftypes.UDP,
 		SourceIP:   srcIp,
 		DestIP:     dstIp,
 		SourcePort: id.RemotePort,

client/firewall/uspfilter/localip.go

@@ -4,89 +4,32 @@ import (
 	"fmt"
 	"net"
 	"net/netip"
-	"sync"
+	"sync/atomic"
 
 	log "github.com/sirupsen/logrus"
 
 	"github.com/netbirdio/netbird/client/firewall/uspfilter/common"
 )
 
-type localIPManager struct {
-	mu sync.RWMutex
-
-	// fixed-size high array for upper byte of a IPv4 address
-	ipv4Bitmap [256]*ipv4LowBitmap
+// localIPSnapshot is an immutable snapshot of local IP addresses, swapped
+// atomically so reads are lock-free.
+type localIPSnapshot struct {
+	ips map[netip.Addr]struct{}
 }
 
-// ipv4LowBitmap is a map for the low 16 bits of a IPv4 address
-type ipv4LowBitmap struct {
-	bitmap [8192]uint32
+type localIPManager struct {
+	snapshot atomic.Pointer[localIPSnapshot]
 }
 
 func newLocalIPManager() *localIPManager {
-	return &localIPManager{}
+	m := &localIPManager{}
+	m.snapshot.Store(&localIPSnapshot{
+		ips: make(map[netip.Addr]struct{}),
+	})
+	return m
 }
 
-func (m *localIPManager) setBitmapBit(ip net.IP) {
-	ipv4 := ip.To4()
-	if ipv4 == nil {
-		return
-	}
-	high := uint16(ipv4[0])
-	low := (uint16(ipv4[1]) << 8) | (uint16(ipv4[2]) << 4) | uint16(ipv4[3])
-
-	index := low / 32
-	bit := low % 32
-
-	if m.ipv4Bitmap[high] == nil {
-		m.ipv4Bitmap[high] = &ipv4LowBitmap{}
-	}
-
-	m.ipv4Bitmap[high].bitmap[index] |= 1 << bit
-}
-
-func (m *localIPManager) setBitInBitmap(ip netip.Addr, bitmap *[256]*ipv4LowBitmap, ipv4Set map[netip.Addr]struct{}, ipv4Addresses *[]netip.Addr) {
-	if !ip.Is4() {
-		return
-	}
-	ipv4 := ip.AsSlice()
-
-	high := uint16(ipv4[0])
-	low := (uint16(ipv4[1]) << 8) | (uint16(ipv4[2]) << 4) | uint16(ipv4[3])
-
-	if bitmap[high] == nil {
-		bitmap[high] = &ipv4LowBitmap{}
-	}
-
-	index := low / 32
-	bit := low % 32
-	bitmap[high].bitmap[index] |= 1 << bit
-
-	if _, exists := ipv4Set[ip]; !exists {
-		ipv4Set[ip] = struct{}{}
-		*ipv4Addresses = append(*ipv4Addresses, ip)
-	}
-}
-
-func (m *localIPManager) checkBitmapBit(ip []byte) bool {
-	high := uint16(ip[0])
-	low := (uint16(ip[1]) << 8) | (uint16(ip[2]) << 4) | uint16(ip[3])
-
-	if m.ipv4Bitmap[high] == nil {
-		return false
-	}
-
-	index := low / 32
-	bit := low % 32
-	return (m.ipv4Bitmap[high].bitmap[index] & (1 << bit)) != 0
-}
-
-func (m *localIPManager) processIP(ip netip.Addr, bitmap *[256]*ipv4LowBitmap, ipv4Set map[netip.Addr]struct{}, ipv4Addresses *[]netip.Addr) error {
-	m.setBitInBitmap(ip, bitmap, ipv4Set, ipv4Addresses)
-	return nil
-}
-
-func (m *localIPManager) processInterface(iface net.Interface, bitmap *[256]*ipv4LowBitmap, ipv4Set map[netip.Addr]struct{}, ipv4Addresses *[]netip.Addr) {
+func processInterface(iface net.Interface, ips map[netip.Addr]struct{}, addresses *[]netip.Addr) {
 	addrs, err := iface.Addrs()
 	if err != nil {
 		log.Debugf("get addresses for interface %s failed: %v", iface.Name, err)
@@ -104,18 +47,19 @@ func (m *localIPManager) processInterface(iface net.Interface, bitmap *[256]*ipv
 			continue
 		}
 
-		addr, ok := netip.AddrFromSlice(ip)
+		parsed, ok := netip.AddrFromSlice(ip)
 		if !ok {
 			log.Warnf("invalid IP address %s in interface %s", ip.String(), iface.Name)
 			continue
 		}
 
-		if err := m.processIP(addr.Unmap(), bitmap, ipv4Set, ipv4Addresses); err != nil {
-			log.Debugf("process IP failed: %v", err)
-		}
+		parsed = parsed.Unmap()
+		ips[parsed] = struct{}{}
+		*addresses = append(*addresses, parsed)
 	}
 }
 
+// UpdateLocalIPs rebuilds the local IP snapshot and swaps it in atomically.
 func (m *localIPManager) UpdateLocalIPs(iface common.IFaceMapper) (err error) {
 	defer func() {
 		if r := recover(); r != nil {
@@ -123,20 +67,20 @@ func (m *localIPManager) UpdateLocalIPs(iface common.IFaceMapper) (err error) {
 		}
 	}()
 
-	var newIPv4Bitmap [256]*ipv4LowBitmap
-	ipv4Set := make(map[netip.Addr]struct{})
-	var ipv4Addresses []netip.Addr
+	ips := make(map[netip.Addr]struct{})
+	var addresses []netip.Addr
 
-	// 127.0.0.0/8
-	newIPv4Bitmap[127] = &ipv4LowBitmap{}
-	for i := 0; i < 8192; i++ {
-		// #nosec G602 -- bitmap is defined as [8192]uint32, loop range is correct
-		newIPv4Bitmap[127].bitmap[i] = 0xFFFFFFFF
-	}
+	// loopback
+	ips[netip.AddrFrom4([4]byte{127, 0, 0, 1})] = struct{}{}
+	ips[netip.IPv6Loopback()] = struct{}{}
 
 	if iface != nil {
-		if err := m.processIP(iface.Address().IP, &newIPv4Bitmap, ipv4Set, &ipv4Addresses); err != nil {
-			return err
+		ip := iface.Address().IP
+		ips[ip] = struct{}{}
+		addresses = append(addresses, ip)
+		if v6 := iface.Address().IPv6; v6.IsValid() {
+			ips[v6] = struct{}{}
+			addresses = append(addresses, v6)
 		}
 	}
 
@@ -144,26 +88,27 @@ func (m *localIPManager) UpdateLocalIPs(iface common.IFaceMapper) (err error) {
 	if err != nil {
 		log.Warnf("failed to get interfaces: %v", err)
 	} else {
+		// TODO: filter out down interfaces (net.FlagUp). Also handle the reverse
+		// case where an interface comes up between refreshes.
 		for _, intf := range interfaces {
-			m.processInterface(intf, &newIPv4Bitmap, ipv4Set, &ipv4Addresses)
+			processInterface(intf, ips, &addresses)
 		}
 	}
 
-	m.mu.Lock()
-	m.ipv4Bitmap = newIPv4Bitmap
-	m.mu.Unlock()
+	m.snapshot.Store(&localIPSnapshot{ips: ips})
 
-	log.Debugf("Local IPv4 addresses: %v", ipv4Addresses)
+	log.Debugf("Local IP addresses: %v", addresses)
 	return nil
 }
 
+// IsLocalIP checks if the given IP is a local address. Lock-free on the read path.
 func (m *localIPManager) IsLocalIP(ip netip.Addr) bool {
-	if !ip.Is4() {
-		return false
+	s := m.snapshot.Load()
+
+	if ip.Is4() && ip.As4()[0] == 127 {
+		return true
 	}
 
-	m.mu.RLock()
-	defer m.mu.RUnlock()
-
-	return m.checkBitmapBit(ip.AsSlice())
+	_, found := s.ips[ip]
+	return found
 }

client/firewall/uspfilter/localip_bench_test.go

@@ -0,0 +1,72 @@
+package uspfilter
+
+import (
+	"net/netip"
+	"testing"
+
+	"github.com/netbirdio/netbird/client/iface/wgaddr"
+)
+
+func setupManager(b *testing.B) *localIPManager {
+	b.Helper()
+	m := newLocalIPManager()
+	mock := &IFaceMock{
+		AddressFunc: func() wgaddr.Address {
+			return wgaddr.Address{
+				IP:      netip.MustParseAddr("100.64.0.1"),
+				Network: netip.MustParsePrefix("100.64.0.0/16"),
+				IPv6:    netip.MustParseAddr("fd00::1"),
+				IPv6Net: netip.MustParsePrefix("fd00::/64"),
+			}
+		},
+	}
+	if err := m.UpdateLocalIPs(mock); err != nil {
+		b.Fatalf("UpdateLocalIPs: %v", err)
+	}
+	return m
+}
+
+func BenchmarkIsLocalIP_v4_hit(b *testing.B) {
+	m := setupManager(b)
+	ip := netip.MustParseAddr("100.64.0.1")
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		m.IsLocalIP(ip)
+	}
+}
+
+func BenchmarkIsLocalIP_v4_miss(b *testing.B) {
+	m := setupManager(b)
+	ip := netip.MustParseAddr("8.8.8.8")
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		m.IsLocalIP(ip)
+	}
+}
+
+func BenchmarkIsLocalIP_v6_hit(b *testing.B) {
+	m := setupManager(b)
+	ip := netip.MustParseAddr("fd00::1")
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		m.IsLocalIP(ip)
+	}
+}
+
+func BenchmarkIsLocalIP_v6_miss(b *testing.B) {
+	m := setupManager(b)
+	ip := netip.MustParseAddr("2001:db8::1")
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		m.IsLocalIP(ip)
+	}
+}
+
+func BenchmarkIsLocalIP_loopback(b *testing.B) {
+	m := setupManager(b)
+	ip := netip.MustParseAddr("127.0.0.1")
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		m.IsLocalIP(ip)
+	}
+}

