Fix TestResetAggregationWindow flakiness on coarse clocks

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

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

.github/workflows/frontend-ui.yml

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

.github/workflows/golangci-lint.yml

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

.goreleaser.yaml

@@ -273,8 +273,8 @@ dockers_v2:
        - netbirdio/netbird
        - ghcr.io/netbirdio/netbird
      tags:
-       - "{{ .Version }}-rootless"
-       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}rootless-latest{{ end }}"
+       - "v{{ .Version }}-rootless"
+       - "{{ if eq .Env.SKIP_PUBLISH \"false\" }}latest{{ end }}"
      dockerfile: client/Dockerfile-rootless
      extra_files:
        - client/netbird-entrypoint.sh

agent-network/README.md

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

client/android/client.go

@@ -247,9 +247,6 @@ func (c *Client) DebugBundle(platformFiles PlatformFiles, anonymize bool) (strin
 		deps.SyncResponse = resp
 
 		if e := cc.Engine(); e != nil {
-			deps.RefreshStatus = func() {
-				e.RunHealthProbes(context.Background(), true)
-			}
 			if cm := e.GetClientMetrics(); cm != nil {
 				deps.ClientMetrics = cm
 			}

client/cmd/login.go

@@ -17,9 +17,7 @@ import (
 	"github.com/netbirdio/netbird/client/internal"
 	"github.com/netbirdio/netbird/client/internal/auth"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
-	nbnet "github.com/netbirdio/netbird/client/net"
 	"github.com/netbirdio/netbird/client/proto"
-	"github.com/netbirdio/netbird/client/server"
 	"github.com/netbirdio/netbird/client/system"
 	"github.com/netbirdio/netbird/util"
 )
@@ -333,14 +331,6 @@ func doForegroundLogin(ctx context.Context, cmd *cobra.Command, setupKey string,
 		return fmt.Errorf("read config file %s: %v", configFilePath, err)
 	}
 
-	// Mirror runInForegroundMode: recover residual state (DNS, firewall,
-	// ssh config, legacy routing) from a previous unclean shutdown and
-	// enable advanced routing before dialing management.
-	if err := server.RestoreResidualState(ctx, profilemanager.NewServiceManager(configFilePath).GetStatePath()); err != nil {
-		log.Warnf("failed to restore residual state: %v", err)
-	}
-	nbnet.Init()
-
 	err = foregroundLogin(ctx, cmd, config, setupKey, activeProf.ID)
 	if err != nil {
 		return fmt.Errorf("foreground login failed: %v", err)

client/cmd/up.go

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

client/firewall/uspfilter/filter.go

@@ -121,7 +121,6 @@ type Manager struct {
 	udpTracker     *conntrack.UDPTracker
 	icmpTracker    *conntrack.ICMPTracker
 	tcpTracker     *conntrack.TCPTracker
-	fragments      *fragmentTracker
 	forwarder      atomic.Pointer[forwarder.Forwarder]
 	pendingCapture atomic.Pointer[forwarder.PacketCapture]
 	logger         *nblog.Logger
@@ -184,41 +183,6 @@ func (d *decoder) decodePacket(data []byte) error {
 	}
 }
 
-// decodeTransport decodes the transport header of a first fragment (which
-// gopacket leaves undecoded) into the decoder and appends its layer type to
-// decoded, so the ACL pipeline can evaluate it like a normal packet. It returns
-// false if the protocol is unsupported or the header is truncated.
-func (d *decoder) decodeTransport(proto layers.IPProtocol, payload []byte) bool {
-	var l4 gopacket.DecodingLayer
-	var layerType gopacket.LayerType
-	var minLen int
-	switch proto {
-	case layers.IPProtocolTCP:
-		l4, layerType, minLen = &d.tcp, layers.LayerTypeTCP, 20
-	case layers.IPProtocolUDP:
-		l4, layerType, minLen = &d.udp, layers.LayerTypeUDP, 8
-	case layers.IPProtocolICMPv4:
-		l4, layerType, minLen = &d.icmp4, layers.LayerTypeICMPv4, 8
-	case layers.IPProtocolICMPv6:
-		l4, layerType, minLen = &d.icmp6, layers.LayerTypeICMPv6, 8
-	default:
-		return false
-	}
-
-	// Reject a fragment too small to hold the full transport header before
-	// decoding: it can't be ACL-evaluated (tiny-fragment attack), and skipping
-	// the decode avoids gopacket allocating an error on the drop path.
-	if len(payload) < minLen {
-		return false
-	}
-
-	if err := l4.DecodeFromBytes(payload, gopacket.NilDecodeFeedback); err != nil {
-		return false
-	}
-	d.decoded = append(d.decoded, layerType)
-	return true
-}
-
 // 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)
@@ -322,8 +286,6 @@ func create(iface common.IFaceMapper, nativeFirewall firewall.Manager, disableSe
 	if err := m.localipmanager.UpdateLocalIPs(iface); err != nil {
 		return nil, fmt.Errorf("update local IPs: %w", err)
 	}
-	m.fragments = newFragmentTracker(m.logger)
-
 	if disableConntrack {
 		log.Info("conntrack is disabled")
 	} else {
@@ -337,7 +299,6 @@ func create(iface common.IFaceMapper, nativeFirewall firewall.Manager, disableSe
 		}
 	}
 	if err := iface.SetFilter(m); err != nil {
-		m.fragments.Close()
 		return nil, fmt.Errorf("set filter: %w", err)
 	}
 	return m, nil
@@ -733,10 +694,6 @@ func (m *Manager) resetState() {
 		m.tcpTracker.Close()
 	}
 
-	if m.fragments != nil {
-		m.fragments.Close()
-	}
-
 	if fwder := m.forwarder.Load(); fwder != nil {
 		fwder.SetCapture(nil)
 		fwder.Stop()
@@ -1089,20 +1046,19 @@ func (m *Manager) filterInbound(packetData []byte, size int) bool {
 		return true
 	}
 
-	// gopacket does not decode the transport header of any IP fragment, so
-	// fragments take a dedicated path: the first fragment's header is decoded
-	// and ACL-evaluated here, and the remaining fragments inherit its verdict.
+	// TODO: pass fragments of routed packets to forwarder
 	if fragment {
-		return m.filterInboundFragment(d, srcIP, dstIP, size)
+		if m.logger.Enabled(nblog.LevelTrace) {
+			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
 	}
 
-	return m.filterInboundDecoded(d, srcIP, dstIP, packetData, size)
-}
-
-// filterInboundDecoded runs the ACL, DNAT and conntrack pipeline on a fully
-// decoded (non-fragment) inbound packet. It returns true if the packet should
-// be dropped.
-func (m *Manager) filterInboundDecoded(d *decoder, srcIP, dstIP netip.Addr, packetData []byte, size int) bool {
 	// 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
@@ -1133,226 +1089,33 @@ func (m *Manager) filterInboundDecoded(d *decoder, srcIP, dstIP netip.Addr, pack
 	return m.handleRoutedTraffic(d, srcIP, dstIP, packetData, size)
 }
 
-// fragmentMeta holds the reassembly identity and layout of an IP fragment,
-// extracted uniformly for IPv4 and IPv6.
-type fragmentMeta struct {
-	key fragmentKey
-	// offset is the fragment offset in 8-byte units (zero for the first
-	// fragment).
-	offset uint16
-	// moreFragments is the More Fragments bit. A first fragment with it unset is
-	// an IPv6 atomic fragment (a complete datagram, RFC 6946): it has no trailing
-	// fragments to inherit a verdict, so it must not be recorded.
-	moreFragments bool
-	proto         layers.IPProtocol
-	// l4payload is the fragmentable payload of this fragment. For the first
-	// fragment it starts with the transport header.
-	l4payload []byte
-	// headerEndOctets is the first fragment's payload length in 8-byte units:
-	// the smallest offset a trailing fragment may start at without overlapping
-	// the inspected transport header.
-	headerEndOctets uint16
-}
-
-// fragmentMetadata extracts the fragment identity and layout from a decoded IP
-// fragment. It returns false for fragments it can't interpret (e.g. an IPv6
-// fragment header shorter than 8 bytes), which are then dropped.
-func fragmentMetadata(d *decoder, srcIP, dstIP netip.Addr) (fragmentMeta, bool) {
-	switch d.decoded[0] {
-	case layers.LayerTypeIPv4:
-		payload := d.ip4.Payload
-		return fragmentMeta{
-			key:             fragmentKey{srcIP: srcIP, dstIP: dstIP, id: uint32(d.ip4.Id), proto: uint8(d.ip4.Protocol)},
-			offset:          d.ip4.FragOffset,
-			moreFragments:   d.ip4.Flags&layers.IPv4MoreFragments != 0,
-			proto:           d.ip4.Protocol,
-			l4payload:       payload,
-			headerEndOctets: octets(len(payload)),
-		}, true
-
-	case layers.LayerTypeIPv6:
-		// IPv6 fragment extension header: 8 bytes, followed by the fragmentable
-		// payload. Layout: next header (1), reserved (1), offset+flags (2), id (4).
-		payload := d.ip6.Payload
-		if len(payload) < 8 {
-			return fragmentMeta{}, false
-		}
-		nextHeader := layers.IPProtocol(payload[0])
-		offsetFlags := binary.BigEndian.Uint16(payload[2:4])
-		id := binary.BigEndian.Uint32(payload[4:8])
-		l4 := payload[8:]
-		return fragmentMeta{
-			key:             fragmentKey{srcIP: srcIP, dstIP: dstIP, id: id, proto: uint8(nextHeader)},
-			offset:          offsetFlags >> 3,
-			moreFragments:   offsetFlags&1 != 0,
-			proto:           nextHeader,
-			l4payload:       l4,
-			headerEndOctets: octets(len(l4)),
-		}, true
-
-	default:
-		return fragmentMeta{}, false
-	}
-}
-
-// octets rounds a byte length up to whole 8-byte units, the granularity of the
-// IP fragment offset field.
-func octets(nbytes int) uint16 {
-	return uint16((nbytes + 7) / 8)
-}
-
-// filterInboundFragment decides the fate of an IP fragment. gopacket stops
-// decoding at the network layer for every fragment, so the first fragment's
-// transport header is decoded and ACL-evaluated here and its verdict recorded;
-// the remaining (headerless) fragments inherit that verdict. Anything that
-// cannot be tied to an allowed, non-overlapping first fragment is dropped.
-func (m *Manager) filterInboundFragment(d *decoder, srcIP, dstIP netip.Addr, size int) bool {
-	meta, ok := fragmentMetadata(d, srcIP, dstIP)
-	if !ok {
-		if m.logger.Enabled(nblog.LevelTrace) {
-			m.logger.Trace2("dropping unsupported fragment: src=%v dst=%v", srcIP, dstIP)
-		}
-		return true
-	}
-
-	if meta.offset != 0 {
-		return m.filterTrailingFragment(meta, srcIP, dstIP)
-	}
-
-	// A new first fragment supersedes any recorded verdict for this datagram, so
-	// a re-sent or overlapping offset-zero fragment can't inherit the old one.
-	m.fragments.poison(meta.key)
-
-	// First fragment: decode its transport header so the ACL can evaluate it. A
-	// decode failure means the fragment is too small to hold the full transport
-	// header (RFC 1858 §3 tiny-fragment attack); it can't be evaluated, so drop it.
-	if !d.decodeTransport(meta.proto, meta.l4payload) {
-		if m.logger.Enabled(nblog.LevelTrace) {
-			m.logger.Trace3("dropping first fragment without full L4 header: src=%v dst=%v id=%v",
-				srcIP, dstIP, meta.key.id)
-		}
-		return true
-	}
-
-	return m.filterFirstFragment(d, meta, srcIP, dstIP, size)
-}
-
-// filterTrailingFragment applies a recorded first-fragment verdict to a
-// non-first fragment.
-func (m *Manager) filterTrailingFragment(meta fragmentMeta, srcIP, dstIP netip.Addr) bool {
-	switch m.fragments.verdict(meta.key, meta.offset) {
-	case fragmentAllow:
-		return false
-	case fragmentOverlap:
-		if m.logger.Enabled(nblog.LevelTrace) {
-			m.logger.Trace3("dropping overlapping fragment rewriting inspected header: src=%v dst=%v id=%v",
-				srcIP, dstIP, meta.key.id)
-		}
-		return true
-	default:
-		if m.logger.Enabled(nblog.LevelTrace) {
-			m.logger.Trace3("dropping fragment with no allowed first fragment: src=%v dst=%v id=%v",
-				srcIP, dstIP, meta.key.id)
-		}
-		return true
-	}
-}
-
-// filterFirstFragment runs the verdict part of the inbound pipeline on a first
-// fragment with its transport header decoded. It mirrors filterInboundDecoded
-// but skips DNAT (port rewriting on fragments is unsupported) and forwarder
-// injection (fragments are left to the stack to reassemble, not forwarded).
-// Allowed fragments have their verdict recorded so the datagram's trailing
-// fragments inherit it.
-func (m *Manager) filterFirstFragment(d *decoder, meta fragmentMeta, srcIP, dstIP netip.Addr, size int) bool {
-	if m.stateful && m.isValidTrackedConnection(d, srcIP, dstIP, size) {
-		m.recordFirstFragment(meta)
-		return false
-	}
-
-	if m.localipmanager.IsLocalIP(dstIP) {
-		ruleID, blocked := m.peerACLsBlock(srcIP, d, nil)
-		if blocked {
-			m.storeDropFlow("Dropping local first fragment (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
-				d, srcIP, dstIP, ruleID, size)
-			return true
-		}
-		m.trackInbound(d, srcIP, dstIP, ruleID, size)
-		m.recordFirstFragment(meta)
-		return false
-	}
-
-	if !m.routingEnabled.Load() {
-		if m.logger.Enabled(nblog.LevelTrace) {
-			m.logger.Trace2("Dropping routed fragment (routing disabled): src=%s dst=%s", srcIP, dstIP)
-		}
-		return true
-	}
-	if m.nativeRouter.Load() {
-		m.trackInbound(d, srcIP, dstIP, nil, size)
-		m.recordFirstFragment(meta)
-		return false
-	}
-
-	// TODO: pass fragments of routed packets to the forwarder; until then
-	// allowed routed fragments go to the native stack.
-	srcPort, dstPort := getPortsFromPacket(d)
-	ruleID, pass := m.routeACLsPass(srcIP, dstIP, d.decoded[1], srcPort, dstPort)
-	if !pass {
-		m.storeDropFlow("Dropping routed first fragment (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
-			d, srcIP, dstIP, ruleID, size)
-		return true
-	}
-
-	m.recordFirstFragment(meta)
-	return false
-}
-
-// recordFirstFragment caches an allowed first fragment's verdict for its
-// trailing fragments to inherit. Atomic fragments (no More Fragments bit) are
-// complete datagrams with no trailing fragments, so they are not cached and
-// cannot exhaust the verdict table.
-func (m *Manager) recordFirstFragment(meta fragmentMeta) {
-	if !meta.moreFragments {
-		return
-	}
-	m.fragments.recordAllowed(meta.key, meta.headerEndOctets)
-}
-
-// storeDropFlow logs and records a netflow drop event for an inbound packet
-// denied by the ACLs. msg is the trace format taking rule id, protocol, source
-// and destination.
-func (m *Manager) storeDropFlow(msg string, d *decoder, srcIP, dstIP netip.Addr, ruleID []byte, size int) {
-	pnum := getProtocolFromPacket(d)
-	srcPort, dstPort := getPortsFromPacket(d)
-
-	if m.logger.Enabled(nblog.LevelTrace) {
-		m.logger.Trace6(msg, ruleID, pnum, srcIP, srcPort, dstIP, dstPort)
-	}
-
-	m.flowLogger.StoreEvent(nftypes.EventFields{
-		FlowID:     uuid.New(),
-		Type:       nftypes.TypeDrop,
-		RuleID:     ruleID,
-		Direction:  nftypes.Ingress,
-		Protocol:   pnum,
-		SourceIP:   srcIP,
-		DestIP:     dstIP,
-		SourcePort: srcPort,
-		DestPort:   dstPort,
-		// TODO: icmp type/code
-		RxPackets: 1,
-		RxBytes:   uint64(size),
-	})
-}
-
 // handleLocalTraffic handles local traffic.
 // If it returns true, the packet should be dropped.
 func (m *Manager) handleLocalTraffic(d *decoder, srcIP, dstIP netip.Addr, packetData []byte, size int) bool {
 	ruleID, blocked := m.peerACLsBlock(srcIP, d, packetData)
 	if blocked {
-		m.storeDropFlow("Dropping local packet (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
-			d, srcIP, dstIP, ruleID, size)
+		pnum := getProtocolFromPacket(d)
+		srcPort, dstPort := getPortsFromPacket(d)
+
+		if m.logger.Enabled(nblog.LevelTrace) {
+			m.logger.Trace6("Dropping local packet (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
+				ruleID, pnum, srcIP, srcPort, dstIP, dstPort)
+		}
+
+		m.flowLogger.StoreEvent(nftypes.EventFields{
+			FlowID:     uuid.New(),
+			Type:       nftypes.TypeDrop,
+			RuleID:     ruleID,
+			Direction:  nftypes.Ingress,
+			Protocol:   pnum,
+			SourceIP:   srcIP,
+			DestIP:     dstIP,
+			SourcePort: srcPort,
+			DestPort:   dstPort,
+			// TODO: icmp type/code
+			RxPackets: 1,
+			RxBytes:   uint64(size),
+		})
 		return true
 	}
 
@@ -1405,8 +1168,27 @@ func (m *Manager) handleRoutedTraffic(d *decoder, srcIP, dstIP netip.Addr, packe
 
 	ruleID, pass := m.routeACLsPass(srcIP, dstIP, protoLayer, srcPort, dstPort)
 	if !pass {
-		m.storeDropFlow("Dropping routed packet (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
-			d, srcIP, dstIP, ruleID, size)
+		proto := getProtocolFromPacket(d)
+
+		if m.logger.Enabled(nblog.LevelTrace) {
+			m.logger.Trace6("Dropping routed packet (ACL denied): rule_id=%s proto=%v src=%s:%d dst=%s:%d",
+				ruleID, proto, srcIP, srcPort, dstIP, dstPort)
+		}
+
+		m.flowLogger.StoreEvent(nftypes.EventFields{
+			FlowID:     uuid.New(),
+			Type:       nftypes.TypeDrop,
+			RuleID:     ruleID,
+			Direction:  nftypes.Ingress,
+			Protocol:   proto,
+			SourceIP:   srcIP,
+			DestIP:     dstIP,
+			SourcePort: srcPort,
+			DestPort:   dstPort,
+			// TODO: icmp type/code
+			RxPackets: 1,
+			RxBytes:   uint64(size),
+		})
 		return true
 	}
 

client/firewall/uspfilter/forwarder/forwarder.go

@@ -5,9 +5,7 @@ import (
 	"fmt"
 	"net"
 	"net/netip"
-	"os"
 	"runtime"
-	"strconv"
 	"sync"
 	"time"
 
@@ -33,11 +31,6 @@ const (
 	defaultMaxInFlight   = 1024
 	iosReceiveWindow     = 16384
 	iosMaxInFlight       = 256
-
-	// envForceTCPRACK overrides the platform default for gVisor's RACK loss
-	// detection. Set to a truthy value to force RACK on, or a falsy value to
-	// force it off, on any platform.
-	envForceTCPRACK = "NB_FORCE_TCP_RACK"
 )
 
 type Forwarder struct {
@@ -159,8 +152,6 @@ func New(iface common.IFaceMapper, logger *nblog.Logger, flowLogger nftypes.Flow
 		maxInFlight = iosMaxInFlight
 	}
 
-	configureTCPRecovery(s)
-
 	tcpForwarder := tcp.NewForwarder(s, receiveWindow, maxInFlight, f.handleTCP)
 	s.SetTransportProtocolHandler(tcp.ProtocolNumber, tcpForwarder.HandlePacket)
 
@@ -475,31 +466,3 @@ func probeRawICMP(network, addr string, logger *nblog.Logger) bool {
 	logger.Debug1("forwarder: raw %s socket access available", network)
 	return true
 }
-
-// configureTCPRecovery disables gVisor's RACK loss detection on Windows, where
-// it interacts poorly with the host and collapses throughput on routed TCP
-// connections (gVisor issue #9778). Other platforms keep the default. The
-// EnvForceTCPRACK environment variable overrides the platform default.
-func configureTCPRecovery(s *stack.Stack) {
-	disableRACK := runtime.GOOS == "windows"
-
-	if val := os.Getenv(envForceTCPRACK); val != "" {
-		force, err := strconv.ParseBool(val)
-		if err != nil {
-			log.Warnf("parse %s: %v", envForceTCPRACK, err)
-		} else {
-			disableRACK = !force
-		}
-	}
-
-	if !disableRACK {
-		return
-	}
-
-	opt := tcpip.TCPRecovery(0)
-	if err := s.SetTransportProtocolOption(tcp.ProtocolNumber, &opt); err != nil {
-		log.Warnf("disable TCP RACK loss detection: %v", err)
-		return
-	}
-	log.Info("forwarder: TCP RACK loss detection disabled")
-}

client/firewall/uspfilter/fragment.go

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

client/firewall/uspfilter/fragment_bench_test.go

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

client/firewall/uspfilter/fragment_test.go

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

client/iface/iface_test.go

@@ -464,8 +464,6 @@ func Test_RemovePeer(t *testing.T) {
 }
 
 func Test_ConnectPeers(t *testing.T) {
-	t.Setenv("NB_DISABLE_EBPF_WG_PROXY", "true")
-
 	peer1ifaceName := fmt.Sprintf("utun%d", WgIntNumber+400)
 	peer1wgIP := netip.MustParsePrefix("10.99.99.17/30")
 	peer1Key, _ := wgtypes.GeneratePrivateKey()
@@ -507,8 +505,12 @@ func Test_ConnectPeers(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	localIP1 := "127.0.0.1"
-	peer1endpoint, err := net.ResolveUDPAddr("udp", fmt.Sprintf("%s:%d", localIP1, peer1wgPort))
+	localIP, err := getLocalIP()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	peer1endpoint, err := net.ResolveUDPAddr("udp", fmt.Sprintf("%s:%d", localIP, peer1wgPort))
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -544,8 +546,7 @@ func Test_ConnectPeers(t *testing.T) {
 		t.Fatal(err)
 	}
 
-	localIP2 := "127.0.0.1"
-	peer2endpoint, err := net.ResolveUDPAddr("udp", fmt.Sprintf("%s:%d", localIP2, peer2wgPort))
+	peer2endpoint, err := net.ResolveUDPAddr("udp", fmt.Sprintf("%s:%d", localIP, peer2wgPort))
 	if err != nil {
 		t.Fatal(err)
 	}
@@ -568,17 +569,17 @@ func Test_ConnectPeers(t *testing.T) {
 	if err != nil {
 		t.Fatal(err)
 	}
-	// The peers use userspace WireGuard (stdnet transport). A tight busy-loop
-	// here starves the wireguard-go goroutines that process the handshake, so
-	// poll on a ticker instead and yield the CPU between checks. WireGuard also
-	// only retries a lost handshake initiation every REKEY_TIMEOUT (5s), which
-	// is why the overall wait can occasionally stretch to tens of seconds.
+	// todo: investigate why in some tests execution we need 30s
 	timeout := 30 * time.Second
 	timeoutChannel := time.After(timeout)
-	ticker := time.NewTicker(500 * time.Millisecond)
-	defer ticker.Stop()
 
 	for {
+		select {
+		case <-timeoutChannel:
+			t.Fatalf("waiting for peer handshake timeout after %s", timeout.String())
+		default:
+		}
+
 		peer, gpErr := getPeer(peer1ifaceName, peer2Key.PublicKey().String())
 		if gpErr != nil {
 			t.Fatal(gpErr)
@@ -587,12 +588,6 @@ func Test_ConnectPeers(t *testing.T) {
 			t.Log("peers successfully handshake")
 			break
 		}
-
-		select {
-		case <-timeoutChannel:
-			t.Fatalf("waiting for peer handshake timeout after %s", timeout.String())
-		case <-ticker.C:
-		}
 	}
 
 }
@@ -620,3 +615,28 @@ func getPeer(ifaceName, peerPubKey string) (wgtypes.Peer, error) {
 	}
 	return wgtypes.Peer{}, fmt.Errorf("peer not found")
 }
+
+func getLocalIP() (string, error) {
+	// Get all interfaces
+	addrs, err := net.InterfaceAddrs()
+	if err != nil {
+		return "", err
+	}
+
+	for _, addr := range addrs {
+		ipNet, ok := addr.(*net.IPNet)
+		if !ok {
+			continue
+		}
+		if ipNet.IP.IsLoopback() {
+			continue
+		}
+
+		if ipNet.IP.To4() == nil {
+			continue
+		}
+		return ipNet.IP.String(), nil
+	}
+
+	return "", fmt.Errorf("no local IP found")
+}

client/iface/netstack/env.go

@@ -3,31 +3,14 @@
 package netstack
 
 import (
-	"net"
+	"fmt"
 	"os"
 	"strconv"
 
 	log "github.com/sirupsen/logrus"
 )
 
-const (
-	EnvUseNetstackMode = "NB_USE_NETSTACK_MODE"
-
-	// EnvSocks5ListenerPort overrides the port the SOCKS5 proxy listens on.
-	EnvSocks5ListenerPort = "NB_SOCKS5_LISTENER_PORT"
-
-	// EnvSocks5ListenerAddress overrides the host/IP the SOCKS5 proxy binds to.
-	// The proxy is a bridge for local host applications into the userspace
-	// WireGuard netstack, so it binds to loopback by default. Override this only
-	// when the proxy must be reachable from other hosts (e.g. a container
-	// gateway); doing so exposes an unauthenticated SOCKS5 proxy on that
-	// address.
-	EnvSocks5ListenerAddress = "NB_SOCKS5_LISTENER_ADDRESS"
-
-	// defaultSocks5Host is the loopback address the SOCKS5 proxy binds to unless
-	// overridden via EnvSocks5ListenerAddress.
-	defaultSocks5Host = "127.0.0.1"
-)
+const EnvUseNetstackMode = "NB_USE_NETSTACK_MODE"
 
 // IsEnabled todo: move these function to cmd layer
 func IsEnabled() bool {
@@ -35,40 +18,24 @@ func IsEnabled() bool {
 }
 
 func ListenAddr() string {
-	return net.JoinHostPort(listenHost(), strconv.Itoa(listenPort()))
-}
-
-// listenHost returns the host/IP the SOCKS5 proxy binds to. It defaults to
-// loopback and only honors EnvSocks5ListenerAddress when it holds a valid IP.
-func listenHost() string {
-	addr := os.Getenv(EnvSocks5ListenerAddress)
-	if addr == "" {
-		return defaultSocks5Host
-	}
-	if net.ParseIP(addr) == nil {
-		log.Warnf("invalid socks5 listener address %q, falling back to default: %s", addr, defaultSocks5Host)
-		return defaultSocks5Host
-	}
-	return addr
-}
-
-// listenPort returns the port the SOCKS5 proxy binds to, defaulting to
-// DefaultSocks5Port when EnvSocks5ListenerPort is unset or invalid.
-func listenPort() int {
-	sPort := os.Getenv(EnvSocks5ListenerPort)
+	sPort := os.Getenv("NB_SOCKS5_LISTENER_PORT")
 	if sPort == "" {
-		return DefaultSocks5Port
+		return listenAddr(DefaultSocks5Port)
 	}
 
 	port, err := strconv.Atoi(sPort)
 	if err != nil {
 		log.Warnf("invalid socks5 listener port, unable to convert it to int, falling back to default: %d", DefaultSocks5Port)
-		return DefaultSocks5Port
+		return listenAddr(DefaultSocks5Port)
 	}
 	if port < 1 || port > 65535 {
 		log.Warnf("invalid socks5 listener port, it should be in the range 1-65535, falling back to default: %d", DefaultSocks5Port)
-		return DefaultSocks5Port
+		return listenAddr(DefaultSocks5Port)
 	}
 