client/firewall/uspfilter/localip_test.go

@@ -72,14 +72,45 @@ func TestLocalIPManager(t *testing.T) {
 			expected: false,
 		},
 		{
-			name: "IPv6 address",
+			name: "IPv6 address matches",
 			setupAddr: wgaddr.Address{
-				IP:      netip.MustParseAddr("fe80::1"),
+				IP:      netip.MustParseAddr("100.64.0.1"),
+				Network: netip.MustParsePrefix("100.64.0.0/16"),
+				IPv6:    netip.MustParseAddr("fd00::1"),
+				IPv6Net: netip.MustParsePrefix("fd00::/64"),
+			},
+			testIP:   netip.MustParseAddr("fd00::1"),
+			expected: true,
+		},
+		{
+			name: "IPv6 address does not match",
+			setupAddr: wgaddr.Address{
+				IP:      netip.MustParseAddr("100.64.0.1"),
+				Network: netip.MustParsePrefix("100.64.0.0/16"),
+				IPv6:    netip.MustParseAddr("fd00::1"),
+				IPv6Net: netip.MustParsePrefix("fd00::/64"),
+			},
+			testIP:   netip.MustParseAddr("fd00::99"),
+			expected: false,
+		},
+		{
+			name: "No aliasing between similar IPs",
+			setupAddr: wgaddr.Address{
+				IP:      netip.MustParseAddr("192.168.1.1"),
 				Network: netip.MustParsePrefix("192.168.1.0/24"),
 			},
-			testIP:   netip.MustParseAddr("fe80::1"),
+			testIP:   netip.MustParseAddr("192.168.0.17"),
 			expected: false,
 		},
+		{
+			name: "IPv6 loopback",
+			setupAddr: wgaddr.Address{
+				IP:      netip.MustParseAddr("100.64.0.1"),
+				Network: netip.MustParsePrefix("100.64.0.0/16"),
+			},
+			testIP:   netip.MustParseAddr("::1"),
+			expected: true,
+		},
 	}
 
 	for _, tt := range tests {
@@ -171,90 +202,3 @@ func TestLocalIPManager_AllInterfaces(t *testing.T) {
 		})
 	}
 }
-
-// MapImplementation is a version using map[string]struct{}
-type MapImplementation struct {
-	localIPs map[string]struct{}
-}
-
-func BenchmarkIPChecks(b *testing.B) {
-	interfaces := make([]net.IP, 16)
-	for i := range interfaces {
-		interfaces[i] = net.IPv4(10, 0, byte(i>>8), byte(i))
-	}
-
-	// Setup bitmap
-	bitmapManager := newLocalIPManager()
-	for _, ip := range interfaces[:8] { // Add half of IPs
-		bitmapManager.setBitmapBit(ip)
-	}
-
-	// Setup map version
-	mapManager := &MapImplementation{
-		localIPs: make(map[string]struct{}),
-	}
-	for _, ip := range interfaces[:8] {
-		mapManager.localIPs[ip.String()] = struct{}{}
-	}
-
-	b.Run("Bitmap_Hit", func(b *testing.B) {
-		ip := interfaces[4]
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			bitmapManager.checkBitmapBit(ip)
-		}
-	})
-
-	b.Run("Bitmap_Miss", func(b *testing.B) {
-		ip := interfaces[12]
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			bitmapManager.checkBitmapBit(ip)
-		}
-	})
-
-	b.Run("Map_Hit", func(b *testing.B) {
-		ip := interfaces[4]
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			// nolint:gosimple
-			_ = mapManager.localIPs[ip.String()]
-		}
-	})
-
-	b.Run("Map_Miss", func(b *testing.B) {
-		ip := interfaces[12]
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			// nolint:gosimple
-			_ = mapManager.localIPs[ip.String()]
-		}
-	})
-}
-
-func BenchmarkWGPosition(b *testing.B) {
-	wgIP := net.ParseIP("10.10.0.1")
-
-	// Create two managers - one checks WG IP first, other checks it last
-	b.Run("WG_First", func(b *testing.B) {
-		bm := newLocalIPManager()
-		bm.setBitmapBit(wgIP)
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			bm.checkBitmapBit(wgIP)
-		}
-	})
-
-	b.Run("WG_Last", func(b *testing.B) {
-		bm := newLocalIPManager()
-		// Fill with other IPs first
-		for i := 0; i < 15; i++ {
-			bm.setBitmapBit(net.IPv4(10, 0, byte(i>>8), byte(i)))
-		}
-		bm.setBitmapBit(wgIP) // Add WG IP last
-		b.ResetTimer()
-		for i := 0; i < b.N; i++ {
-			bm.checkBitmapBit(wgIP)
-		}
-	})
-}

client/firewall/uspfilter/nat.go

@@ -13,8 +13,6 @@ import (
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
 )
 
-var ErrIPv4Only = errors.New("only IPv4 is supported for DNAT")
-
 var (
 	errInvalidIPHeaderLength = errors.New("invalid IP header length")
 )
@@ -25,10 +23,33 @@ const (
 	destinationPortOffset = 2
 
 	// IP address offsets in IPv4 header
-	sourceIPOffset      = 12
-	destinationIPOffset = 16
+	ipv4SrcOffset = 12
+	ipv4DstOffset = 16
+
+	// IP address offsets in IPv6 header
+	ipv6SrcOffset = 8
+	ipv6DstOffset = 24
+
+	// IPv6 fixed header length
+	ipv6HeaderLen = 40
 )
 
+// ipHeaderLen returns the IP header length based on the decoded layer type.
+func ipHeaderLen(d *decoder) (int, error) {
+	switch d.decoded[0] {
+	case layers.LayerTypeIPv4:
+		n := int(d.ip4.IHL) * 4
+		if n < 20 {
+			return 0, errInvalidIPHeaderLength
+		}
+		return n, nil
+	case layers.LayerTypeIPv6:
+		return ipv6HeaderLen, nil
+	default:
+		return 0, fmt.Errorf("unknown IP layer: %v", d.decoded[0])
+	}
+}
+
 // ipv4Checksum calculates IPv4 header checksum.
 func ipv4Checksum(header []byte) uint16 {
 	if len(header) < 20 {
@@ -234,14 +255,13 @@ func (m *Manager) translateOutboundDNAT(packetData []byte, d *decoder) bool {
 		return false
 	}
 
-	dstIP := netip.AddrFrom4([4]byte{packetData[16], packetData[17], packetData[18], packetData[19]})
-
+	_, dstIP := extractPacketIPs(packetData, d)
 	translatedIP, exists := m.getDNATTranslation(dstIP)
 	if !exists {
 		return false
 	}
 
-	if err := m.rewritePacketIP(packetData, d, translatedIP, destinationIPOffset); err != nil {
+	if err := m.rewritePacketIP(packetData, d, translatedIP, false); err != nil {
 		m.logger.Error1("failed to rewrite packet destination: %v", err)
 		return false
 	}
@@ -256,14 +276,13 @@ func (m *Manager) translateInboundReverse(packetData []byte, d *decoder) bool {
 		return false
 	}
 
-	srcIP := netip.AddrFrom4([4]byte{packetData[12], packetData[13], packetData[14], packetData[15]})
-
+	srcIP, _ := extractPacketIPs(packetData, d)
 	originalIP, exists := m.findReverseDNATMapping(srcIP)
 	if !exists {
 		return false
 	}
 
-	if err := m.rewritePacketIP(packetData, d, originalIP, sourceIPOffset); err != nil {
+	if err := m.rewritePacketIP(packetData, d, originalIP, true); err != nil {
 		m.logger.Error1("failed to rewrite packet source: %v", err)
 		return false
 	}
@@ -272,38 +291,96 @@ func (m *Manager) translateInboundReverse(packetData []byte, d *decoder) bool {
 	return true
 }
 
-// rewritePacketIP replaces an IP address (source or destination) in the packet and updates checksums.
-func (m *Manager) rewritePacketIP(packetData []byte, d *decoder, newIP netip.Addr, ipOffset int) error {
+// extractPacketIPs extracts src and dst IP addresses directly from raw packet bytes.
+func extractPacketIPs(packetData []byte, d *decoder) (src, dst netip.Addr) {
+	switch d.decoded[0] {
+	case layers.LayerTypeIPv4:
+		src = netip.AddrFrom4([4]byte{packetData[ipv4SrcOffset], packetData[ipv4SrcOffset+1], packetData[ipv4SrcOffset+2], packetData[ipv4SrcOffset+3]})
+		dst = netip.AddrFrom4([4]byte{packetData[ipv4DstOffset], packetData[ipv4DstOffset+1], packetData[ipv4DstOffset+2], packetData[ipv4DstOffset+3]})
+	case layers.LayerTypeIPv6:
+		src = netip.AddrFrom16([16]byte(packetData[ipv6SrcOffset : ipv6SrcOffset+16]))
+		dst = netip.AddrFrom16([16]byte(packetData[ipv6DstOffset : ipv6DstOffset+16]))
+	}
+	return src, dst
+}
+
+// rewritePacketIP replaces a source (isSource=true) or destination IP address in the packet and updates checksums.
+func (m *Manager) rewritePacketIP(packetData []byte, d *decoder, newIP netip.Addr, isSource bool) error {
+	hdrLen, err := ipHeaderLen(d)
+	if err != nil {
+		return err
+	}
+
+	switch d.decoded[0] {
+	case layers.LayerTypeIPv4:
+		return m.rewriteIPv4(packetData, d, newIP, hdrLen, isSource)
+	case layers.LayerTypeIPv6:
+		return m.rewriteIPv6(packetData, d, newIP, hdrLen, isSource)
+	default:
+		return fmt.Errorf("unknown IP layer: %v", d.decoded[0])
+	}
+}
+
+func (m *Manager) rewriteIPv4(packetData []byte, d *decoder, newIP netip.Addr, hdrLen int, isSource bool) error {
 	if !newIP.Is4() {
-		return ErrIPv4Only
+		return fmt.Errorf("cannot write IPv6 address into IPv4 packet")
+	}
+
+	offset := ipv4DstOffset
+	if isSource {
+		offset = ipv4SrcOffset
 	}
 
 	var oldIP [4]byte
-	copy(oldIP[:], packetData[ipOffset:ipOffset+4])
+	copy(oldIP[:], packetData[offset:offset+4])
 	newIPBytes := newIP.As4()
+	copy(packetData[offset:offset+4], newIPBytes[:])
 
-	copy(packetData[ipOffset:ipOffset+4], newIPBytes[:])
-
-	ipHeaderLen := int(d.ip4.IHL) * 4
-	if ipHeaderLen < 20 || ipHeaderLen > len(packetData) {
-		return errInvalidIPHeaderLength
-	}
-
+	// Recalculate IPv4 header checksum
 	binary.BigEndian.PutUint16(packetData[10:12], 0)
-	ipChecksum := ipv4Checksum(packetData[:ipHeaderLen])
-	binary.BigEndian.PutUint16(packetData[10:12], ipChecksum)
+	binary.BigEndian.PutUint16(packetData[10:12], ipv4Checksum(packetData[:hdrLen]))
 
+	// Update transport checksums incrementally
 	if len(d.decoded) > 1 {
 		switch d.decoded[1] {
 		case layers.LayerTypeTCP:
-			m.updateTCPChecksum(packetData, ipHeaderLen, oldIP[:], newIPBytes[:])
+			m.updateTCPChecksum(packetData, hdrLen, oldIP[:], newIPBytes[:])
 		case layers.LayerTypeUDP:
-			m.updateUDPChecksum(packetData, ipHeaderLen, oldIP[:], newIPBytes[:])
+			m.updateUDPChecksum(packetData, hdrLen, oldIP[:], newIPBytes[:])
 		case layers.LayerTypeICMPv4:
-			m.updateICMPChecksum(packetData, ipHeaderLen)
+			m.updateICMPChecksum(packetData, hdrLen)
 		}
 	}
+	return nil
+}
 
+func (m *Manager) rewriteIPv6(packetData []byte, d *decoder, newIP netip.Addr, hdrLen int, isSource bool) error {
+	if !newIP.Is6() {
+		return fmt.Errorf("cannot write IPv4 address into IPv6 packet")
+	}
+
+	offset := ipv6DstOffset
+	if isSource {
+		offset = ipv6SrcOffset
+	}
+
+	var oldIP [16]byte
+	copy(oldIP[:], packetData[offset:offset+16])
+	newIPBytes := newIP.As16()
+	copy(packetData[offset:offset+16], newIPBytes[:])
+
+	// IPv6 has no header checksum, only update transport checksums
+	if len(d.decoded) > 1 {
+		switch d.decoded[1] {
+		case layers.LayerTypeTCP:
+			m.updateTCPChecksum(packetData, hdrLen, oldIP[:], newIPBytes[:])
+		case layers.LayerTypeUDP:
+			m.updateUDPChecksum(packetData, hdrLen, oldIP[:], newIPBytes[:])
+		case layers.LayerTypeICMPv6:
+			// ICMPv6 checksum includes pseudo-header with addresses, use incremental update
+			m.updateICMPv6Checksum(packetData, hdrLen, oldIP[:], newIPBytes[:])
+		}
+	}
 	return nil
 }
 
@@ -351,6 +428,20 @@ func (m *Manager) updateICMPChecksum(packetData []byte, ipHeaderLen int) {
 	binary.BigEndian.PutUint16(icmpData[2:4], checksum)
 }
 
+// updateICMPv6Checksum updates ICMPv6 checksum after address change.
+// ICMPv6 uses a pseudo-header (like TCP/UDP), so incremental update applies.
+func (m *Manager) updateICMPv6Checksum(packetData []byte, ipHeaderLen int, oldIP, newIP []byte) {
+	icmpStart := ipHeaderLen
+	if len(packetData) < icmpStart+4 {
+		return
+	}
+
+	checksumOffset := icmpStart + 2
+	oldChecksum := binary.BigEndian.Uint16(packetData[checksumOffset : checksumOffset+2])
+	newChecksum := incrementalUpdate(oldChecksum, oldIP, newIP)
+	binary.BigEndian.PutUint16(packetData[checksumOffset:checksumOffset+2], newChecksum)
+}
+
 // incrementalUpdate performs incremental checksum update per RFC 1624.
 func incrementalUpdate(oldChecksum uint16, oldBytes, newBytes []byte) uint16 {
 	sum := uint32(^oldChecksum)
@@ -421,6 +512,7 @@ func (m *Manager) addPortRedirection(targetIP netip.Addr, protocol gopacket.Laye
 }
 
 // AddInboundDNAT adds an inbound DNAT rule redirecting traffic from NetBird peers to local services.
+// TODO: also delegate to nativeFirewall when available for kernel WG mode
 func (m *Manager) AddInboundDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
 	var layerType gopacket.LayerType
 	switch protocol {
@@ -466,6 +558,22 @@ func (m *Manager) RemoveInboundDNAT(localAddr netip.Addr, protocol firewall.Prot
 	return m.removePortRedirection(localAddr, layerType, sourcePort, targetPort)
 }
 
+// AddOutputDNAT delegates to the native firewall if available.
+func (m *Manager) AddOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	if m.nativeFirewall == nil {
+		return fmt.Errorf("output DNAT not supported without native firewall")
+	}
+	return m.nativeFirewall.AddOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
+// RemoveOutputDNAT delegates to the native firewall if available.
+func (m *Manager) RemoveOutputDNAT(localAddr netip.Addr, protocol firewall.Protocol, sourcePort, targetPort uint16) error {
+	if m.nativeFirewall == nil {
+		return nil
+	}
+	return m.nativeFirewall.RemoveOutputDNAT(localAddr, protocol, sourcePort, targetPort)
+}
+
 // translateInboundPortDNAT applies port-specific DNAT translation to inbound packets.
 func (m *Manager) translateInboundPortDNAT(packetData []byte, d *decoder, srcIP, dstIP netip.Addr) bool {
 	if !m.portDNATEnabled.Load() {
@@ -515,12 +623,12 @@ func (m *Manager) applyPortRule(packetData []byte, d *decoder, srcIP, dstIP neti
 
 // rewriteTCPPort rewrites a TCP port (source or destination) and updates checksum.
 func (m *Manager) rewriteTCPPort(packetData []byte, d *decoder, newPort uint16, portOffset int) error {
-	ipHeaderLen := int(d.ip4.IHL) * 4
-	if ipHeaderLen < 20 || ipHeaderLen > len(packetData) {
-		return errInvalidIPHeaderLength
+	hdrLen, err := ipHeaderLen(d)
+	if err != nil {
+		return err
 	}
 
-	tcpStart := ipHeaderLen
+	tcpStart := hdrLen
 	if len(packetData) < tcpStart+4 {
 		return fmt.Errorf("packet too short for TCP header")
 	}
@@ -546,12 +654,12 @@ func (m *Manager) rewriteTCPPort(packetData []byte, d *decoder, newPort uint16,
 
 // rewriteUDPPort rewrites a UDP port (source or destination) and updates checksum.
 func (m *Manager) rewriteUDPPort(packetData []byte, d *decoder, newPort uint16, portOffset int) error {
-	ipHeaderLen := int(d.ip4.IHL) * 4
-	if ipHeaderLen < 20 || ipHeaderLen > len(packetData) {
-		return errInvalidIPHeaderLength
+	hdrLen, err := ipHeaderLen(d)
+	if err != nil {
+		return err
 	}
 
-	udpStart := ipHeaderLen
+	udpStart := hdrLen
 	if len(packetData) < udpStart+8 {
 		return fmt.Errorf("packet too short for UDP header")
 	}

client/firewall/uspfilter/nat_bench_test.go

@@ -342,12 +342,17 @@ func BenchmarkDNATMemoryAllocations(b *testing.B) {
 
 		// Parse the packet fresh each time to get a clean decoder
 		d := &decoder{decoded: []gopacket.LayerType{}}
-		d.parser = gopacket.NewDecodingLayerParser(
+		d.parser4 = gopacket.NewDecodingLayerParser(
 			layers.LayerTypeIPv4,
 			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 		)
-		d.parser.IgnoreUnsupported = true
-		err = d.parser.DecodeLayers(testPacket, &d.decoded)
+		d.parser4.IgnoreUnsupported = true
+		d.parser6 = gopacket.NewDecodingLayerParser(
+			layers.LayerTypeIPv6,
+			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+		)
+		d.parser6.IgnoreUnsupported = true
+		err = d.decodePacket(testPacket)
 		assert.NoError(b, err)
 
 		manager.translateOutboundDNAT(testPacket, d)
@@ -371,12 +376,17 @@ func BenchmarkDirectIPExtraction(b *testing.B) {
 	b.Run("decoder_extraction", func(b *testing.B) {
 		// Create decoder once for comparison
 		d := &decoder{decoded: []gopacket.LayerType{}}
-		d.parser = gopacket.NewDecodingLayerParser(
+		d.parser4 = gopacket.NewDecodingLayerParser(
 			layers.LayerTypeIPv4,
 			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 		)
-		d.parser.IgnoreUnsupported = true
-		err := d.parser.DecodeLayers(packet, &d.decoded)
+		d.parser4.IgnoreUnsupported = true
+		d.parser6 = gopacket.NewDecodingLayerParser(
+			layers.LayerTypeIPv6,
+			&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+		)
+		d.parser6.IgnoreUnsupported = true
+		err := d.decodePacket(packet)
 		assert.NoError(b, err)
 
 		for i := 0; i < b.N; i++ {

client/firewall/uspfilter/nat_test.go

@@ -86,13 +86,18 @@ func parsePacket(t testing.TB, packetData []byte) *decoder {
 	d := &decoder{
 		decoded: []gopacket.LayerType{},
 	}
-	d.parser = gopacket.NewDecodingLayerParser(
+	d.parser4 = gopacket.NewDecodingLayerParser(
 		layers.LayerTypeIPv4,
 		&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
 	)
-	d.parser.IgnoreUnsupported = true
+	d.parser4.IgnoreUnsupported = true
+	d.parser6 = gopacket.NewDecodingLayerParser(
+		layers.LayerTypeIPv6,
+		&d.eth, &d.ip4, &d.ip6, &d.icmp4, &d.icmp6, &d.tcp, &d.udp,
+	)
+	d.parser6.IgnoreUnsupported = true
 