-	return port
+	return listenAddr(port)
+}
+
+func listenAddr(port int) string {
+	return fmt.Sprintf("0.0.0.0:%d", port)
 }

client/iface/netstack/env_test.go

@@ -1,63 +0,0 @@
-//go:build !js
-
-package netstack
-
-import (
-	"net"
-	"strconv"
-	"testing"
-)
-
-func TestListenAddr_DefaultsToLoopback(t *testing.T) {
-	// No env overrides: must bind loopback, never all interfaces.
-	got := ListenAddr()
-	want := net.JoinHostPort("127.0.0.1", strconv.Itoa(DefaultSocks5Port))
-	if got != want {
-		t.Fatalf("ListenAddr() = %q, want %q", got, want)
-	}
-}
-
-func TestListenAddr_AddressOverride(t *testing.T) {
-	tests := []struct {
-		name string
-		env  string
-		want string
-	}{
-		{name: "valid override honored", env: "0.0.0.0", want: "0.0.0.0"},
-		{name: "valid specific ip honored", env: "10.0.0.5", want: "10.0.0.5"},
-		{name: "ipv6 loopback bracketed", env: "::1", want: "::1"},
-		{name: "invalid falls back to loopback", env: "not-an-ip", want: "127.0.0.1"},
-		{name: "empty falls back to loopback", env: "", want: "127.0.0.1"},
-	}
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			t.Setenv(EnvSocks5ListenerAddress, tc.env)
-			want := net.JoinHostPort(tc.want, strconv.Itoa(DefaultSocks5Port))
-			if got := ListenAddr(); got != want {
-				t.Fatalf("ListenAddr() = %q, want %q", got, want)
-			}
-		})
-	}
-}
-
-func TestListenAddr_PortOverride(t *testing.T) {
-	tests := []struct {
-		name string
-		env  string
-		want int
-	}{
-		{name: "valid port honored", env: "1081", want: 1081},
-		{name: "non-numeric falls back", env: "abc", want: DefaultSocks5Port},
-		{name: "out of range falls back", env: "70000", want: DefaultSocks5Port},
-		{name: "zero falls back", env: "0", want: DefaultSocks5Port},
-	}
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			t.Setenv(EnvSocks5ListenerPort, tc.env)
-			want := net.JoinHostPort("127.0.0.1", strconv.Itoa(tc.want))
-			if got := ListenAddr(); got != want {
-				t.Fatalf("ListenAddr() = %q, want %q", got, want)
-			}
-		})
-	}
-}

client/internal/auth/auth.go

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

client/internal/auth/device_flow.go

@@ -259,18 +259,12 @@ func (d *DeviceAuthorizationFlow) WaitToken(ctx context.Context, info AuthFlowIn
 	ticker := time.NewTicker(interval)
 	defer ticker.Stop()
 
-	log.Infof("device flow: waiting for user authorization, polling token endpoint every %s, code expires in %s", interval, timeout)
-
-	start := time.Now()
-	polls := 0
-
 	for {
 		select {
 		case <-waitCtx.Done():
 			return TokenInfo{}, waitCtx.Err()
 		case <-ticker.C:
 
-			polls++
 			tokenResponse, err := d.requestToken(info)
 			if err != nil {
 				return TokenInfo{}, fmt.Errorf("parsing token response failed with error: %v", err)
@@ -278,12 +272,10 @@ func (d *DeviceAuthorizationFlow) WaitToken(ctx context.Context, info AuthFlowIn
 
 			if tokenResponse.Error != "" {
 				if tokenResponse.Error == "authorization_pending" {
-					log.Tracef("device flow: authorization still pending after poll %d", polls)
 					continue
 				} else if tokenResponse.Error == "slow_down" {
 					interval += (3 * time.Second)
 					ticker.Reset(interval)
-					log.Infof("device flow: IdP requested slow_down, polling interval increased to %s", interval)
 					continue
 				}
 
@@ -299,12 +291,11 @@ func (d *DeviceAuthorizationFlow) WaitToken(ctx context.Context, info AuthFlowIn
 				UseIDToken:   d.providerConfig.UseIDToken,
 			}
 
-			err = validateTokenAudience(tokenInfo.GetTokenToUse(), d.providerConfig.Audience)
+			err = isValidAccessToken(tokenInfo.GetTokenToUse(), d.providerConfig.Audience)
 			if err != nil {
 				return TokenInfo{}, fmt.Errorf("validate access token failed with error: %v", err)
 			}
 
-			log.Infof("device flow: user authorization confirmed after %d polls in %s", polls, time.Since(start).Round(time.Second))
 			return tokenInfo, err
 		}
 	}

client/internal/auth/pkce_flow.go

@@ -188,8 +188,6 @@ func (p *PKCEAuthorizationFlow) WaitToken(ctx context.Context, info AuthFlowInfo
 	waitCtx, cancel := context.WithTimeout(ctx, timeout)
 	defer cancel()
 
-	log.Infof("pkce flow: waiting for authorization callback on %s, timeout %s", p.oAuthConfig.RedirectURL, timeout)
-
 	tokenChan := make(chan *oauth2.Token, 1)
 	errChan := make(chan error, 1)
 
@@ -223,7 +221,6 @@ func (p *PKCEAuthorizationFlow) WaitToken(ctx context.Context, info AuthFlowInfo
 func (p *PKCEAuthorizationFlow) startServer(server *http.Server, tokenChan chan<- *oauth2.Token, errChan chan<- error) {
 	mux := http.NewServeMux()
 	mux.HandleFunc("/", func(w http.ResponseWriter, req *http.Request) {
-		log.Infof("pkce flow: received authorization callback from IdP")
 		cert := p.providerConfig.ClientCertPair
 		if cert != nil {
 			tr := &http.Transport{
@@ -274,18 +271,11 @@ func (p *PKCEAuthorizationFlow) handleRequest(req *http.Request) (*oauth2.Token,
 		return nil, fmt.Errorf("authentication failed: missing code")
 	}
 
-	exchangeStart := time.Now()
-	token, err := p.oAuthConfig.Exchange(
+	return p.oAuthConfig.Exchange(
 		req.Context(),
 		code,
 		oauth2.SetAuthURLParam("code_verifier", p.codeVerifier),
 	)
-	if err != nil {
-		return nil, err
-	}
-
-	log.Infof("pkce flow: authorization code exchanged for token in %s", time.Since(exchangeStart).Round(time.Millisecond))
-	return token, nil
 }
 
 func (p *PKCEAuthorizationFlow) parseOAuthToken(token *oauth2.Token) (TokenInfo, error) {
@@ -306,7 +296,7 @@ func (p *PKCEAuthorizationFlow) parseOAuthToken(token *oauth2.Token) (TokenInfo,
 		audience = p.providerConfig.ClientID
 	}
 
-	if err := validateTokenAudience(tokenInfo.GetTokenToUse(), audience); err != nil {
+	if err := isValidAccessToken(tokenInfo.GetTokenToUse(), audience); err != nil {
 		return TokenInfo{}, fmt.Errorf("authentication failed: invalid access token - %w", err)
 	}
 
@@ -320,11 +310,6 @@ func (p *PKCEAuthorizationFlow) parseOAuthToken(token *oauth2.Token) (TokenInfo,
 	return tokenInfo, nil
 }
 
-// parseEmailFromIDToken extracts the email (or name) claim from an ID token
-// without verifying its signature. The value is best-effort and used only as a
-// UX convenience (login hint prefill and display); it never drives an
-// authorization decision. The authoritative identity is established server-side
-// from the signature-verified token.
 func parseEmailFromIDToken(token string) (string, error) {
 	parts := strings.Split(token, ".")
 	if len(parts) < 2 {

client/internal/auth/sessionwatch/watcher.go

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

client/internal/auth/sessionwatch/watcher_test.go

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

client/internal/auth/util.go

@@ -20,26 +20,14 @@ func randomBytesInHex(count int) (string, error) {
 	return hex.EncodeToString(buf), nil
 }
 
-// validateTokenAudience checks that the token is a well-formed JWT whose
-// audience claim matches the expected audience.
-//
-// It does NOT verify the token's cryptographic signature and therefore must not
-// be treated as an authenticity check. The token is obtained by the client
-// directly from the IdP token endpoint over TLS, and its signature is verified
-// server-side by the management server against the IdP's JWKS
-// (see shared/auth/jwt/validator.go). This function is only a client-side
-// sanity check that the returned token targets the expected audience.
-func validateTokenAudience(token string, audience string) error {
+// isValidAccessToken is a simple validation of the access token
+func isValidAccessToken(token string, audience string) error {
 	if token == "" {
 		return fmt.Errorf("token received is empty")
 	}
 
-	parts := strings.Split(token, ".")
-	if len(parts) != 3 {
-		return fmt.Errorf("token is not a well-formed JWT")
-	}
-
-	claimsString, err := base64.RawURLEncoding.DecodeString(parts[1])
+	encodedClaims := strings.Split(token, ".")[1]
+	claimsString, err := base64.RawURLEncoding.DecodeString(encodedClaims)
 	if err != nil {
 		return err
 	}

client/internal/auth/util_test.go

@@ -1,108 +0,0 @@
-package auth
-
-import (
-	"encoding/base64"
-	"encoding/json"
-	"testing"
-)
-
-// makeJWT builds an unsigned JWT-shaped string (header.payload.signature) with
-// the given claims payload. The signature part is arbitrary because
-// validateTokenAudience intentionally does not verify it.
-func makeJWT(t *testing.T, claims map[string]interface{}) string {
-	t.Helper()
-	header := base64.RawURLEncoding.EncodeToString([]byte(`{"alg":"RS256","typ":"JWT"}`))
-	payloadBytes, err := json.Marshal(claims)
-	if err != nil {
-		t.Fatalf("marshal claims: %v", err)
-	}
-	payload := base64.RawURLEncoding.EncodeToString(payloadBytes)
-	return header + "." + payload + ".unverified-signature"
-}
-
-func TestValidateTokenAudience(t *testing.T) {
-	tests := []struct {
-		name     string
-		token    string
-		audience string
-		wantErr  bool
-	}{
-		{
-			name:     "empty token",
-			token:    "",
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "not a JWT - no dots",
-			token:    "notajwt",
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "not a JWT - two parts only",
-			token:    "header.payload",
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "matching string audience",
-			token:    makeJWT(t, map[string]interface{}{"aud": "netbird"}),
-			audience: "netbird",
-			wantErr:  false,
-		},
-		{
-			name:     "mismatching string audience",
-			token:    makeJWT(t, map[string]interface{}{"aud": "other"}),
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "matching audience in array",
-			token:    makeJWT(t, map[string]interface{}{"aud": []interface{}{"other", "netbird"}}),
-			audience: "netbird",
-			wantErr:  false,
-		},
-		{
-			name:     "mismatching audience array",
-			token:    makeJWT(t, map[string]interface{}{"aud": []interface{}{"a", "b"}}),
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "missing audience claim",
-			token:    makeJWT(t, map[string]interface{}{"sub": "user"}),
-			audience: "netbird",
-			wantErr:  true,
-		},
-		{
-			name:     "invalid base64 payload",
-			token:    "header.!!!not-base64!!!.sig",
-			audience: "netbird",
-			wantErr:  true,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			err := validateTokenAudience(tc.token, tc.audience)
-			if tc.wantErr && err == nil {
-				t.Fatalf("expected error, got nil")
-			}
-			if !tc.wantErr && err != nil {
-				t.Fatalf("expected no error, got %v", err)
-			}
-		})
-	}
-}
-
-// TestValidateTokenAudienceNoPanic guards the regression where a non-empty
-// token without the JWT dot structure caused an index-out-of-range panic.
-func TestValidateTokenAudienceNoPanic(t *testing.T) {
-	inputs := []string{"a", ".", "a.", "aaaa", "no-dots-here"}
-	for _, in := range inputs {
-		if err := validateTokenAudience(in, "netbird"); err == nil {
-			t.Fatalf("expected error for malformed token %q, got nil", in)
-		}
-	}
-}

client/internal/conn_mgr.go

@@ -34,8 +34,6 @@ const (
 // - Handling connection establishment based on peer signaling
 //
 // The implementation is not thread-safe; it is protected by engine.syncMsgMux.
-// The only exception is ActivatePeer, which is safe for concurrent use so the
-// DNS warm-up path can call it without contending on the engine mutex.
 type ConnMgr struct {
 	peerStore        *peerstore.Store
 	statusRecorder   *peer.Status
@@ -44,26 +42,12 @@ type ConnMgr struct {
 	rosenpassEnabled bool
 
 	lazyConnMgr *manager.Manager
-	// lazyConnMgrMu guards the lazyConnMgr pointer for readers outside the
-	// engine loop (ActivatePeer). Writers hold it in addition to
-	// engine.syncMsgMux; all other reads stay under engine.syncMsgMux only.
-	lazyConnMgrMu sync.RWMutex
-
-	// reconcileRoutedIPs re-applies a peer's routed allowed IPs after its lazy wake endpoint is
-	// (re)armed (Mode A at arm time). Injected by the engine; nil disables the reconcile.
-	reconcileRoutedIPs func(peerKey string) error
 
 	wg            sync.WaitGroup
 	lazyCtx       context.Context
 	lazyCtxCancel context.CancelFunc
 }
 
-// SetRoutedIPsReconciler injects the callback used to re-apply a peer's routed allowed IPs when
-// its lazy wake endpoint is (re)armed. Must be called before the lazy manager starts.
-func (e *ConnMgr) SetRoutedIPsReconciler(fn func(peerKey string) error) {
-	e.reconcileRoutedIPs = fn
-}
-
 func NewConnMgr(engineConfig *EngineConfig, statusRecorder *peer.Status, peerStore *peerstore.Store, iface lazyconn.WGIface) *ConnMgr {
 	e := &ConnMgr{
 		peerStore:        peerStore,
@@ -125,7 +109,7 @@ func (e *ConnMgr) UpdatedRemoteFeatureFlag(ctx context.Context, enabled bool) er
 			return nil
 		}
 
-		log.Infof("lazy connection manager is enabled by the management feature flag")
+		log.Warnf("lazy connection manager is enabled by management feature flag")
 		e.initLazyManager(ctx)
 		e.statusRecorder.UpdateLazyConnection(true)
 		return e.addPeersToLazyConnManager()
@@ -254,20 +238,12 @@ func (e *ConnMgr) RemovePeerConn(peerKey string) {
 	conn.Log.Infof("removed peer from lazy conn manager")
 }
 
-// ActivatePeer wakes an idle lazy connection. Unlike the rest of ConnMgr it is
-// safe for concurrent use: the lazy manager pointer is read under lazyConnMgrMu
-// and the manager itself is internally synchronized, so callers outside the
-// engine loop (DNS warm-up) do not need engine.syncMsgMux.
 func (e *ConnMgr) ActivatePeer(ctx context.Context, conn *peer.Conn) {
-	e.lazyConnMgrMu.RLock()
-	lazyConnMgr := e.lazyConnMgr
-	started := lazyConnMgr != nil && e.lazyCtxCancel != nil
-	e.lazyConnMgrMu.RUnlock()
-	if !started {
+	if !e.isStartedWithLazyMgr() {
 		return
 	}
 
-	if found := lazyConnMgr.ActivatePeer(conn.GetKey()); found {
+	if found := e.lazyConnMgr.ActivatePeer(conn.GetKey()); found {
 		if err := conn.Open(ctx); err != nil {
 			conn.Log.Errorf("failed to open connection: %v", err)
 		}
@@ -292,22 +268,16 @@ func (e *ConnMgr) Close() {
 
 	e.lazyCtxCancel()
 	e.wg.Wait()
-
-	e.lazyConnMgrMu.Lock()
 	e.lazyConnMgr = nil
-	e.lazyConnMgrMu.Unlock()
 }
 
 func (e *ConnMgr) initLazyManager(engineCtx context.Context) {
 	cfg := manager.Config{
 		InactivityThreshold: inactivityThresholdEnv(),
-		ReconcileAllowedIPs: e.reconcileRoutedIPs,
 	}
-
-	e.lazyConnMgrMu.Lock()
 	e.lazyConnMgr = manager.NewManager(cfg, engineCtx, e.peerStore, e.iface)
+
 	e.lazyCtx, e.lazyCtxCancel = context.WithCancel(engineCtx)
-	e.lazyConnMgrMu.Unlock()
 
 	e.wg.Add(1)
 	go func() {
@@ -346,10 +316,7 @@ func (e *ConnMgr) closeManager(ctx context.Context) {
 
 	e.lazyCtxCancel()
 	e.wg.Wait()
-
-	e.lazyConnMgrMu.Lock()
 	e.lazyConnMgr = nil
-	e.lazyConnMgrMu.Unlock()
 
 	for _, peerID := range e.peerStore.PeersPubKey() {
 		e.peerStore.PeerConnOpen(ctx, peerID)

client/internal/conn_mgr_test.go

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

client/internal/connect.go

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

client/internal/debug/debug.go

@@ -480,6 +480,7 @@ func (g *BundleGenerator) addStatus() error {
 
 		fullStatus := g.statusRecorder.GetFullStatus()
 		protoFullStatus := nbstatus.ToProtoFullStatus(fullStatus)
+		protoFullStatus.Events = g.statusRecorder.GetEventHistory()
 		overview := nbstatus.ConvertToStatusOutputOverview(protoFullStatus, nbstatus.ConvertOptions{
 			Anonymize:     g.anonymize,
 			ProfileName:   profName,
@@ -676,7 +677,6 @@ func (g *BundleGenerator) addCommonConfigFields(configContent *strings.Builder)
 	configContent.WriteString(fmt.Sprintf("BlockLANAccess: %v\n", g.internalConfig.BlockLANAccess))
 	configContent.WriteString(fmt.Sprintf("BlockInbound: %v\n", g.internalConfig.BlockInbound))
 	configContent.WriteString(fmt.Sprintf("DisableIPv6: %v\n", g.internalConfig.DisableIPv6))
-	configContent.WriteString(fmt.Sprintf("SyncMessageVersion: %v\n", g.internalConfig.SyncMessageVersion))
 
 	if g.internalConfig.DisableNotifications != nil {
 		configContent.WriteString(fmt.Sprintf("DisableNotifications: %v\n", *g.internalConfig.DisableNotifications))

client/internal/debug/debug_test.go

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

client/internal/dns/local/local.go

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

client/internal/dns/local/local_test.go

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

client/internal/dns/local/warmup_test.go

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

client/internal/dns/mock_server.go

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

client/internal/dns/server.go

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

client/internal/dns/service_listener.go

@@ -292,16 +292,18 @@ func (s *serviceViaListener) generateFreePort() (uint16, error) {
 		return customPort, nil
 	}
 
-	probeListener, err := net.ListenUDP("udp4", &net.UDPAddr{})
+	udpAddr := net.UDPAddrFromAddrPort(netip.MustParseAddrPort("0.0.0.0:0"))
+	probeListener, err := net.ListenUDP("udp", udpAddr)
 	if err != nil {
 		log.Debugf("failed to bind random port for DNS: %s", err)
 		return 0, err
 	}
 
-	port := uint16(probeListener.LocalAddr().(*net.UDPAddr).Port)
-	if err = probeListener.Close(); err != nil {
+	addrPort := netip.MustParseAddrPort(probeListener.LocalAddr().String()) // might panic if address is incorrect
+	err = probeListener.Close()
+	if err != nil {
 		log.Debugf("failed to free up DNS port: %s", err)
 		return 0, err
 	}
-	return port, nil
+	return addrPort.Port(), nil
 }

client/internal/dns_peer_activator.go

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

client/internal/dns_peer_activator_test.go

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

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

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

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

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

client/internal/engine.go

@@ -50,7 +50,6 @@ import (
 	icemaker "github.com/netbirdio/netbird/client/internal/peer/ice"
 	"github.com/netbirdio/netbird/client/internal/peerstore"
 	"github.com/netbirdio/netbird/client/internal/portforward"
-	"github.com/netbirdio/netbird/client/internal/pqkem"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	"github.com/netbirdio/netbird/client/internal/relay"
 	"github.com/netbirdio/netbird/client/internal/rosenpass"
@@ -65,10 +64,7 @@ import (
 	"github.com/netbirdio/netbird/route"
 	mgm "github.com/netbirdio/netbird/shared/management/client"
 	"github.com/netbirdio/netbird/shared/management/domain"
-	sharedgrpc "github.com/netbirdio/netbird/shared/management/grpc"
-	nbnetworkmap "github.com/netbirdio/netbird/shared/management/networkmap"
 	mgmProto "github.com/netbirdio/netbird/shared/management/proto"
-	types "github.com/netbirdio/netbird/shared/management/types"
 	"github.com/netbirdio/netbird/shared/netiputil"
 	auth "github.com/netbirdio/netbird/shared/relay/auth/hmac"
 	relayClient "github.com/netbirdio/netbird/shared/relay/client"
@@ -151,7 +147,6 @@ type EngineConfig struct {
 	BlockLANAccess      bool
 	BlockInbound        bool
 	DisableIPv6         bool
-	SyncMessageVersion  *int
 
 	// LazyConnection is the MDM-sourced lazy-connection override; StateUnset defers to
 	// the env var and management feature flag.
@@ -198,10 +193,6 @@ type Engine struct {
 	// rpManager is a Rosenpass manager
 	rpManager *rosenpass.Manager
 
-	// pqkemManager runs the ML-KEM post-quantum PSK exchange (gated by NB_ENABLE_PQ_MLKEM).
-	// It owns the data-path transport and peer endpoint routing.
-	pqkemManager *pqkem.Manager
-
 	// syncMsgMux is used to guarantee sequential Management Service message processing
 	syncMsgMux *sync.Mutex
 
@@ -229,13 +220,6 @@ type Engine struct {
 	// networkSerial is the latest CurrentSerial (state ID) of the network sent by the Management service
 	networkSerial uint64
 
-	// latestComponents is the most-recent NetworkMapComponents decoded from
-	// a NetworkMapEnvelope (capability=3 peers only). Held alongside the
-	// NetworkMap that Calculate() produced from it so future incremental
-	// updates have a base to apply changes against. nil for legacy-format
-	// peers. Guarded by syncMsgMux.
-	latestComponents *types.NetworkMapComponents
-
 	networkMonitor *networkmonitor.NetworkMonitor
 
 	sshServer sshServer
@@ -567,7 +551,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 		} else {
 			log.Infof("running rosenpass in strict mode")
 		}
-		e.rpManager, err = rosenpass.NewManager(e.config.PreSharedKey, e.config.WgIfaceName, publicKey)
+		e.rpManager, err = rosenpass.NewManager(e.config.PreSharedKey, e.config.WgIfaceName)
 		if err != nil {
 			return fmt.Errorf("create rosenpass manager: %w", err)
 		}
@@ -656,19 +640,6 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 		e.rpManager.SetInterface(e.wgInterface)
 	}
 
-	// Start the ML-KEM PQ manager after the interface is up so its dedicated UDP
-	// transport can bind on the WG overlay IP.
-	if pqkem.Enabled() {
-		tr, pqErr := newPQTransport(e.config.WgAddr.IP)
-		if pqErr != nil {
-			log.Errorf("pqkem: transport bind failed, exchange disabled: %v", pqErr)
-		} else {
-			e.pqkemManager = pqkem.NewManager(pqkem.LocalID(publicKey.String()), pqCallbackHandler{wg: e.wgInterface}, pqkem.NewLogger())
-			e.pqkemManager.Start(tr)
-			log.Infof("pqkem: enabled (udp port %d on overlay %s)", e.pqkemManager.LocalPort(), e.config.WgAddr.IP)
-		}
-	}
-
 	// if inbound conns are blocked there is no need to create the ACL manager
 	if e.firewall != nil && !e.config.BlockInbound {
 		e.acl = acl.NewDefaultManager(e.firewall)
@@ -681,24 +652,8 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	iceCfg := e.createICEConfig()
 
 	e.connMgr = NewConnMgr(e.config, e.statusRecorder, e.peerStore, wgIface)
-	e.connMgr.SetRoutedIPsReconciler(func(peerKey string) error {
-		if e.routeManager == nil {
-			return nil
-		}
-		return e.routeManager.ReconcilePeerAllowedIPs(peerKey)
-	})
 	e.connMgr.Start(e.ctx)
 
-	// Wire DNS-time lazy-connection warm-up now that the connection manager
-	// exists (it does not at DNS-server construction time). A DNS answer that
-	// points at an idle peer then wakes it before the client's first request.
-	e.dnsServer.SetPeerActivator(&dnsPeerActivator{
-		connMgr:   e.connMgr,
-		peerStore: e.peerStore,
-		status:    e.statusRecorder,
-		ctx:       e.ctx,
-	})
-
 	e.srWatcher = guard.NewSRWatcher(e.signal, e.relayManager, e.mobileDep.IFaceDiscover, iceCfg)
 	e.srWatcher.Start(peer.IsForceRelayed())
 
@@ -932,10 +887,6 @@ func (e *Engine) removePeer(peerKey string) error {
 
 	e.connMgr.RemovePeerConn(peerKey)
 
-	if e.pqkemManager != nil {
-		e.pqkemManager.RemovePeer(pqkem.RemoteID(peerKey))
-	}
-
 	err := e.statusRecorder.RemovePeer(peerKey)
 	if err != nil {
 		log.Warnf("received error when removing peer %s from status recorder: %v", peerKey, err)
@@ -1012,12 +963,8 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 
 	e.ApplySessionDeadline(update.GetSessionExpiresAt())
 
-	// Envelope sync responses carry PeerConfig at the top level; legacy
-	// NetworkMap syncs carry it under NetworkMap.PeerConfig.
-	if pc := update.GetPeerConfig(); pc != nil {
-		e.handleAutoUpdateVersion(pc.GetAutoUpdate())
-	} else if nm := update.GetNetworkMap(); nm != nil && nm.GetPeerConfig() != nil {
-		e.handleAutoUpdateVersion(nm.GetPeerConfig().GetAutoUpdate())
+	if update.NetworkMap != nil && update.NetworkMap.PeerConfig != nil {
+		e.handleAutoUpdateVersion(update.NetworkMap.PeerConfig.AutoUpdate)
 	}
 
 	done := e.phase("netbird_config")
@@ -1027,47 +974,12 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 		return err
 	}
 