-	err := d.parser.DecodeLayers(packetData, &d.decoded)
+	err := d.decodePacket(packetData)
 	require.NoError(t, err)
 	return d
 }

client/firewall/uspfilter/rule.go

@@ -18,9 +18,7 @@ type PeerRule struct {
 	protoLayer gopacket.LayerType
 	sPort      *firewall.Port
 	dPort      *firewall.Port
-	drop       bool
-
-	udpHook func([]byte) bool
+	drop bool
 }
 
 // ID returns the rule id

client/firewall/uspfilter/tracer.go

@@ -112,10 +112,13 @@ func (t *PacketTrace) AddResultWithForwarder(stage PacketStage, message string,
 }
 
 func (p *PacketBuilder) Build() ([]byte, error) {
-	ip := p.buildIPLayer()
-	pktLayers := []gopacket.SerializableLayer{ip}
+	ipLayer, err := p.buildIPLayer()
+	if err != nil {
+		return nil, err
+	}
+	pktLayers := []gopacket.SerializableLayer{ipLayer}
 
-	transportLayer, err := p.buildTransportLayer(ip)
+	transportLayer, err := p.buildTransportLayer(ipLayer)
 	if err != nil {
 		return nil, err
 	}
@@ -129,30 +132,43 @@ func (p *PacketBuilder) Build() ([]byte, error) {
 	return serializePacket(pktLayers)
 }
 
-func (p *PacketBuilder) buildIPLayer() *layers.IPv4 {
+func (p *PacketBuilder) buildIPLayer() (gopacket.SerializableLayer, error) {
+	if p.SrcIP.Is4() != p.DstIP.Is4() {
+		return nil, fmt.Errorf("mixed address families: src=%s dst=%s", p.SrcIP, p.DstIP)
+	}
+	proto := getIPProtocolNumber(p.Protocol, p.SrcIP.Is6())
+	if p.SrcIP.Is6() {
+		return &layers.IPv6{
+			Version:    6,
+			HopLimit:   64,
+			NextHeader: proto,
+			SrcIP:      p.SrcIP.AsSlice(),
+			DstIP:      p.DstIP.AsSlice(),
+		}, nil
+	}
 	return &layers.IPv4{
 		Version:  4,
 		TTL:      64,
-		Protocol: layers.IPProtocol(getIPProtocolNumber(p.Protocol)),
+		Protocol: proto,
 		SrcIP:    p.SrcIP.AsSlice(),
 		DstIP:    p.DstIP.AsSlice(),
-	}
+	}, nil
 }
 
-func (p *PacketBuilder) buildTransportLayer(ip *layers.IPv4) ([]gopacket.SerializableLayer, error) {
+func (p *PacketBuilder) buildTransportLayer(ipLayer gopacket.SerializableLayer) ([]gopacket.SerializableLayer, error) {
 	switch p.Protocol {
 	case "tcp":
-		return p.buildTCPLayer(ip)
+		return p.buildTCPLayer(ipLayer)
 	case "udp":
-		return p.buildUDPLayer(ip)
+		return p.buildUDPLayer(ipLayer)
 	case "icmp":
-		return p.buildICMPLayer()
+		return p.buildICMPLayer(ipLayer)
 	default:
 		return nil, fmt.Errorf("unsupported protocol: %s", p.Protocol)
 	}
 }
 
-func (p *PacketBuilder) buildTCPLayer(ip *layers.IPv4) ([]gopacket.SerializableLayer, error) {
+func (p *PacketBuilder) buildTCPLayer(ipLayer gopacket.SerializableLayer) ([]gopacket.SerializableLayer, error) {
 	tcp := &layers.TCP{
 		SrcPort: layers.TCPPort(p.SrcPort),
 		DstPort: layers.TCPPort(p.DstPort),
@@ -164,24 +180,44 @@ func (p *PacketBuilder) buildTCPLayer(ip *layers.IPv4) ([]gopacket.SerializableL
 		PSH:     p.TCPState != nil && p.TCPState.PSH,
 		URG:     p.TCPState != nil && p.TCPState.URG,
 	}
-	if err := tcp.SetNetworkLayerForChecksum(ip); err != nil {
-		return nil, fmt.Errorf("set network layer for TCP checksum: %w", err)
+	if nl, ok := ipLayer.(gopacket.NetworkLayer); ok {
+		if err := tcp.SetNetworkLayerForChecksum(nl); err != nil {
+			return nil, fmt.Errorf("set network layer for TCP checksum: %w", err)
+		}
 	}
 	return []gopacket.SerializableLayer{tcp}, nil
 }
 
-func (p *PacketBuilder) buildUDPLayer(ip *layers.IPv4) ([]gopacket.SerializableLayer, error) {
+func (p *PacketBuilder) buildUDPLayer(ipLayer gopacket.SerializableLayer) ([]gopacket.SerializableLayer, error) {
 	udp := &layers.UDP{
 		SrcPort: layers.UDPPort(p.SrcPort),
 		DstPort: layers.UDPPort(p.DstPort),
 	}
-	if err := udp.SetNetworkLayerForChecksum(ip); err != nil {
-		return nil, fmt.Errorf("set network layer for UDP checksum: %w", err)
+	if nl, ok := ipLayer.(gopacket.NetworkLayer); ok {
+		if err := udp.SetNetworkLayerForChecksum(nl); err != nil {
+			return nil, fmt.Errorf("set network layer for UDP checksum: %w", err)
+		}
 	}
 	return []gopacket.SerializableLayer{udp}, nil
 }
 
-func (p *PacketBuilder) buildICMPLayer() ([]gopacket.SerializableLayer, error) {
+func (p *PacketBuilder) buildICMPLayer(ipLayer gopacket.SerializableLayer) ([]gopacket.SerializableLayer, error) {
+	if p.SrcIP.Is6() || p.DstIP.Is6() {
+		icmp := &layers.ICMPv6{
+			TypeCode: layers.CreateICMPv6TypeCode(p.ICMPType, p.ICMPCode),
+		}
+		if nl, ok := ipLayer.(gopacket.NetworkLayer); ok {
+			_ = icmp.SetNetworkLayerForChecksum(nl)
+		}
+		if p.ICMPType == layers.ICMPv6TypeEchoRequest || p.ICMPType == layers.ICMPv6TypeEchoReply {
+			echo := &layers.ICMPv6Echo{
+				Identifier: 1,
+				SeqNumber:  1,
+			}
+			return []gopacket.SerializableLayer{icmp, echo}, nil
+		}
+		return []gopacket.SerializableLayer{icmp}, nil
+	}
 	icmp := &layers.ICMPv4{
 		TypeCode: layers.CreateICMPv4TypeCode(p.ICMPType, p.ICMPCode),
 	}
@@ -204,14 +240,17 @@ func serializePacket(layers []gopacket.SerializableLayer) ([]byte, error) {
 	return buf.Bytes(), nil
 }
 
-func getIPProtocolNumber(protocol fw.Protocol) int {
+func getIPProtocolNumber(protocol fw.Protocol, isV6 bool) layers.IPProtocol {
 	switch protocol {
 	case fw.ProtocolTCP:
-		return int(layers.IPProtocolTCP)
+		return layers.IPProtocolTCP
 	case fw.ProtocolUDP:
-		return int(layers.IPProtocolUDP)
+		return layers.IPProtocolUDP
 	case fw.ProtocolICMP:
-		return int(layers.IPProtocolICMPv4)
+		if isV6 {
+			return layers.IPProtocolICMPv6
+		}
+		return layers.IPProtocolICMPv4
 	default:
 		return 0
 	}
@@ -234,7 +273,7 @@ func (m *Manager) TracePacket(packetData []byte, direction fw.RuleDirection) *Pa
 	trace := &PacketTrace{Direction: direction}
 
 	// Initial packet decoding
-	if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+	if err := d.decodePacket(packetData); err != nil {
 		trace.AddResult(StageReceived, fmt.Sprintf("Failed to decode packet: %v", err), false)
 		return trace
 	}
@@ -256,6 +295,8 @@ func (m *Manager) TracePacket(packetData []byte, direction fw.RuleDirection) *Pa
 		trace.DestinationPort = uint16(d.udp.DstPort)
 	case layers.LayerTypeICMPv4:
 		trace.Protocol = "ICMP"
+	case layers.LayerTypeICMPv6:
+		trace.Protocol = "ICMPv6"
 	}
 
 	trace.AddResult(StageReceived, fmt.Sprintf("Received %s packet: %s:%d -> %s:%d",
@@ -319,6 +360,13 @@ func (m *Manager) buildConntrackStateMessage(d *decoder) string {
 			flags&conntrack.TCPFin != 0)
 	case layers.LayerTypeICMPv4:
 		msg += fmt.Sprintf(" (ICMP ID=%d, Seq=%d)", d.icmp4.Id, d.icmp4.Seq)
+	case layers.LayerTypeICMPv6:
+		var id, seq uint16
+		if len(d.icmp6.Payload) >= 4 {
+			id = uint16(d.icmp6.Payload[0])<<8 | uint16(d.icmp6.Payload[1])
+			seq = uint16(d.icmp6.Payload[2])<<8 | uint16(d.icmp6.Payload[3])
+		}
+		msg += fmt.Sprintf(" (ICMPv6 ID=%d, Seq=%d)", id, seq)
 	}
 	return msg
 }
@@ -415,7 +463,7 @@ func (m *Manager) traceOutbound(packetData []byte, trace *PacketTrace) *PacketTr
 	d := m.decoders.Get().(*decoder)
 	defer m.decoders.Put(d)
 