-	// Decode the network map from either the components envelope or the
-	// legacy proto.NetworkMap before the posture-check gating below, so the
-	// "is there a network map" decision covers both wire shapes.
-	var (
-		nm         *mgmProto.NetworkMap
-		components *types.NetworkMapComponents
-	)
-	if version := update.GetVersion(); version == int32(sharedgrpc.ComponentNetworkMap) {
-		// Components-format peer: decode the envelope back to typed
-		// components, run Calculate() locally, and convert to the wire
-		// NetworkMap shape the rest of the engine consumes. Components are
-		// retained so future incremental updates can apply deltas instead
-		// of doing a full reconstruction.
-		envelope := update.GetNetworkMapEnvelope()
-		if envelope == nil {
-			return fmt.Errorf("received a SyncReponse indicating use of components network map, but components are missing")
-		}
-
-		localKey := e.config.WgPrivateKey.PublicKey().String()
-		dnsName := ""
-		if pc := update.GetPeerConfig(); pc != nil {
-			// PeerConfig.Fqdn = "<dns_label>.<dns_domain>" — extract the
-			// shared domain by stripping the peer's own label prefix. Falls
-			// back to empty if the FQDN doesn't have the expected shape.
-			dnsName = extractDNSDomainFromFQDN(pc.GetFqdn())
-		}
-		result, err := nbnetworkmap.EnvelopeToNetworkMap(e.ctx, envelope, localKey, dnsName)
-		if err != nil {
-			return fmt.Errorf("decode network map envelope: %w", err)
-		}
-		nm = result.NetworkMap
-		components = result.Components
-	} else {
-		nm = update.GetNetworkMap()
-	}
-
 	// Posture checks are bound to the network map presence:
 	//   NetworkMap != nil, checks present -> apply the received checks
 	//   NetworkMap != nil, checks nil     -> posture checks were removed, clear them
 	//   NetworkMap == nil                 -> config-only update (e.g. relay token rotation),
 	//                                        leave the previously applied checks untouched
+	nm := update.GetNetworkMap()
 	if nm == nil {
 		return nil
 	}
@@ -1080,14 +992,6 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 	}
 
 	done = e.phase("persist")
-	// Only retain the components view when the server sent the envelope
-	// path. A legacy proto.NetworkMap means components == nil; writing it
-	// here would clobber a previously-cached snapshot, breaking the
-	// incremental-delta base on a future envelope sync.
-	if components != nil {
-		e.latestComponents = components
-	}
-
 	e.persistSyncResponse(update)
 	done()
 
@@ -1101,19 +1005,6 @@ func (e *Engine) handleSync(update *mgmProto.SyncResponse) error {
 	return nil
 }
 
-// extractDNSDomainFromFQDN returns the trailing dotted domain part of the
-// receiving peer's FQDN — the same value the management server fills as
-// dnsName when it builds the legacy NetworkMap. "peer42.netbird.cloud" →
-// "netbird.cloud". An empty string is returned for unrecognized formats.
-func extractDNSDomainFromFQDN(fqdn string) string {
-	for i := 0; i < len(fqdn); i++ {
-		if fqdn[i] == '.' && i+1 < len(fqdn) {
-			return fqdn[i+1:]
-		}
-	}
-	return ""
-}
-
 // updateNetbirdConfig applies the management-provided NetBird configuration:
 // STUN/TURN and relay servers, flow logging and DNS settings. A nil config is a no-op,
 // which is the case for sync updates carrying only a network map.
@@ -1273,7 +1164,6 @@ func (e *Engine) applyInfoFlags(info *system.Info) {
 		e.config.BlockLANAccess,
 		e.config.BlockInbound,
 		e.config.DisableIPv6,
-		e.config.SyncMessageVersion,
 		e.config.EnableSSHRoot,
 		e.config.EnableSSHSFTP,
 		e.config.EnableSSHLocalPortForwarding,
@@ -1922,9 +1812,6 @@ func (e *Engine) createPeerConn(pubKey string, allowedIPs []netip.Prefix, agentV
 		},
 		ICEConfig: e.createICEConfig(),
 	}
-	if e.pqkemManager != nil {
-		config.PQ = pqHandshaker{mgr: e.pqkemManager}
-	}
 
 	serviceDependencies := peer.ServiceDependencies{
 		StatusRecorder:     e.statusRecorder,
@@ -2108,10 +1995,6 @@ func (e *Engine) close() {
 		_ = e.rpManager.Close()
 	}
 
-	if e.pqkemManager != nil {
-		e.pqkemManager.Stop()
-	}
-
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 	defer cancel()
 	if err := e.portForwardManager.GracefullyStop(ctx); err != nil {
@@ -2149,7 +2032,6 @@ func (e *Engine) readInitialSettings() ([]*route.Route, *nbdns.Config, bool, err
 		e.config.BlockLANAccess,
 		e.config.BlockInbound,
 		e.config.DisableIPv6,
-		e.config.SyncMessageVersion,
 		e.config.EnableSSHRoot,
 		e.config.EnableSSHSFTP,
 		e.config.EnableSSHLocalPortForwarding,
@@ -2723,14 +2605,13 @@ func (e *Engine) updateForwardRules(rules []*mgmProto.ForwardingRule) ([]firewal
 
 func (e *Engine) toExcludedLazyPeers(rules []firewallManager.ForwardRule, peers []*mgmProto.RemotePeerConfig) map[string]bool {
 	excludedPeers := make(map[string]bool)
-
-	// Ingress forward targets: inbound forwarded traffic is initiated remotely and
-	// cannot wake a lazy connection, so the peer routing the target must stay
-	// permanently connected. AllowedIPs are already parsed on the peer conn, so
-	// reuse those typed prefixes instead of re-parsing the network map strings.
 	for _, r := range rules {
+		ip := r.TranslatedAddress
 		for _, p := range peers {
-			if e.peerRoutesAddr(p, r.TranslatedAddress) {
+			for _, allowedIP := range p.GetAllowedIps() {
+				if allowedIP != ip.String() {
+					continue
+				}
 				log.Infof("exclude forwarder peer from lazy connection: %s", p.GetWgPubKey())
 				excludedPeers[p.GetWgPubKey()] = true
 			}
@@ -2740,27 +2621,6 @@ func (e *Engine) toExcludedLazyPeers(rules []firewallManager.ForwardRule, peers
 	return excludedPeers
 }
 
-// peerRoutesAddr reports whether the peer is a router for addr, matched against
-// the peer's already-parsed AllowedIPs from the store (the same typed value the
-// lazy manager consumes) rather than re-parsing the network map strings.
-func (e *Engine) peerRoutesAddr(p *mgmProto.RemotePeerConfig, addr netip.Addr) bool {
-	prefixes, ok := e.peerStore.AllowedIPs(p.GetWgPubKey())
-	if !ok {
-		return false
-	}
-	return prefixesContain(prefixes, addr)
-}
-
-// prefixesContain reports whether addr falls within any of the prefixes.
-func prefixesContain(prefixes []netip.Prefix, addr netip.Addr) bool {
-	for _, prefix := range prefixes {
-		if prefix.Contains(addr) {
-			return true
-		}
-	}
-	return false
-}
-
 // isChecksEqual checks if two slices of checks are equal.
 func isChecksEqual(checks1, checks2 []*mgmProto.Checks) bool {
 	normalize := func(checks []*mgmProto.Checks) []string {
@@ -2924,8 +2784,6 @@ func convertToOfferAnswer(msg *sProto.Message) (*peer.OfferAnswer, error) {
 		Version:         msg.GetBody().GetNetBirdVersion(),
 		RosenpassPubKey: rosenpassPubKey,
 		RosenpassAddr:   rosenpassAddr,
-		MlkemPayload:    msg.GetBody().GetMlkemPayload(),
-		MlkemPort:       int(msg.GetBody().GetMlkemPort()),
 		RelaySrvAddress: msg.GetBody().GetRelayServerAddress(),
 		RelaySrvIP:      relayIP,
 		SessionID:       sessionID,

client/internal/engine_lazy_exclude_test.go

@@ -1,87 +0,0 @@
-package internal
-
-import (
-	"net/netip"
-	"testing"
-
-	"github.com/stretchr/testify/require"
-
-	firewallManager "github.com/netbirdio/netbird/client/firewall/manager"
-	"github.com/netbirdio/netbird/client/internal/peer"
-	"github.com/netbirdio/netbird/client/internal/peerstore"
-	mgmProto "github.com/netbirdio/netbird/shared/management/proto"
-)
-
-func TestPrefixesContain(t *testing.T) {
-	tests := []struct {
-		name     string
-		prefixes []string
-		addr     string
-		want     bool
-	}{
-		{name: "own overlay /32 matches", prefixes: []string{"100.110.8.145/32"}, addr: "100.110.8.145", want: true},
-		{name: "addr inside routed subnet", prefixes: []string{"10.121.0.0/16"}, addr: "10.121.208.4", want: true},
-		{name: "addr outside subnet", prefixes: []string{"10.121.0.0/16"}, addr: "10.122.0.1", want: false},
-		{name: "different /32", prefixes: []string{"100.110.8.145/32"}, addr: "100.110.8.146", want: false},
-		{name: "ipv6 /128 matches", prefixes: []string{"fd00::1/128"}, addr: "fd00::1", want: true},
-		{name: "no prefixes", prefixes: nil, addr: "10.121.208.4", want: false},
-	}
-
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			prefixes := make([]netip.Prefix, 0, len(tt.prefixes))
-			for _, p := range tt.prefixes {
-				prefixes = append(prefixes, netip.MustParsePrefix(p))
-			}
-			require.Equal(t, tt.want, prefixesContain(prefixes, netip.MustParseAddr(tt.addr)))
-		})
-	}
-}
-
-// TestToExcludedLazyPeers_ForwardTarget guards a regression: the forward-target
-// peer (the peer routing a ForwardRule.TranslatedAddress) must be excluded from
-// lazy connections, matched via the peer's already-parsed AllowedIPs.
-func TestToExcludedLazyPeers_ForwardTarget(t *testing.T) {
-	const targetPeerKey = "cccccccccccccccccccccccccccccccccccccccccc0="
-	const otherPeerKey = "dddddddddddddddddddddddddddddddddddddddddd0="
-
-	store := peerstore.NewConnStore()
-	store.AddPeerConn(targetPeerKey, newTestConn(t, targetPeerKey, "100.110.8.145/32"))
-	store.AddPeerConn(otherPeerKey, newTestConn(t, otherPeerKey, "100.110.9.10/32"))
-
-	e := &Engine{peerStore: store}
-
-	peers := []*mgmProto.RemotePeerConfig{
-		{WgPubKey: targetPeerKey, AllowedIps: []string{"100.110.8.145/32"}},
-		{WgPubKey: otherPeerKey, AllowedIps: []string{"100.110.9.10/32"}},
-	}
-	rules := []firewallManager.ForwardRule{
-		{TranslatedAddress: netip.MustParseAddr("100.110.8.145")},
-	}
-
-	excluded := e.toExcludedLazyPeers(rules, peers)
-
-	require.True(t, excluded[targetPeerKey], "forward-target peer must be excluded from lazy connections")
-	require.False(t, excluded[otherPeerKey], "non-target peer must not be excluded")
-	require.Len(t, excluded, 1)
-}
-
-func TestToExcludedLazyPeers_NoRules(t *testing.T) {
-	e := &Engine{peerStore: peerstore.NewConnStore()}
-
-	peers := []*mgmProto.RemotePeerConfig{
-		{WgPubKey: "peer-a", AllowedIps: []string{"100.110.8.145/32"}},
-	}
-
-	require.Empty(t, e.toExcludedLazyPeers(nil, peers))
-}
-
-func newTestConn(t *testing.T, key, allowedIP string) *peer.Conn {
-	t.Helper()
-	conn, err := peer.NewConn(peer.ConnConfig{
-		Key:      key,
-		WgConfig: peer.WgConfig{AllowedIps: []netip.Prefix{netip.MustParsePrefix(allowedIP)}},
-	}, peer.ServiceDependencies{})
-	require.NoError(t, err)
-	return conn
-}

client/internal/engine_session_deadline_test.go

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

client/internal/lazyconn/manager/manager.go

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

client/internal/netflow/store/event_aggregation_test.go

@@ -175,9 +175,10 @@ func TestFlowAggregationOfUnknownProtocols(t *testing.T) {
 }
 
 func TestResetAggregationWindow(t *testing.T) {
-	now := time.Now()
-	nowFunc := func() time.Time { return now }
-	store := NewAggregatingMemoryStoreWithTimeFunc(nowFunc)
+	store := NewAggregatingMemoryStore()
+	// Backdate the window start so the reset produces a different timestamp
+	// even on platforms with coarse clock granularity.
+	store.WindowStart = store.WindowStart.Add(-time.Second)
 	store.StoreEvent(&types.Event{
 		ID:        uuid.New(),
 		Timestamp: time.Now(),
@@ -200,7 +201,6 @@ func TestResetAggregationWindow(t *testing.T) {
 		},
 	})
 
-	now = now.Add(1 * time.Second)
 	reset := store.ResetAggregationWindow()
 	previousEvents, ok := reset.(*AggregatingMemory)
 	assert.True(t, ok)

client/internal/netflow/store/memory.go

@@ -29,7 +29,6 @@ type AggregatingMemory struct {
 	WindowStart time.Time
 	WindowEnd   time.Time
 	rnd         *v2.PCG
-	nowFunc     func() time.Time
 }
 
 func (m *Memory) StoreEvent(event *types.Event) {
@@ -63,19 +62,14 @@ func (m *Memory) DeleteEvents(ids []uuid.UUID) {
 }
 
 func NewAggregatingMemoryStore() *AggregatingMemory {
-	return NewAggregatingMemoryStoreWithTimeFunc(defaultNowFunc)
-}
-
-// used in tests when deterministic (less random) time intervals are required
-func NewAggregatingMemoryStoreWithTimeFunc(nowFunc func() time.Time) *AggregatingMemory {
-	return &AggregatingMemory{WindowStart: nowFunc(), Memory: Memory{events: make(map[uuid.UUID]*types.Event)}, nowFunc: nowFunc, rnd: v2.NewPCG(rand.Uint64(), rand.Uint64())}
+	return &AggregatingMemory{WindowStart: time.Now(), Memory: Memory{events: make(map[uuid.UUID]*types.Event)}, rnd: v2.NewPCG(rand.Uint64(), rand.Uint64())}
 }
 
 func (am *AggregatingMemory) ResetAggregationWindow() types.FlowEventAggregator {
 	am.mux.Lock()
 	defer am.mux.Unlock()
 
-	now := am.nowFunc()
+	now := time.Now()
 	toret := AggregatingMemory{WindowStart: am.WindowStart, WindowEnd: now, Memory: Memory{events: am.events}, rnd: v2.NewPCG(rand.Uint64(), rand.Uint64())}
 
 	am.events = make(map[uuid.UUID]*types.Event)
@@ -158,7 +152,3 @@ func (am *AggregatingMemory) GetAggregatedEvents() []*types.Event {
 
 	return slices.Collect(maps.Values(aggregated)) // could return an iterator instead here
 }
-
-func defaultNowFunc() time.Time {
-	return time.Now()
-}

client/internal/peer/conn.go

@@ -30,11 +30,6 @@ import (
 	relayClient "github.com/netbirdio/netbird/shared/relay/client"
 )
 
-// wgTimeoutEscalationThreshold is the number of consecutive WireGuard
-// handshake timeouts after which the rosenpass state for the peer is
-// considered desynced and gets reset.
-const wgTimeoutEscalationThreshold = 3
-
 // MetricsRecorder is an interface for recording peer connection metrics
 type MetricsRecorder interface {
 	RecordConnectionStages(
@@ -74,32 +69,6 @@ type RosenpassConfig struct {
 	PermissiveMode bool
 }
 
-// PQHandshaker attaches post-quantum ML-KEM material to signalling offers/answers and
-// feeds received material back. It is implemented by the engine over the pqkem
-// manager and is nil when the PQ exchange is disabled. remoteKey is the peer's
-// WireGuard public key.
-type PQHandshaker interface {
-	// OfferPayload returns the KEM offer to embed in an outgoing offer (nil if this
-	// peer is not the KEM initiator) and the local PQ data-path port to announce.
-	OfferPayload(remoteKey string) (payload []byte, port int)
-	// AnswerPayload processes a received KEM offer (nil if absent) and returns the KEM
-	// answer to embed in the outgoing answer (nil if none) and the local PQ port.
-	AnswerPayload(remoteKey string, recvOffer []byte) (payload []byte, port int)
-	// OnAnswer feeds a received KEM answer (nil if absent).
-	OnAnswer(remoteKey string, recvAnswer []byte)
-	// PSK returns the peer's latest derived post-quantum PSK to program at WG
-	// peer-config time (the pull path). ok is false until one has been derived.
-	PSK(remoteKey string) (wgtypes.Key, bool)
-	// SetRemoteAddr registers the peer's data-path endpoint learned from signalling:
-	// its WG overlay IP with the advertised pq UDP port.
-	SetRemoteAddr(remoteKey string, addr netip.AddrPort)
-	// OnDataPathRekeyed signals a fresh WireGuard handshake for the peer; it clocks the
-	// next chained PSK rotation pushed over the data path.
-	OnDataPathRekeyed(remoteKey string)
-	// OnDataPathDown signals the peer's tunnel went down.
-	OnDataPathDown(remoteKey string)
-}
-
 // ConnConfig is a peer Connection configuration
 type ConnConfig struct {
 	// Key is a public key of a remote peer
@@ -117,9 +86,6 @@ type ConnConfig struct {
 
 	RosenpassConfig RosenpassConfig
 
-	// PQ carries post-quantum ML-KEM material on offers/answers; nil when disabled.
-	PQ PQHandshaker
-
 	// ICEConfig ICE protocol configuration
 	ICEConfig icemaker.Config
 }
@@ -152,9 +118,6 @@ type Conn struct {
 	wgWatcher       *WGWatcher
 	wgWatcherWg     sync.WaitGroup
 	wgWatcherCancel context.CancelFunc
-	// wgTimeouts counts consecutive WireGuard handshake timeouts without a
-	// successful handshake in between. Guarded by mu.
-	wgTimeouts int
 
 	// used to store the remote Rosenpass key for Relayed connection in case of connection update from ice
 	rosenpassRemoteKey []byte
@@ -232,6 +195,7 @@ func NewConn(config ConnConfig, services ServiceDependencies) (*Conn, error) {
 		statusICE:          worker.NewAtomicStatus(),
 		dumpState:          dumpState,
 		endpointUpdater:    NewEndpointUpdater(connLog, config.WgConfig, isController(config)),
+		wgWatcher:          NewWGWatcher(connLog, config.WgConfig.WgInterface, config.Key, dumpState),
 		metricsRecorder:    services.MetricsRecorder,
 	}
 
@@ -699,21 +663,16 @@ func (conn *Conn) onGuardEvent() {
 	}
 }
 
-func (conn *Conn) onWGDisconnected(watcherCtx context.Context) {
+func (conn *Conn) onWGDisconnected() {
 	conn.mu.Lock()
 	defer conn.mu.Unlock()
 
-	// watcherCtx guards against a stale watcher tearing down a connection that already superseded it.
-	if conn.ctx.Err() != nil || watcherCtx.Err() != nil {
+	if conn.ctx.Err() != nil {
 		return
 	}
 
 	conn.Log.Warnf("WireGuard handshake timeout detected, closing current connection")
 
-	if conn.config.PQ != nil {
-		conn.config.PQ.OnDataPathDown(conn.config.Key)
-	}
-
 	// Close the active connection based on current priority
 	switch conn.currentConnPriority {
 	case conntype.Relay:
@@ -724,29 +683,6 @@ func (conn *Conn) onWGDisconnected(watcherCtx context.Context) {
 	default:
 		conn.Log.Debugf("No active connection to close on WG timeout")
 	}
-
-	conn.escalateWGTimeoutLocked()
-}
-
-// escalateWGTimeoutLocked resets the peer's rosenpass state after repeated
-// handshake timeouts. With rosenpass enabled, persistent timeouts mean the
-// preshared keys have desynced; the renewal exchange runs over the dead
-// tunnel and cannot resync them. Reporting the peer disconnected drops its
-// rosenpass state, so the next connection configuration programs the
-// rendezvous key and the tunnel can bootstrap again. Callers must hold mu.
-func (conn *Conn) escalateWGTimeoutLocked() {
-	if conn.config.RosenpassConfig.PubKey == nil {
-		return
-	}
-
-	conn.wgTimeouts++
-	if conn.wgTimeouts < wgTimeoutEscalationThreshold || conn.onDisconnected == nil {
-		return
-	}
-	conn.wgTimeouts = 0
-
-	conn.Log.Warnf("%d consecutive WireGuard handshake timeouts, resetting rosenpass state for peer", wgTimeoutEscalationThreshold)
-	conn.onDisconnected(conn.config.WgConfig.RemoteKey)
 }
 
 func (conn *Conn) updateRelayStatus(relayServerAddr string, rosenpassPubKey []byte, updateTime time.Time) {
@@ -866,39 +802,25 @@ func (conn *Conn) isConnectedOnAllWay() (status guard.ConnStatus) {
 	})
 }
 
-// enableWgWatcherIfNeeded starts a fresh watcher instance per connection attempt, so its
-// lifecycle stays bound to conn.mu and enable/disable can't race an old goroutine's shutdown.
-// Caller must hold conn.mu.
 func (conn *Conn) enableWgWatcherIfNeeded(enabledTime time.Time) {
-	if conn.wgWatcher != nil {
+	if !conn.wgWatcher.PrepareInitialHandshake() {
 		return
 	}
 
-	watcher := NewWGWatcher(conn.Log, conn.config.WgConfig.WgInterface, conn.config.Key, conn.dumpState)
-	watcher.PrepareInitialHandshake()
-
 	wgWatcherCtx, wgWatcherCancel := context.WithCancel(conn.ctx)
-	conn.wgWatcher = watcher
 	conn.wgWatcherCancel = wgWatcherCancel
-
 	conn.wgWatcherWg.Add(1)
 	go func() {
 		defer conn.wgWatcherWg.Done()
-		onDisconnected := func() { conn.onWGDisconnected(wgWatcherCtx) }
-		watcher.EnableWgWatcher(wgWatcherCtx, enabledTime, onDisconnected, conn.onWGHandshakeSuccess, conn.onWGCheckSuccess)
+		conn.wgWatcher.EnableWgWatcher(wgWatcherCtx, enabledTime, conn.onWGDisconnected, conn.onWGHandshakeSuccess)
 	}()
 }
 
-// disableWgWatcherIfNeeded cancels and drops the watcher once no transport is active. It never
-// waits for the goroutine: the timeout path reentrantly calls back here under conn.mu, so
-// blocking would deadlock. Caller must hold conn.mu.
 func (conn *Conn) disableWgWatcherIfNeeded() {
-	if conn.currentConnPriority != conntype.None || conn.wgWatcher == nil {
-		return
+	if conn.currentConnPriority == conntype.None && conn.wgWatcherCancel != nil {
+		conn.wgWatcherCancel()
+		conn.wgWatcherCancel = nil
 	}
-	conn.wgWatcherCancel()
-	conn.wgWatcher = nil
-	conn.wgWatcherCancel = nil
 }
 
 func (conn *Conn) newProxy(remoteConn net.Conn) (wgproxy.Proxy, error) {
@@ -921,9 +843,7 @@ func (conn *Conn) resetEndpoint() {
 		return
 	}
 	conn.Log.Infof("reset wg endpoint")
-	if conn.wgWatcher != nil {
-		conn.wgWatcher.Reset()
-	}
+	conn.wgWatcher.Reset()
 	if err := conn.endpointUpdater.RemoveEndpointAddress(); err != nil {
 		conn.Log.Warnf("failed to remove endpoint address before update: %v", err)
 	}
@@ -972,20 +892,6 @@ func (conn *Conn) onWGHandshakeSuccess(when time.Time) {
 	conn.recordConnectionMetrics()
 }
 
-// onWGCheckSuccess is called for every watcher check that observed a fresh
-// handshake, including handshakes of connections that were already up when
-// the watcher started.
-func (conn *Conn) onWGCheckSuccess() {
-	conn.mu.Lock()
-	conn.wgTimeouts = 0
-	conn.mu.Unlock()
-
-	// A fresh WireGuard handshake is the clock for the post-quantum PSK rotation.
-	if conn.config.PQ != nil {
-		conn.config.PQ.OnDataPathRekeyed(conn.config.Key)
-	}
-}
-
 // recordConnectionMetrics records connection stage timestamps as metrics
 func (conn *Conn) recordConnectionMetrics() {
 	if conn.metricsRecorder == nil {
@@ -1025,15 +931,6 @@ func (conn *Conn) AgentVersionString() string {
 }
 
 func (conn *Conn) presharedKey(remoteRosenpassKey []byte) *wgtypes.Key {
-	// Post-quantum: once the ML-KEM exchange has derived a PSK for this peer, program
-	// it here so the peer's next WireGuard handshake adopts it. Applied at peer-config
-	// time (bootstrap / reconnect); steady-state rotation is pushed separately.
-	if conn.config.PQ != nil {
-		if psk, ok := conn.config.PQ.PSK(conn.config.Key); ok {
-			return &psk
-		}
-	}
-
 	if conn.config.RosenpassConfig.PubKey == nil {
 		return conn.config.WgConfig.PreSharedKey
 	}

client/internal/peer/conn_test.go

@@ -7,7 +7,6 @@ import (
 	"testing"
 	"time"
 
-	log "github.com/sirupsen/logrus"
 	"github.com/stretchr/testify/assert"
 
 	"github.com/netbirdio/netbird/client/iface"
@@ -305,84 +304,3 @@ func TestConn_presharedKey_RosenpassManaged(t *testing.T) {
 		t.Fatalf("expected non-nil presharedKey before Rosenpass manages PSK")
 	}
 }
-
-func newWGTimeoutTestConn(rosenpassEnabled bool, disconnected *[]string) *Conn {
-	cfg := ConnConfig{
-		Key:      "LLHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
-		LocalKey: "RRHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU=",
-		WgConfig: WgConfig{RemoteKey: "LLHf3Ma6z6mdLbriAJbqhX7+nM/B71lgw2+91q3LfhU="},
-	}
-	if rosenpassEnabled {
-		cfg.RosenpassConfig = RosenpassConfig{PubKey: []byte("dummykey")}
-	}
-
-	conn := &Conn{
-		ctx:           context.Background(),
-		config:        cfg,
-		Log:           log.WithField("peer", cfg.Key),
-		metricsStages: &MetricsStages{},
-	}
-	conn.SetOnDisconnected(func(remotePeer string) {
-		*disconnected = append(*disconnected, remotePeer)
-	})
-	return conn
-}
-
-// TestConn_onWGDisconnected_EscalatesToRosenpassReset: repeated handshake
-// timeouts with rosenpass enabled mean the preshared keys have desynced. The
-// renewal exchange runs over the dead tunnel and cannot resync them, so after
-// wgTimeoutEscalationThreshold consecutive timeouts the conn must report the
-// peer disconnected, dropping its rosenpass state so the next configuration
-// programs the rendezvous key.
-func TestConn_onWGDisconnected_EscalatesToRosenpassReset(t *testing.T) {
-	var disconnected []string
-	conn := newWGTimeoutTestConn(true, &disconnected)
-
-	for i := 0; i < wgTimeoutEscalationThreshold-1; i++ {
-		conn.onWGDisconnected(conn.ctx)
-	}
-	assert.Empty(t, disconnected, "escalation must not fire below the threshold")
-
-	conn.onWGDisconnected(conn.ctx)
-	assert.Equal(t, []string{conn.config.WgConfig.RemoteKey}, disconnected,
-		"reaching the threshold must report the peer disconnected once")
-
-	for i := 0; i < wgTimeoutEscalationThreshold-1; i++ {
-		conn.onWGDisconnected(conn.ctx)
-	}
-	assert.Len(t, disconnected, 1, "escalation must restart counting after firing")
-
-	conn.onWGDisconnected(conn.ctx)
-	assert.Len(t, disconnected, 2, "continued timeouts must escalate again")
-}
-
-// TestConn_onWGDisconnected_CheckSuccessResetsEscalation: a successful
-// handshake between timeouts means the tunnel recovered; the counter must
-// start over.
-func TestConn_onWGDisconnected_CheckSuccessResetsEscalation(t *testing.T) {
-	var disconnected []string
-	conn := newWGTimeoutTestConn(true, &disconnected)
-
-	for i := 0; i < wgTimeoutEscalationThreshold-1; i++ {
-		conn.onWGDisconnected(conn.ctx)
-	}
-	conn.onWGCheckSuccess()
-
-	for i := 0; i < wgTimeoutEscalationThreshold-1; i++ {
-		conn.onWGDisconnected(conn.ctx)
-	}
-	assert.Empty(t, disconnected, "handshake success must reset the timeout count")
-}
-
-// TestConn_onWGDisconnected_NoEscalationWithoutRosenpass: without rosenpass
-// there is no per-peer key state to reset; repeated timeouts must not report
-// disconnects.
-func TestConn_onWGDisconnected_NoEscalationWithoutRosenpass(t *testing.T) {
-	var disconnected []string
-	conn := newWGTimeoutTestConn(false, &disconnected)
-
-	for i := 0; i < wgTimeoutEscalationThreshold*3; i++ {
-		conn.onWGDisconnected(conn.ctx)
-	}
-	assert.Empty(t, disconnected, "escalation must be limited to rosenpass connections")
-}

client/internal/peer/handshaker.go

@@ -39,16 +39,6 @@ type OfferAnswer struct {
 	// This value is the local Rosenpass server address when sending the message
 	RosenpassAddr string
 