-	if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+	if err := d.decodePacket(packetData); err != nil {
 		trace.AddResult(StageCompleted, "Packet dropped - decode error", false)
 		return trace
 	}
@@ -434,7 +482,7 @@ func (m *Manager) traceOutbound(packetData []byte, trace *PacketTrace) *PacketTr
 func (m *Manager) handleInboundDNAT(trace *PacketTrace, packetData []byte, d *decoder, srcIP, dstIP *netip.Addr) bool {
 	portDNATApplied := m.traceInboundPortDNAT(trace, packetData, d)
 	if portDNATApplied {
-		if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+		if err := d.decodePacket(packetData); err != nil {
 			trace.AddResult(StageInboundPortDNAT, "Failed to re-decode after port DNAT", false)
 			return true
 		}
@@ -444,7 +492,7 @@ func (m *Manager) handleInboundDNAT(trace *PacketTrace, packetData []byte, d *de
 
 	nat1to1Applied := m.traceInbound1to1NAT(trace, packetData, d)
 	if nat1to1Applied {
-		if err := d.parser.DecodeLayers(packetData, &d.decoded); err != nil {
+		if err := d.decodePacket(packetData); err != nil {
 			trace.AddResult(StageInbound1to1NAT, "Failed to re-decode after 1:1 NAT", false)
 			return true
 		}
@@ -509,7 +557,7 @@ func (m *Manager) traceInbound1to1NAT(trace *PacketTrace, packetData []byte, d *
 		return false
 	}
 
-	srcIP := netip.AddrFrom4([4]byte{packetData[12], packetData[13], packetData[14], packetData[15]})
+	srcIP, _ := extractPacketIPs(packetData, d)
 
 	translated := m.translateInboundReverse(packetData, d)
 	if translated {
@@ -539,7 +587,7 @@ func (m *Manager) traceOutbound1to1NAT(trace *PacketTrace, packetData []byte, d
 		return false
 	}
 
-	dstIP := netip.AddrFrom4([4]byte{packetData[16], packetData[17], packetData[18], packetData[19]})
+	_, dstIP := extractPacketIPs(packetData, d)
 
 	translated := m.translateOutboundDNAT(packetData, d)
 	if translated {

client/firewall/uspfilter/tracer_test.go

@@ -399,21 +399,17 @@ func TestTracePacket(t *testing.T) {
 		{
 			name: "UDPTraffic_WithHook",
 			setup: func(m *Manager) {
-				hookFunc := func([]byte) bool {
-					return true
-				}
-				m.AddUDPPacketHook(true, netip.MustParseAddr("1.1.1.1"), 53, hookFunc)
+				m.SetUDPPacketHook(netip.MustParseAddr("100.10.255.254"), 53, func([]byte) bool {
+					return true // drop (intercepted by hook)
+				})
 			},
 			packetBuilder: func() *PacketBuilder {
-				return createPacketBuilder("1.1.1.1", "100.10.0.100", "udp", 12345, 53, fw.RuleDirectionIN)
+				return createPacketBuilder("100.10.0.100", "100.10.255.254", "udp", 12345, 53, fw.RuleDirectionOUT)
 			},
 			expectedStages: []PacketStage{
 				StageReceived,
-				StageInboundPortDNAT,
-				StageInbound1to1NAT,
-				StageConntrack,
-				StageRouting,
-				StagePeerACL,
+				StageOutbound1to1NAT,
+				StageOutboundPortReverse,
 				StageCompleted,
 			},
 			expectedAllow: false,

client/grpc/dialer.go

@@ -28,7 +28,7 @@ func Backoff(ctx context.Context) backoff.BackOff {
 
 // CreateConnection creates a gRPC client connection with the appropriate transport options.
 // The component parameter specifies the WebSocket proxy component path (e.g., "/management", "/signal").
-func CreateConnection(ctx context.Context, addr string, tlsEnabled bool, component string) (*grpc.ClientConn, error) {
+func CreateConnection(ctx context.Context, addr string, tlsEnabled bool, component string, extraOpts ...grpc.DialOption) (*grpc.ClientConn, error) {
 	transportOption := grpc.WithTransportCredentials(insecure.NewCredentials())
 	// for js, the outer websocket layer takes care of tls
 	if tlsEnabled && runtime.GOOS != "js" {
@@ -46,9 +46,7 @@ func CreateConnection(ctx context.Context, addr string, tlsEnabled bool, compone
 	connCtx, cancel := context.WithTimeout(ctx, 30*time.Second)
 	defer cancel()
 
-	conn, err := grpc.DialContext(
-		connCtx,
-		addr,
+	opts := []grpc.DialOption{
 		transportOption,
 		WithCustomDialer(tlsEnabled, component),
 		grpc.WithBlock(),
@@ -56,7 +54,10 @@ func CreateConnection(ctx context.Context, addr string, tlsEnabled bool, compone
 			Time:    30 * time.Second,
 			Timeout: 10 * time.Second,
 		}),
-	)
+	}
+	opts = append(opts, extraOpts...)
+
+	conn, err := grpc.DialContext(connCtx, addr, opts...)
 	if err != nil {
 		return nil, fmt.Errorf("dial context: %w", err)
 	}

client/iface/configurer/uapi.go

@@ -5,20 +5,18 @@ package configurer
 import (
 	"net"
 
-	log "github.com/sirupsen/logrus"
 	"golang.zx2c4.com/wireguard/ipc"
 )
 
 func openUAPI(deviceName string) (net.Listener, error) {
 	uapiSock, err := ipc.UAPIOpen(deviceName)
 	if err != nil {
-		log.Errorf("failed to open uapi socket: %v", err)
 		return nil, err
 	}
 
 	listener, err := ipc.UAPIListen(deviceName, uapiSock)
 	if err != nil {
-		log.Errorf("failed to listen on uapi socket: %v", err)
+		_ = uapiSock.Close()
 		return nil, err
 	}
 

client/iface/configurer/usp.go

@@ -54,6 +54,14 @@ func NewUSPConfigurer(device *device.Device, deviceName string, activityRecorder
 	return wgCfg
 }
 
+func NewUSPConfigurerNoUAPI(device *device.Device, deviceName string, activityRecorder *bind.ActivityRecorder) *WGUSPConfigurer {
+	return &WGUSPConfigurer{
+		device:           device,
+		deviceName:       deviceName,
+		activityRecorder: activityRecorder,
+	}
+}
+
 func (c *WGUSPConfigurer) ConfigureInterface(privateKey string, port int) error {
 	log.Debugf("adding Wireguard private key")
 	key, err := wgtypes.ParseKey(privateKey)
@@ -111,7 +119,7 @@ func (c *WGUSPConfigurer) UpdatePeer(peerKey string, allowedIps []netip.Prefix,
 		if err != nil {
 			return fmt.Errorf("failed to parse endpoint address: %w", err)
 		}
-		addrPort := netip.AddrPortFrom(addr, uint16(endpoint.Port))
+		addrPort := netip.AddrPortFrom(addr.Unmap(), uint16(endpoint.Port))
 		c.activityRecorder.UpsertAddress(peerKey, addrPort)
 	}
 	return nil

client/iface/device/adapter.go

@@ -2,7 +2,7 @@ package device
 
 // TunAdapter is an interface for create tun device from external service
 type TunAdapter interface {
-	ConfigureInterface(address string, mtu int, dns string, searchDomains string, routes string) (int, error)
+	ConfigureInterface(address string, addressV6 string, mtu int, dns string, searchDomains string, routes string) (int, error)
 	UpdateAddr(address string) error
 	ProtectSocket(fd int32) bool
 }

client/iface/device/device_android.go

@@ -63,7 +63,7 @@ func (t *WGTunDevice) Create(routes []string, dns string, searchDomains []string
 		searchDomainsToString = ""
 	}
 
-	fd, err := t.tunAdapter.ConfigureInterface(t.address.String(), int(t.mtu), dns, searchDomainsToString, routesString)
+	fd, err := t.tunAdapter.ConfigureInterface(t.address.String(), t.address.IPv6String(), int(t.mtu), dns, searchDomainsToString, routesString)
 	if err != nil {
 		log.Errorf("failed to create Android interface: %s", err)
 		return nil, err

client/iface/device/device_darwin.go

@@ -131,23 +131,32 @@ func (t *TunDevice) Device() *device.Device {
 
 // assignAddr Adds IP address to the tunnel interface and network route based on the range provided
 func (t *TunDevice) assignAddr() error {
-	cmd := exec.Command("ifconfig", t.name, "inet", t.address.IP.String(), t.address.IP.String())
-	if out, err := cmd.CombinedOutput(); err != nil {
-		log.Errorf("adding address command '%v' failed with output: %s", cmd.String(), out)
-		return err
+	if out, err := exec.Command("ifconfig", t.name, "inet", t.address.IP.String(), t.address.IP.String()).CombinedOutput(); err != nil {
+		return fmt.Errorf("add v4 address: %s: %w", string(out), err)
 	}
 
-	// dummy ipv6 so routing works
-	cmd = exec.Command("ifconfig", t.name, "inet6", "fe80::/64")
-	if out, err := cmd.CombinedOutput(); err != nil {
-		log.Debugf("adding address command '%v' failed with output: %s", cmd.String(), out)
+	// Assign a dummy link-local so macOS enables IPv6 on the tun device.
+	// When a real overlay v6 is present, use that instead.
+	v6Addr := "fe80::/64"
+	if t.address.HasIPv6() {
+		v6Addr = t.address.IPv6String()
+	}
+	if out, err := exec.Command("ifconfig", t.name, "inet6", v6Addr).CombinedOutput(); err != nil {
+		log.Warnf("failed to assign IPv6 address %s, continuing v4-only: %s: %v", v6Addr, string(out), err)
+		t.address.ClearIPv6()
 	}
 