-	// MlkemPayload carries the post-quantum X25519MLKEM768 handshake message
-	// (pqkem-framed offer on an OFFER, answer on an ANSWER) that seeds the
-	// WireGuard PSK. Opaque here — the pqkem library frames and parses it. Nil
-	// when the peer does not run the ML-KEM PQ exchange.
-	MlkemPayload []byte
-
-	// MlkemPort is the peer's ML-KEM PQ service UDP port (bound on its WG overlay
-	// IP) where data-path rekey messages are sent. Zero when not running the exchange.
-	MlkemPort int
-
 	// relay server address
 	RelaySrvAddress string
 	// RelaySrvIP is the IP the remote peer is connected to on its
@@ -130,8 +120,6 @@ func (h *Handshaker) Listen(ctx context.Context) {
 
 			h.updateRemoteICEState(&remoteOfferAnswer)
 
-			h.pqRegisterEndpoint(remoteOfferAnswer.MlkemPort)
-
 			if h.relayListener != nil {
 				h.relayListener.Notify(&remoteOfferAnswer)
 			}
@@ -140,7 +128,7 @@ func (h *Handshaker) Listen(ctx context.Context) {
 				h.iceListener(&remoteOfferAnswer)
 			}
 
-			if err := h.sendAnswer(&remoteOfferAnswer); err != nil {
+			if err := h.sendAnswer(); err != nil {
 				h.log.Errorf("failed to send remote offer confirmation: %s", err)
 				continue
 			}
@@ -154,8 +142,6 @@ func (h *Handshaker) Listen(ctx context.Context) {
 
 			h.updateRemoteICEState(&remoteOfferAnswer)
 
-			h.pqRegisterEndpoint(remoteOfferAnswer.MlkemPort)
-
 			if h.relayListener != nil {
 				h.relayListener.Notify(&remoteOfferAnswer)
 			}
@@ -163,10 +149,6 @@ func (h *Handshaker) Listen(ctx context.Context) {
 			if h.iceListener != nil && h.RemoteICESupported() {
 				h.iceListener(&remoteOfferAnswer)
 			}
-
-			if h.config.PQ != nil {
-				h.config.PQ.OnAnswer(h.config.Key, remoteOfferAnswer.MlkemPayload)
-			}
 		case <-ctx.Done():
 			h.log.Infof("stop listening for remote offers and answers")
 			return
@@ -174,16 +156,6 @@ func (h *Handshaker) Listen(ctx context.Context) {
 	}
 }
 
-// pqRegisterEndpoint feeds the post-quantum handshaker the peer's data-path endpoint
-// (its WG overlay IP plus the advertised pq UDP port) learned from a remote offer/answer.
-func (h *Handshaker) pqRegisterEndpoint(remotePort int) {
-	if h.config.PQ == nil || remotePort <= 0 || remotePort > 65535 || len(h.config.WgConfig.AllowedIps) == 0 {
-		return
-	}
-	addr := netip.AddrPortFrom(h.config.WgConfig.AllowedIps[0].Addr(), uint16(remotePort))
-	h.config.PQ.SetRemoteAddr(h.config.Key, addr)
-}
-
 func (h *Handshaker) SendOffer() error {
 	h.mu.Lock()
 	defer h.mu.Unlock()
@@ -223,23 +195,13 @@ func (h *Handshaker) sendOffer() error {
 	}
 
 	offer := h.buildOfferAnswer()
-	if h.config.PQ != nil {
-		offer.MlkemPayload, offer.MlkemPort = h.config.PQ.OfferPayload(h.config.Key)
-	}
 	h.log.Debugf("sending offer with serial: %s", offer.SessionIDString())
 
 	return h.signaler.SignalOffer(offer, h.config.Key)
 }
 
-func (h *Handshaker) sendAnswer(remoteOffer *OfferAnswer) error {
+func (h *Handshaker) sendAnswer() error {
 	answer := h.buildOfferAnswer()
-	if h.config.PQ != nil {
-		var recvOffer []byte
-		if remoteOffer != nil {
-			recvOffer = remoteOffer.MlkemPayload
-		}
-		answer.MlkemPayload, answer.MlkemPort = h.config.PQ.AnswerPayload(h.config.Key, recvOffer)
-	}
 	h.log.Debugf("sending answer with serial: %s", answer.SessionIDString())
 
 	return h.signaler.SignalAnswer(answer, h.config.Key)

client/internal/peer/signaler.go

@@ -63,8 +63,6 @@ func (s *Signaler) signalOfferAnswer(offerAnswer OfferAnswer, remoteKey string,
 		},
 		RosenpassPubKey: offerAnswer.RosenpassPubKey,
 		RosenpassAddr:   offerAnswer.RosenpassAddr,
-		MlkemPayload:    offerAnswer.MlkemPayload,
-		MlkemPort:       offerAnswer.MlkemPort,
 		RelaySrvAddress: offerAnswer.RelaySrvAddress,
 		RelaySrvIP:      offerAnswer.RelaySrvIP,
 		SessionID:       sessionIDBytes,

client/internal/peer/status.go

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

client/internal/peer/wg_watcher.go

@@ -3,6 +3,7 @@ package peer
 import (
 	"context"
 	"fmt"
+	"sync"
 	"time"
 
 	log "github.com/sirupsen/logrus"
@@ -23,14 +24,14 @@ type WGInterfaceStater interface {
 	GetStats() (map[string]configurer.WGStats, error)
 }
 
-// WGWatcher is single-shot: one instance per connection attempt, run once, then discarded.
-// Lifecycle is owned by Conn under conn.mu, so it keeps no "enabled" state to go stale.
 type WGWatcher struct {
 	log           *log.Entry
 	wgIfaceStater WGInterfaceStater
 	peerKey       string
 	stateDump     *stateDump
 
+	enabled   bool
+	muEnabled sync.Mutex
 	// initialHandshake is not thread-safe; never call PrepareInitialHandshake and EnableWgWatcher concurrently.
 	initialHandshake time.Time
 
@@ -47,23 +48,36 @@ func NewWGWatcher(log *log.Entry, wgIfaceStater WGInterfaceStater, peerKey strin
 	}
 }
 
-// PrepareInitialHandshake reads the peer's current WireGuard handshake time. It must be
-// called before the peer is (re)configured on the WireGuard interface, so the captured
-// baseline reflects the state prior to this connection attempt instead of racing with
-// that configuration.
-func (w *WGWatcher) PrepareInitialHandshake() {
+// PrepareInitialHandshake reserves the watcher and reads the peer's current WireGuard
+// handshake time. It must be called before the peer is (re)configured on the WireGuard
+// interface, so the captured baseline reflects the state prior to this connection attempt
+// instead of racing with that configuration. Returns ok=false if the watcher is already
+// running, in which case EnableWgWatcher must not be called.
+func (w *WGWatcher) PrepareInitialHandshake() (ok bool) {
+	w.muEnabled.Lock()
+	if w.enabled {
+		w.muEnabled.Unlock()
+		return false
+	}
+
 	w.log.Debugf("enable WireGuard watcher")
+	w.enabled = true
+	w.muEnabled.Unlock()
+
 	handshake, _ := w.wgState()
 	w.initialHandshake = handshake
+	return true
 }
 
 // EnableWgWatcher runs the WireGuard watcher loop using the handshake baseline captured by
 // PrepareInitialHandshake. The watcher runs until ctx is cancelled. Caller is responsible
-// for context lifecycle management. onHandshakeSuccessFn is called only for the first
-// handshake observed by this run, onCheckSuccessFn for every check that observed a fresh
-// handshake, including the first.
-func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time), onCheckSuccessFn func()) {
-	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, onCheckSuccessFn, enabledTime, w.initialHandshake)
+// for context lifecycle management.
+func (w *WGWatcher) EnableWgWatcher(ctx context.Context, enabledTime time.Time, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time)) {
+	w.periodicHandshakeCheck(ctx, onDisconnectedFn, onHandshakeSuccessFn, enabledTime, w.initialHandshake)
+
+	w.muEnabled.Lock()
+	w.enabled = false
+	w.muEnabled.Unlock()
 }
 
 // Reset signals the watcher that the WireGuard peer has been reset and a new
@@ -76,7 +90,7 @@ func (w *WGWatcher) Reset() {
 }
 
 // wgStateCheck help to check the state of the WireGuard handshake and relay connection
-func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time), onCheckSuccessFn func(), enabledTime time.Time, initialHandshake time.Time) {
+func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn func(), onHandshakeSuccessFn func(when time.Time), enabledTime time.Time, initialHandshake time.Time) {
 	w.log.Infof("WireGuard watcher started")
 
 	timer := time.NewTimer(wgHandshakeOvertime)
@@ -89,7 +103,6 @@ func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn
 		case <-timer.C:
 			handshake, ok := w.handshakeCheck(lastHandshake)
 			if !ok {
-				// early ctx cancel check return
 				if ctx.Err() != nil {
 					return
 				}
@@ -104,10 +117,6 @@ func (w *WGWatcher) periodicHandshakeCheck(ctx context.Context, onDisconnectedFn
 				}
 			}
 
-			if onCheckSuccessFn != nil && ctx.Err() == nil {
-				onCheckSuccessFn()
-			}
-
 			lastHandshake = *handshake
 
 			resetTime := time.Until(handshake.Add(checkPeriod))
@@ -138,9 +147,9 @@ func (w *WGWatcher) handshakeCheck(lastHandshake time.Time) (*time.Time, bool) {
 
 	w.log.Tracef("previous handshake, handshake: %v, %v", lastHandshake, handshake)
 
-	// the current known handshake did not change
+	// the current know handshake did not change
 	if handshake.Equal(lastHandshake) {
-		w.log.Warnf("WireGuard handshake not updated: %v", handshake)
+		w.log.Warnf("WireGuard handshake timed out: %v", handshake)
 		return nil, false
 	}
 

client/internal/peer/wg_watcher_test.go

@@ -7,6 +7,7 @@ import (
 	"time"
 
 	log "github.com/sirupsen/logrus"
+	"github.com/stretchr/testify/require"
 
 	"github.com/netbirdio/netbird/client/iface/configurer"
 )
@@ -23,72 +24,6 @@ func (m *MocWgIface) disconnect() {
 	m.stop = true
 }
 
-type mockHandshakeStats struct {
-	mu        sync.Mutex
-	handshake time.Time
-}
-
-func (m *mockHandshakeStats) GetStats() (map[string]configurer.WGStats, error) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-	return map[string]configurer.WGStats{"": {LastHandshake: m.handshake}}, nil
-}
-
-func (m *mockHandshakeStats) advance() {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-	m.handshake = time.Now()
-}
-
-// TestWGWatcher_CheckSuccessCallback: onCheckSuccessFn must fire for a fresh
-// handshake even when the watcher started with an existing handshake baseline,
-// the case where onHandshakeSuccessFn stays silent.
-func TestWGWatcher_CheckSuccessCallback(t *testing.T) {
-	// checkPeriod bounds how stale a handshake may be before the watcher treats it
-	// as a suspended-machine timeout. The first check fires after wgHandshakeOvertime,
-	// so keep checkPeriod well above any scheduling jitter to avoid a false timeout
-	// converting the expected success into a disconnect on a loaded runner.
-	checkPeriod = 1 * time.Minute
-	wgHandshakeOvertime = 1 * time.Second
-
-	mlog := log.WithField("peer", "tet")
-	// Use an old baseline so advance() yields a strictly newer handshake even on
-	// platforms with coarse clock resolution (Windows), where two time.Now() calls
-	// microseconds apart can return the same instant and read as a timed-out handshake.
-	stats := &mockHandshakeStats{handshake: time.Now().Add(-time.Hour)}
-	watcher := NewWGWatcher(mlog, stats, "", newStateDump("peer", mlog, &Status{}))
-
-	ctx, cancel := context.WithCancel(context.Background())
-	defer cancel()
-
-	watcher.PrepareInitialHandshake()
-
-	firstHandshake := make(chan struct{}, 1)
-	checkSuccess := make(chan struct{}, 1)
-	go watcher.EnableWgWatcher(ctx, time.Now(), func() {}, func(when time.Time) {
-		firstHandshake <- struct{}{}
-	}, func() {
-		select {
-		case checkSuccess <- struct{}{}:
-		default:
-		}
-	})
-
-	stats.advance()
-
-	select {
-	case <-checkSuccess:
-	case <-time.After(10 * time.Second):
-		t.Errorf("timeout waiting for check success callback")
-	}
-
-	select {
-	case <-firstHandshake:
-		t.Errorf("first-handshake callback must not fire for a non-zero baseline")
-	default:
-	}
-}
-
 func TestWGWatcher_EnableWgWatcher(t *testing.T) {
 	checkPeriod = 5 * time.Second
 	wgHandshakeOvertime = 1 * time.Second
@@ -100,7 +35,8 @@ func TestWGWatcher_EnableWgWatcher(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
 