-	routeCmd := exec.Command("route", "add", "-net", t.address.Network.String(), "-interface", t.name)
-	if out, err := routeCmd.CombinedOutput(); err != nil {
-		log.Errorf("adding route command '%v' failed with output: %s", routeCmd.String(), out)
-		return err
+	if out, err := exec.Command("route", "add", "-net", t.address.Network.String(), "-interface", t.name).CombinedOutput(); err != nil {
+		return fmt.Errorf("add route %s via %s: %s: %w", t.address.Network, t.name, string(out), err)
 	}
+
+	if t.address.HasIPv6() {
+		if out, err := exec.Command("route", "add", "-inet6", "-net", t.address.IPv6Net.String(), "-interface", t.name).CombinedOutput(); err != nil {
+			log.Warnf("failed to add route %s via %s, continuing v4-only: %s: %v", t.address.IPv6Net, t.name, string(out), err)
+			t.address.ClearIPv6()
+		}
+	}
+
 	return nil
 }
 

client/iface/device/device_filter.go

@@ -15,14 +15,17 @@ type PacketFilter interface {
 	// FilterInbound filter incoming packets from external sources to host
 	FilterInbound(packetData []byte, size int) bool
 
-	// AddUDPPacketHook calls hook when UDP packet from given direction matched
-	//
-	// Hook function returns flag which indicates should be the matched package dropped or not.
-	// Hook function receives raw network packet data as argument.
-	AddUDPPacketHook(in bool, ip netip.Addr, dPort uint16, hook func(packet []byte) bool) string
+	// SetUDPPacketHook registers a hook for outbound UDP packets matching the given IP and port.
+	// Hook function returns true if the packet should be dropped.
+	// Only one UDP hook is supported; calling again replaces the previous hook.
+	// Pass nil hook to remove.
+	SetUDPPacketHook(ip netip.Addr, dPort uint16, hook func(packet []byte) bool)
 
-	// RemovePacketHook removes hook by ID
-	RemovePacketHook(hookID string) error
+	// SetTCPPacketHook registers a hook for outbound TCP packets matching the given IP and port.
+	// Hook function returns true if the packet should be dropped.
+	// Only one TCP hook is supported; calling again replaces the previous hook.
+	// Pass nil hook to remove.
+	SetTCPPacketHook(ip netip.Addr, dPort uint16, hook func(packet []byte) bool)
 }
 
 // FilteredDevice to override Read or Write of packets

client/iface/device/device_ios.go

@@ -151,8 +151,11 @@ func (t *TunDevice) MTU() uint16 {
 	return t.mtu
 }
 
-func (t *TunDevice) UpdateAddr(_ wgaddr.Address) error {
-	// todo implement
+// UpdateAddr updates the device address. On iOS the tunnel is managed by the
+// NetworkExtension, so we only store the new value. The extension picks up the
+// change on the next tunnel reconfiguration.
+func (t *TunDevice) UpdateAddr(addr wgaddr.Address) error {
+	t.address = addr
 	return nil
 }
 

client/iface/device/device_kernel_unix.go

@@ -173,7 +173,7 @@ func (t *TunKernelDevice) FilteredDevice() *FilteredDevice {
 
 // assignAddr Adds IP address to the tunnel interface
 func (t *TunKernelDevice) assignAddr() error {
-	return t.link.assignAddr(t.address)
+	return t.link.assignAddr(&t.address)
 }
 
 func (t *TunKernelDevice) GetNet() *netstack.Net {

client/iface/device/device_netstack.go

@@ -3,6 +3,7 @@ package device
 import (
 	"errors"
 	"fmt"
+	"net/netip"
 
 	log "github.com/sirupsen/logrus"
 	"golang.zx2c4.com/wireguard/conn"
@@ -63,8 +64,12 @@ func (t *TunNetstackDevice) create() (WGConfigurer, error) {
 		return nil, fmt.Errorf("last ip: %w", err)
 	}
 
-	log.Debugf("netstack using address: %s", t.address.IP)
-	t.nsTun = nbnetstack.NewNetStackTun(t.listenAddress, t.address.IP, dnsAddr, int(t.mtu))
+	addresses := []netip.Addr{t.address.IP}
+	if t.address.HasIPv6() {
+		addresses = append(addresses, t.address.IPv6)
+	}
+	log.Debugf("netstack using addresses: %v", addresses)
+	t.nsTun = nbnetstack.NewNetStackTun(t.listenAddress, addresses, dnsAddr, int(t.mtu))
 	log.Debugf("netstack using dns address: %s", dnsAddr)
 	tunIface, net, err := t.nsTun.Create()
 	if err != nil {
@@ -79,7 +84,7 @@ func (t *TunNetstackDevice) create() (WGConfigurer, error) {
 		device.NewLogger(wgLogLevel(), "[netbird] "),
 	)
 
-	t.configurer = configurer.NewUSPConfigurer(t.device, t.name, t.bind.ActivityRecorder())
+	t.configurer = configurer.NewUSPConfigurerNoUAPI(t.device, t.name, t.bind.ActivityRecorder())
 	err = t.configurer.ConfigureInterface(t.key, t.port)
 	if err != nil {
 		if cErr := tunIface.Close(); cErr != nil {

client/iface/device/device_usp_unix.go

@@ -16,7 +16,7 @@ import (
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 )
 
-type USPDevice struct {
+type TunDevice struct {
 	name    string
 	address wgaddr.Address
 	port    int
@@ -30,10 +30,10 @@ type USPDevice struct {
 	configurer     WGConfigurer
 }
 
-func NewUSPDevice(name string, address wgaddr.Address, port int, key string, mtu uint16, iceBind *bind.ICEBind) *USPDevice {
+func NewTunDevice(name string, address wgaddr.Address, port int, key string, mtu uint16, iceBind *bind.ICEBind) *TunDevice {
 	log.Infof("using userspace bind mode")
 
-	return &USPDevice{
+	return &TunDevice{
 		name:    name,
 		address: address,
 		port:    port,
@@ -43,7 +43,7 @@ func NewUSPDevice(name string, address wgaddr.Address, port int, key string, mtu
 	}
 }
 
-func (t *USPDevice) Create() (WGConfigurer, error) {
+func (t *TunDevice) Create() (WGConfigurer, error) {
 	log.Info("create tun interface")
 	tunIface, err := tun.CreateTUN(t.name, int(t.mtu))
 	if err != nil {
@@ -75,7 +75,7 @@ func (t *USPDevice) Create() (WGConfigurer, error) {
 	return t.configurer, nil
 }
 
-func (t *USPDevice) Up() (*udpmux.UniversalUDPMuxDefault, error) {
+func (t *TunDevice) Up() (*udpmux.UniversalUDPMuxDefault, error) {
 	if t.device == nil {
 		return nil, fmt.Errorf("device is not ready yet")
 	}
@@ -95,12 +95,12 @@ func (t *USPDevice) Up() (*udpmux.UniversalUDPMuxDefault, error) {
 	return udpMux, nil
 }
 
-func (t *USPDevice) UpdateAddr(address wgaddr.Address) error {
+func (t *TunDevice) UpdateAddr(address wgaddr.Address) error {
 	t.address = address
 	return t.assignAddr()
 }
 
-func (t *USPDevice) Close() error {
+func (t *TunDevice) Close() error {
 	if t.configurer != nil {
 		t.configurer.Close()
 	}
@@ -115,39 +115,39 @@ func (t *USPDevice) Close() error {
 	return nil
 }
 
-func (t *USPDevice) WgAddress() wgaddr.Address {
+func (t *TunDevice) WgAddress() wgaddr.Address {
 	return t.address
 }
 
-func (t *USPDevice) MTU() uint16 {
+func (t *TunDevice) MTU() uint16 {
 	return t.mtu
 }
 
-func (t *USPDevice) DeviceName() string {
+func (t *TunDevice) DeviceName() string {
 	return t.name
 }
 
-func (t *USPDevice) FilteredDevice() *FilteredDevice {
+func (t *TunDevice) FilteredDevice() *FilteredDevice {
 	return t.filteredDevice
 }
 
 // Device returns the wireguard device
-func (t *USPDevice) Device() *device.Device {
+func (t *TunDevice) Device() *device.Device {
 	return t.device
 }
 
 // assignAddr Adds IP address to the tunnel interface
-func (t *USPDevice) assignAddr() error {
+func (t *TunDevice) assignAddr() error {
 	link := newWGLink(t.name)
 
-	return link.assignAddr(t.address)
+	return link.assignAddr(&t.address)
 }
 
-func (t *USPDevice) GetNet() *netstack.Net {
+func (t *TunDevice) GetNet() *netstack.Net {
 	return nil
 }
 
 // GetICEBind returns the ICEBind instance
-func (t *USPDevice) GetICEBind() EndpointManager {
+func (t *TunDevice) GetICEBind() EndpointManager {
 	return t.iceBind
 }

client/iface/device/device_windows.go

@@ -87,7 +87,19 @@ func (t *TunDevice) Create() (WGConfigurer, error) {
 	err = nbiface.Set()
 	if err != nil {
 		t.device.Close()
-		return nil, fmt.Errorf("got error when getting setting the interface mtu: %s", err)
+		return nil, fmt.Errorf("set IPv4 interface MTU: %s", err)
+	}
+
+	if t.address.HasIPv6() {
+		nbiface6, err := luid.IPInterface(windows.AF_INET6)
+		if err != nil {
+			log.Warnf("failed to get IPv6 interface for MTU: %v", err)
+		} else {
+			nbiface6.NLMTU = uint32(t.mtu)
+			if err := nbiface6.Set(); err != nil {
+				log.Warnf("failed to set IPv6 interface MTU: %v", err)
+			}
+		}
 	}
 	err = t.assignAddr()
 	if err != nil {
@@ -178,8 +190,21 @@ func (t *TunDevice) GetInterfaceGUIDString() (string, error) {
 // assignAddr Adds IP address to the tunnel interface and network route based on the range provided
 func (t *TunDevice) assignAddr() error {
 	luid := winipcfg.LUID(t.nativeTunDevice.LUID())
-	log.Debugf("adding address %s to interface: %s", t.address.IP, t.name)
-	return luid.SetIPAddresses([]netip.Prefix{netip.MustParsePrefix(t.address.String())})
+
+	v4Prefix := t.address.Prefix()
+	if t.address.HasIPv6() {
+		v6Prefix := t.address.IPv6Prefix()
+		log.Debugf("adding addresses %s, %s to interface: %s", v4Prefix, v6Prefix, t.name)
+		if err := luid.SetIPAddresses([]netip.Prefix{v4Prefix, v6Prefix}); err != nil {
+			log.Warnf("failed to assign dual-stack addresses, retrying v4-only: %v", err)
+			t.address.ClearIPv6()
+			return luid.SetIPAddresses([]netip.Prefix{v4Prefix})
+		}
+		return nil
+	}
+
+	log.Debugf("adding address %s to interface: %s", v4Prefix, t.name)
+	return luid.SetIPAddresses([]netip.Prefix{v4Prefix})
 }
 
 func (t *TunDevice) GetNet() *netstack.Net {

client/iface/device/kernel_module.go

@@ -1,8 +0,0 @@
-//go:build (!linux && !freebsd) || android
-
-package device
-
-// WireGuardModuleIsLoaded check if we can load WireGuard mod (linux only)
-func WireGuardModuleIsLoaded() bool {
-	return false
-}

client/iface/device/kernel_module_freebsd.go

@@ -1,18 +0,0 @@
-package device
-
-// WireGuardModuleIsLoaded check if kernel support wireguard
-func WireGuardModuleIsLoaded() bool {
-	// Despite the fact FreeBSD natively support Wireguard (https://github.com/WireGuard/wireguard-freebsd)
-	//  we are currently do not use it, since it is required to add wireguard kernel support to
-	//   - https://github.com/netbirdio/netbird/tree/main/sharedsock
-	//   - https://github.com/mdlayher/socket
-	// TODO: implement kernel space
-	return false
-}
-
-// ModuleTunIsLoaded check if tun module exist, if is not attempt to load it
-func ModuleTunIsLoaded() bool {
-	// Assume tun supported by freebsd kernel by default
-	// TODO: implement check for module loaded in kernel or build-it
-	return true
-}

client/iface/device/kernel_module_nonlinux.go

@@ -0,0 +1,13 @@
+//go:build !linux || android
+
+package device
+
+// WireGuardModuleIsLoaded reports whether the kernel WireGuard module is available.
+func WireGuardModuleIsLoaded() bool {
+	return false
+}
+
+// ModuleTunIsLoaded reports whether the tun device is available.
+func ModuleTunIsLoaded() bool {
+	return true
+}

client/iface/device/wg_link_freebsd.go

@@ -2,6 +2,7 @@ package device
 
 import (
 	"fmt"
+	"os/exec"
 
 	log "github.com/sirupsen/logrus"
 
@@ -57,32 +58,32 @@ func (l *wgLink) up() error {
 	return nil
 }
 
-func (l *wgLink) assignAddr(address wgaddr.Address) error {
+func (l *wgLink) assignAddr(address *wgaddr.Address) error {
 	link, err := freebsd.LinkByName(l.name)
 	if err != nil {
 		return fmt.Errorf("link by name: %w", err)
 	}
 
-	ip := address.IP.String()
-
-	// Convert prefix length to hex netmask
 	prefixLen := address.Network.Bits()
-	if !address.IP.Is4() {
-		return fmt.Errorf("IPv6 not supported for interface assignment")
-	}
-
 	maskBits := uint32(0xffffffff) << (32 - prefixLen)
 	mask := fmt.Sprintf("0x%08x", maskBits)
 
-	log.Infof("assign addr %s mask %s to %s interface", ip, mask, l.name)
+	log.Infof("assign addr %s mask %s to %s interface", address.IP, mask, l.name)
 
-	err = link.AssignAddr(ip, mask)
-	if err != nil {
+	if err := link.AssignAddr(address.IP.String(), mask); err != nil {
 		return fmt.Errorf("assign addr: %w", err)
 	}
 
-	err = link.Up()
-	if err != nil {
+	if address.HasIPv6() {
+		log.Infof("assign IPv6 addr %s to %s interface", address.IPv6String(), l.name)
+		cmd := exec.Command("ifconfig", l.name, "inet6", address.IPv6String())
+		if out, err := cmd.CombinedOutput(); err != nil {
+			log.Warnf("failed to assign IPv6 address %s to %s, continuing v4-only: %s: %v", address.IPv6String(), l.name, string(out), err)
+			address.ClearIPv6()
+		}
+	}
+
+	if err := link.Up(); err != nil {
 		return fmt.Errorf("up: %w", err)
 	}
 

client/iface/device/wg_link_linux.go

@@ -4,6 +4,8 @@ package device
 
 import (
 	"fmt"
+	"net"
+	"net/netip"
 	"os"
 
 	log "github.com/sirupsen/logrus"
@@ -92,7 +94,7 @@ func (l *wgLink) up() error {
 	return nil
 }
 
-func (l *wgLink) assignAddr(address wgaddr.Address) error {
+func (l *wgLink) assignAddr(address *wgaddr.Address) error {
 	//delete existing addresses
 	list, err := netlink.AddrList(l, 0)
 	if err != nil {
@@ -110,20 +112,16 @@ func (l *wgLink) assignAddr(address wgaddr.Address) error {
 	}
 
 	name := l.attrs.Name
-	addrStr := address.String()
 
-	log.Debugf("adding address %s to interface: %s", addrStr, name)
-
-	addr, err := netlink.ParseAddr(addrStr)
-	if err != nil {
-		return fmt.Errorf("parse addr: %w", err)
+	if err := l.addAddr(name, address.Prefix()); err != nil {
+		return err
 	}
 
-	err = netlink.AddrAdd(l, addr)
-	if os.IsExist(err) {
-		log.Infof("interface %s already has the address: %s", name, addrStr)
-	} else if err != nil {
-		return fmt.Errorf("add addr: %w", err)
+	if address.HasIPv6() {
+		if err := l.addAddr(name, address.IPv6Prefix()); err != nil {
+			log.Warnf("failed to assign IPv6 address %s to %s, continuing v4-only: %v", address.IPv6Prefix(), name, err)
+			address.ClearIPv6()
+		}
 	}
 
 	// On linux, the link must be brought up
@@ -133,3 +131,22 @@ func (l *wgLink) assignAddr(address wgaddr.Address) error {
 
 	return nil
 }
+
+func (l *wgLink) addAddr(ifaceName string, prefix netip.Prefix) error {
+	log.Debugf("adding address %s to interface: %s", prefix, ifaceName)
+
+	addr := &netlink.Addr{
+		IPNet: &net.IPNet{
+			IP:   prefix.Addr().AsSlice(),
+			Mask: net.CIDRMask(prefix.Bits(), prefix.Addr().BitLen()),
+		},
+	}
+
+	if err := netlink.AddrAdd(l, addr); os.IsExist(err) {
+		log.Infof("interface %s already has the address: %s", ifaceName, prefix)
+	} else if err != nil {
+		return fmt.Errorf("add addr %s: %w", prefix, err)
+	}
+
+	return nil
+}

client/iface/iface.go

@@ -57,7 +57,7 @@ type wgProxyFactory interface {
 
 type WGIFaceOpts struct {
 	IFaceName    string
-	Address      string
+	Address      wgaddr.Address
 	WGPort       int
 	WGPrivKey    string
 	MTU          uint16
@@ -141,16 +141,11 @@ func (w *WGIface) Up() (*udpmux.UniversalUDPMuxDefault, error) {
 }
 
 // UpdateAddr updates address of the interface
-func (w *WGIface) UpdateAddr(newAddr string) error {
+func (w *WGIface) UpdateAddr(newAddr wgaddr.Address) error {
 	w.mu.Lock()
 	defer w.mu.Unlock()
 
-	addr, err := wgaddr.ParseWGAddress(newAddr)
-	if err != nil {
-		return err
-	}
-
-	return w.tun.UpdateAddr(addr)
+	return w.tun.UpdateAddr(newAddr)
 }
 
 // UpdatePeer updates existing Wireguard Peer or creates a new one if doesn't exist

client/iface/iface_new.go

@@ -1,33 +1,28 @@
+//go:build !linux && !ios && !android && !js
+
 package iface
 
 import (
 	"github.com/netbirdio/netbird/client/iface/bind"
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/netstack"
-	wgaddr "github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/iface/wgproxy"
 )
 
 // NewWGIFace Creates a new WireGuard interface instance
 func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
+	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, opts.Address, opts.MTU)
 
 	var tun WGTunDevice
 	if netstack.IsEnabled() {
-		tun = device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr())
+		tun = device.NewNetstackDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr())
 	} else {
-		tun = device.NewTunDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind)
+		tun = device.NewTunDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind)
 	}
 
-	wgIFace := &WGIface{
+	return &WGIface{
 		userspaceBind:  true,
 		tun:            tun,
 		wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
-	}
-	return wgIFace, nil
-
+	}, nil
 }

client/iface/iface_new_android.go

@@ -4,23 +4,17 @@ import (
 	"github.com/netbirdio/netbird/client/iface/bind"
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/netstack"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/iface/wgproxy"
 )
 
 // NewWGIFace Creates a new WireGuard interface instance
 func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
-	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
+	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, opts.Address, opts.MTU)
 
 	if netstack.IsEnabled() {
 		wgIFace := &WGIface{
 			userspaceBind:  true,
-			tun:            device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr()),
+			tun:            device.NewNetstackDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr()),
 			wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
 		}
 		return wgIFace, nil
@@ -28,7 +22,7 @@ func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
 
 	wgIFace := &WGIface{
 		userspaceBind:  true,
-		tun:            device.NewTunDevice(wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, opts.MobileArgs.TunAdapter, opts.DisableDNS),
+		tun:            device.NewTunDevice(opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, opts.MobileArgs.TunAdapter, opts.DisableDNS),
 		wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
 	}
 	return wgIFace, nil

client/iface/iface_new_darwin.go

@@ -1,35 +0,0 @@
-//go:build !ios
-
-package iface
-
-import (
-	"github.com/netbirdio/netbird/client/iface/bind"
-	"github.com/netbirdio/netbird/client/iface/device"
-	"github.com/netbirdio/netbird/client/iface/netstack"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
-	"github.com/netbirdio/netbird/client/iface/wgproxy"
-)
-
-// NewWGIFace Creates a new WireGuard interface instance
-func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
-	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
-
-	var tun WGTunDevice
-	if netstack.IsEnabled() {
-		tun = device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr())
-	} else {
-		tun = device.NewTunDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind)
-	}
-
-	wgIFace := &WGIface{
-		userspaceBind:  true,
-		tun:            tun,
-		wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
-	}
-	return wgIFace, nil
-}

client/iface/iface_new_freebsd.go

@@ -1,41 +0,0 @@
-//go:build freebsd
-
-package iface
-
-import (
-	"fmt"
-
-	"github.com/netbirdio/netbird/client/iface/bind"
-	"github.com/netbirdio/netbird/client/iface/device"
-	"github.com/netbirdio/netbird/client/iface/netstack"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
-	"github.com/netbirdio/netbird/client/iface/wgproxy"
-)
-
-// NewWGIFace Creates a new WireGuard interface instance
-func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
-	wgIFace := &WGIface{}
-
-	if netstack.IsEnabled() {
-		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
-		wgIFace.tun = device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr())
-		wgIFace.userspaceBind = true
-		wgIFace.wgProxyFactory = wgproxy.NewUSPFactory(iceBind, opts.MTU)
-		return wgIFace, nil
-	}
-
-	if device.ModuleTunIsLoaded() {
-		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
-		wgIFace.tun = device.NewUSPDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind)
-		wgIFace.userspaceBind = true
-		wgIFace.wgProxyFactory = wgproxy.NewUSPFactory(iceBind, opts.MTU)
-		return wgIFace, nil
-	}
-
-	return nil, fmt.Errorf("couldn't check or load tun module")
-}

client/iface/iface_new_ios.go

@@ -5,21 +5,15 @@ package iface
 import (
 	"github.com/netbirdio/netbird/client/iface/bind"
 	"github.com/netbirdio/netbird/client/iface/device"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/iface/wgproxy"
 )
 
 // NewWGIFace Creates a new WireGuard interface instance
 func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
-	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
+	iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, opts.Address, opts.MTU)
 
 	wgIFace := &WGIface{
-		tun:            device.NewTunDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, opts.MobileArgs.TunFd),
+		tun:            device.NewTunDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, opts.MobileArgs.TunFd),
 		userspaceBind:  true,
 		wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
 	}

client/iface/iface_new_js.go

@@ -4,21 +4,15 @@ import (
 	"github.com/netbirdio/netbird/client/iface/bind"
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/netstack"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/iface/wgproxy"
 )
 
 // NewWGIFace creates a new WireGuard interface for WASM (always uses netstack mode)
 func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
 	relayBind := bind.NewRelayBindJS()
 
 	wgIface := &WGIface{
-		tun:            device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, relayBind, netstack.ListenAddr()),
+		tun:            device.NewNetstackDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, relayBind, netstack.ListenAddr()),
 		userspaceBind:  true,
 		wgProxyFactory: wgproxy.NewUSPFactory(relayBind, opts.MTU),
 	}

client/iface/iface_new_linux.go

@@ -3,44 +3,40 @@
 package iface
 
 import (
-	"fmt"
+	"errors"
 
 	"github.com/netbirdio/netbird/client/iface/bind"
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/netstack"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/iface/wgproxy"
 )
 
 // NewWGIFace Creates a new WireGuard interface instance
 func NewWGIFace(opts WGIFaceOpts) (*WGIface, error) {
-	wgAddress, err := wgaddr.ParseWGAddress(opts.Address)
-	if err != nil {
-		return nil, err
-	}
-
-	wgIFace := &WGIface{}
-
 	if netstack.IsEnabled() {
-		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
-		wgIFace.tun = device.NewNetstackDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr())
-		wgIFace.userspaceBind = true
-		wgIFace.wgProxyFactory = wgproxy.NewUSPFactory(iceBind, opts.MTU)
-		return wgIFace, nil
+		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, opts.Address, opts.MTU)
+		return &WGIface{
+			tun:            device.NewNetstackDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind, netstack.ListenAddr()),
+			userspaceBind:  true,
+			wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
+		}, nil
 	}
 
 	if device.WireGuardModuleIsLoaded() {
-		wgIFace.tun = device.NewKernelDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, opts.TransportNet)
-		wgIFace.wgProxyFactory = wgproxy.NewKernelFactory(opts.WGPort, opts.MTU)
-		return wgIFace, nil
-	}
-	if device.ModuleTunIsLoaded() {
-		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, wgAddress, opts.MTU)
-		wgIFace.tun = device.NewUSPDevice(opts.IFaceName, wgAddress, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind)
-		wgIFace.userspaceBind = true
-		wgIFace.wgProxyFactory = wgproxy.NewUSPFactory(iceBind, opts.MTU)
-		return wgIFace, nil
+		return &WGIface{
+			tun:            device.NewKernelDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, opts.TransportNet),
+			wgProxyFactory: wgproxy.NewKernelFactory(opts.WGPort, opts.MTU),
+		}, nil
 	}
 
-	return nil, fmt.Errorf("couldn't check or load tun module")
+	if device.ModuleTunIsLoaded() {
+		iceBind := bind.NewICEBind(opts.TransportNet, opts.FilterFn, opts.Address, opts.MTU)
+		return &WGIface{
+			tun:            device.NewTunDevice(opts.IFaceName, opts.Address, opts.WGPort, opts.WGPrivKey, opts.MTU, iceBind),
+			userspaceBind:  true,
+			wgProxyFactory: wgproxy.NewUSPFactory(iceBind, opts.MTU),
+		}, nil
+	}
+
+	return nil, errors.New("tun module not available")
 }

client/iface/iface_test.go

@@ -16,6 +16,7 @@ import (
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
 
 	"github.com/netbirdio/netbird/client/iface/device"
+	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/internal/stdnet"
 )
 
@@ -48,7 +49,7 @@ func TestWGIface_UpdateAddr(t *testing.T) {
 
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      addr,
+		Address:      wgaddr.MustParseWGAddress(addr),
 		WGPort:       wgPort,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -84,7 +85,7 @@ func TestWGIface_UpdateAddr(t *testing.T) {
 
 	//update WireGuard address
 	addr = "100.64.0.2/8"
-	err = iface.UpdateAddr(addr)
+	err = iface.UpdateAddr(wgaddr.MustParseWGAddress(addr))
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -130,7 +131,7 @@ func Test_CreateInterface(t *testing.T) {
 	}
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      wgIP,
+		Address:      wgaddr.MustParseWGAddress(wgIP),
 		WGPort:       33100,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -174,7 +175,7 @@ func Test_Close(t *testing.T) {
 
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      wgIP,
+		Address:      wgaddr.MustParseWGAddress(wgIP),
 		WGPort:       wgPort,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -219,7 +220,7 @@ func TestRecreation(t *testing.T) {
 
 			opts := WGIFaceOpts{
 				IFaceName:    ifaceName,
-				Address:      wgIP,
+				Address:      wgaddr.MustParseWGAddress(wgIP),
 				WGPort:       wgPort,
 				WGPrivKey:    key,
 				MTU:          DefaultMTU,
@@ -291,7 +292,7 @@ func Test_ConfigureInterface(t *testing.T) {
 	}
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      wgIP,
+		Address:      wgaddr.MustParseWGAddress(wgIP),
 		WGPort:       wgPort,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -347,7 +348,7 @@ func Test_UpdatePeer(t *testing.T) {
 
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      wgIP,
+		Address:      wgaddr.MustParseWGAddress(wgIP),
 		WGPort:       33100,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -417,7 +418,7 @@ func Test_RemovePeer(t *testing.T) {
 
 	opts := WGIFaceOpts{
 		IFaceName:    ifaceName,
-		Address:      wgIP,
+		Address:      wgaddr.MustParseWGAddress(wgIP),
 		WGPort:       33100,
 		WGPrivKey:    key,
 		MTU:          DefaultMTU,
@@ -482,7 +483,7 @@ func Test_ConnectPeers(t *testing.T) {
 
 	optsPeer1 := WGIFaceOpts{
 		IFaceName:    peer1ifaceName,
-		Address:      peer1wgIP.String(),
+		Address:      wgaddr.MustParseWGAddress(peer1wgIP.String()),
 		WGPort:       peer1wgPort,
 		WGPrivKey:    peer1Key.String(),
 		MTU:          DefaultMTU,
@@ -522,7 +523,7 @@ func Test_ConnectPeers(t *testing.T) {
 
 	optsPeer2 := WGIFaceOpts{
 		IFaceName:    peer2ifaceName,
-		Address:      peer2wgIP.String(),
+		Address:      wgaddr.MustParseWGAddress(peer2wgIP.String()),
 		WGPort:       peer2wgPort,
 		WGPrivKey:    peer2Key.String(),
 		MTU:          DefaultMTU,