-	watcher.PrepareInitialHandshake()
+	ok := watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should not be enabled yet")
 
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
@@ -108,7 +44,7 @@ func TestWGWatcher_EnableWgWatcher(t *testing.T) {
 		onDisconnected <- struct{}{}
 	}, func(when time.Time) {
 		mlog.Infof("onHandshakeSuccess: %v", when)
-	}, nil)
+	})
 
 	// wait for initial reading
 	time.Sleep(2 * time.Second)
@@ -130,13 +66,14 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	watcher := NewWGWatcher(mlog, mocWgIface, "", newStateDump("peer", mlog, &Status{}))
 
 	ctx, cancel := context.WithCancel(context.Background())
-	watcher.PrepareInitialHandshake()
+	ok := watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should not be enabled yet")
 
 	wg := &sync.WaitGroup{}
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
-		watcher.EnableWgWatcher(ctx, time.Now(), func() {}, func(when time.Time) {}, nil)
+		watcher.EnableWgWatcher(ctx, time.Now(), func() {}, func(when time.Time) {})
 	}()
 	cancel()
 
@@ -146,12 +83,13 @@ func TestWGWatcher_ReEnable(t *testing.T) {
 	ctx, cancel = context.WithCancel(context.Background())
 	defer cancel()
 
-	watcher.PrepareInitialHandshake()
+	ok = watcher.PrepareInitialHandshake()
+	require.True(t, ok, "watcher should be re-enabled after the previous run stopped")
 
 	onDisconnected := make(chan struct{}, 1)
 	go watcher.EnableWgWatcher(ctx, time.Now(), func() {
 		onDisconnected <- struct{}{}
-	}, func(when time.Time) {}, nil)
+	}, func(when time.Time) {})
 
 	time.Sleep(2 * time.Second)
 	mocWgIface.disconnect()

client/internal/pqkem/bench_test.go

@@ -1,59 +0,0 @@
-package pqkem
-
-import (
-	"crypto/ecdh"
-	"crypto/mlkem"
-	"crypto/rand"
-	"testing"
-)
-
-func BenchmarkX25519Keygen(b *testing.B) {
-	c := ecdh.X25519()
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if _, err := c.GenerateKey(rand.Reader); err != nil {
-			b.Fatal(err)
-		}
-	}
-}
-
-func BenchmarkX25519ECDH(b *testing.B) {
-	c := ecdh.X25519()
-	a, _ := c.GenerateKey(rand.Reader)
-	p, _ := c.GenerateKey(rand.Reader)
-	pub := p.PublicKey()
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if _, err := a.ECDH(pub); err != nil {
-			b.Fatal(err)
-		}
-	}
-}
-
-func BenchmarkMLKEMKeygen(b *testing.B) {
-	for i := 0; i < b.N; i++ {
-		if _, err := mlkem.GenerateKey768(); err != nil {
-			b.Fatal(err)
-		}
-	}
-}
-
-func BenchmarkMLKEMEncaps(b *testing.B) {
-	dk, _ := mlkem.GenerateKey768()
-	ek := dk.EncapsulationKey()
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		_, _ = ek.Encapsulate()
-	}
-}
-
-func BenchmarkMLKEMDecaps(b *testing.B) {
-	dk, _ := mlkem.GenerateKey768()
-	_, ct := dk.EncapsulationKey().Encapsulate()
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		if _, err := dk.Decapsulate(ct); err != nil {
-			b.Fatal(err)
-		}
-	}
-}

client/internal/pqkem/callbacks.go

@@ -1,18 +0,0 @@
-package pqkem
-
-// CallbackHandler is implemented by the host and invoked by the library. The
-// library only reports events; the host owns the reaction. Keeping this an
-// interface — rather than touching the transport or keying directly — is what lets
-// the KEM code be extracted as a standalone library.
-type CallbackHandler interface {
-	// OnNewPSKReady fires when a fresh post-quantum PSK has been derived for a peer
-	// and must be programmed into the consumer's secure channel. It is invoked at
-	// the commit point of each side: the initiator on receiving the answer, the
-	// responder on receiving the confirm.
-	OnNewPSKReady(remoteID RemoteID, psk PSK) error
-
-	// OnRekeyFailed fires when an exchange fails to converge within the allotted
-	// time. The host should tear the peer connection down so it re-establishes, and
-	// log a WARN. The library reports the event; it does not dictate the reaction.
-	OnRekeyFailed(remoteID RemoteID) error
-}

client/internal/pqkem/convergence.go

@@ -1,224 +0,0 @@
-package pqkem
-
-import (
-	"context"
-	"time"
-)
-
-// startExchange creates a fresh initiator exchange (acknowledging ackID, zero for a
-// bootstrap) and returns the framed offer for the caller to send — pushed over the
-// data path for a chained rekey, or handed to the host for signalling when viaSignal
-// is set. Any previous in-flight exchange for the peer is cancelled.
-func (m *Manager) startExchange(remoteID RemoteID, viaSignal bool, ackID ExchangeID) ([]byte, error) {
-	init, err := NewInitiator()
-	if err != nil {
-		return nil, err
-	}
-	id, err := newExchangeID()
-	if err != nil {
-		return nil, err
-	}
-	raw, err := (&OfferMsg{ExchangeID: id, AckID: ackID, KEMOffer: init.Offer()}).Encode()
-	if err != nil {
-		return nil, err
-	}
-
-	ctx, cancel := context.WithCancel(m.rootCtx)
-	m.mu.Lock()
-	if old := m.exchanges[remoteID]; old != nil && old.cancel != nil {
-		old.cancel()
-	}
-	m.exchanges[remoteID] = &exchangeCtl{
-		id:        id,
-		state:     stateAwaitingAnswer,
-		startedAt: time.Now(),
-		cancel:    cancel,
-		lastSent:  raw,
-		initiator: init,
-		viaSignal: viaSignal,
-	}
-	m.mu.Unlock()
-
-	m.wait.Add(1)
-	go m.initiatorLoop(ctx, remoteID, id)
-	return raw, nil
-}
-
-// processOffer (responder) first acknowledges the previous exchange the offer names
-// (that offer riding the data path under the freshly adopted key proves it worked),
-// then derives the PSK for the new offer, commits it optimistically, and returns the
-// framed answer. A duplicate offer returns the cached answer without re-deriving.
-func (m *Manager) processOffer(remoteID RemoteID, o *OfferMsg) ([]byte, error) {
-	if o.AckID != (ExchangeID{}) {
-		m.ackConverged(remoteID, o.AckID)
-	}
-
-	m.mu.Lock()
-	if ex := m.exchanges[remoteID]; ex != nil && ex.id == o.ExchangeID {
-		state, last := ex.state, ex.lastSent
-		m.mu.Unlock()
-		if state == stateReserved {
-			return nil, nil
-		}
-		return last, nil
-	}
-	// Reserve the slot so a concurrent duplicate offer bails.
-	m.exchanges[remoteID] = &exchangeCtl{id: o.ExchangeID, state: stateReserved, startedAt: time.Now()}
-	m.mu.Unlock()
-
-	answerBytes, psk, err := Respond(o.KEMOffer, m.binding(remoteID))
-	if err != nil {
-		return nil, err
-	}
-	raw, err := (&AnswerMsg{ExchangeID: o.ExchangeID, KEMAnswer: answerBytes}).Encode()
-	if err != nil {
-		return nil, err
-	}
-
-	m.mu.Lock()
-	ex := m.exchanges[remoteID]
-	if ex == nil || ex.id != o.ExchangeID {
-		m.mu.Unlock()
-		return nil, nil
-	}
-	ex.state = stateAwaitingAck
-	ex.lastSent = raw
-	ex.pendingPSK = psk
-	m.psks[remoteID] = psk
-	m.mu.Unlock()
-
-	// Commit optimistically so our data path can rekey to the new PSK.
-	if err := m.cbHandler.OnNewPSKReady(remoteID, psk); err != nil {
-		return nil, err
-	}
-	return raw, nil
-}
-
-// processAnswer (initiator) derives and commits the PSK and parks in
-// stateAwaitingRekey; the next offer (chained from OnDataPathRekeyed) will acknowledge
-// this exchange. Only valid in stateAwaitingAnswer; advancing the state under the
-// lock makes a concurrent/duplicate answer bail.
-func (m *Manager) processAnswer(remoteID RemoteID, a *AnswerMsg) error {
-	m.mu.Lock()
-	ex := m.exchanges[remoteID]
-	if ex == nil || ex.id != a.ExchangeID || ex.state != stateAwaitingAnswer {
-		m.mu.Unlock()
-		return nil
-	}
-	ex.state = stateAwaitingRekey
-	init := ex.initiator
-	ex.initiator = nil
-	m.mu.Unlock()
-
-	psk, err := init.Finish(a.KEMAnswer, m.binding(remoteID))
-	if err != nil {
-		return err
-	}
-
-	// The initiator has converged: the responder must have derived the key to answer.
-	m.mu.Lock()
-	m.established[remoteID] = true
-	m.failures[remoteID] = 0
-	m.psks[remoteID] = psk
-	m.mu.Unlock()
-
-	return m.cbHandler.OnNewPSKReady(remoteID, psk)
-}
-
-// ackConverged (responder) records convergence of the exchange named by ackID: a
-// later offer acknowledging it proves both sides operate on that exchange's key. Only
-// acts on a matching stateAwaitingAck exchange; anything else is ignored.
-func (m *Manager) ackConverged(remoteID RemoteID, ackID ExchangeID) {
-	m.mu.Lock()
-	ex := m.exchanges[remoteID]
-	if ex == nil || ex.id != ackID || ex.state != stateAwaitingAck {
-		m.mu.Unlock()
-		return
-	}
-	delete(m.exchanges, remoteID)
-	m.established[remoteID] = true
-	m.failures[remoteID] = 0
-	_ = time.Since(ex.startedAt) // convergence latency (metrics hook, later step)
-	m.mu.Unlock()
-}
-
-// initiatorLoop enforces the offer->answer convergence deadline and retransmits the
-// initiator's outstanding data-path offer while awaiting the answer (a
-// signalling-bootstrapped offer is retransmitted by the host, so it is not resent
-// here). Exhausting the deadline before the answer arrives is a failure. Once the
-// answer is in (state past awaitingAnswer) the loop exits: the next rotation is driven
-// by OnDataPathRekeyed, and the idle wait for it has no deadline.
-func (m *Manager) initiatorLoop(ctx context.Context, remoteID RemoteID, id ExchangeID) {
-	defer m.wait.Done()
-	t := time.NewTicker(m.retryInterval)
-	defer t.Stop()
-
-	attempts := 0
-	for {
-		select {
-		case <-ctx.Done():
-			return
-		case <-t.C:
-			m.mu.Lock()
-			ex := m.exchanges[remoteID]
-			if ex == nil || ex.id != id {
-				m.mu.Unlock()
-				return
-			}
-
-			switch ex.state {
-			case stateAwaitingAnswer:
-				if attempts >= m.maxRetries {
-					delete(m.exchanges, remoteID)
-					fail := m.registerFailureLocked(remoteID)
-					m.mu.Unlock()
-					m.raiseFailure(remoteID, fail)
-					return
-				}
-				viaSignal := ex.viaSignal
-				msg := ex.lastSent
-				attempts++
-				m.mu.Unlock()
-				if !viaSignal {
-					if err := m.pushDataPath(remoteID, msg); err != nil {
-						m.logger.Warn("pqkem: offer retransmit failed", "peer", remoteID, "err", err)
-					}
-				}
-
-			default:
-				// Past awaiting the answer (converged) or superseded: the loop's job
-				// is done. The next rotation is driven externally by OnDataPathRekeyed,
-				// so there is no deadline while idle-waiting for it (that wait can be
-				// as long as the transport's natural rekey interval).
-				m.mu.Unlock()
-				return
-			}
-		}
-	}
-}
-
-// registerFailureLocked applies policy B and reports whether OnRekeyFailed is due:
-// an initial exchange (peer never established) fails immediately; a rekey tolerates
-// up to maxRekeyFailures consecutive misses (we stay on the still-valid previous
-// PSK) before failing. Assumes m.mu is held.
-func (m *Manager) registerFailureLocked(remoteID RemoteID) bool {
-	if !m.established[remoteID] {
-		return true
-	}
-	m.failures[remoteID]++
-	if m.failures[remoteID] >= m.maxRekeyFailures {
-		m.failures[remoteID] = 0
-		return true
-	}
-	return false
-}
-
-func (m *Manager) raiseFailure(remoteID RemoteID, fail bool) {
-	if !fail {
-		m.logger.Warn("pqkem: rekey attempt timed out, will retry next cycle", "peer", remoteID)
-		return
-	}
-	if err := m.cbHandler.OnRekeyFailed(remoteID); err != nil {
-		m.logger.Error("pqkem: OnRekeyFailed handler error", "peer", remoteID, "err", err)
-	}
-}

client/internal/pqkem/convergence_test.go

@@ -1,74 +0,0 @@
-package pqkem
-
-import (
-	"net/netip"
-	"testing"
-	"time"
-
-	"github.com/stretchr/testify/require"
-)
-
-// dropTransport is a pqkem.Transport that silently discards everything.
-type dropTransport struct{}
-
-func (dropTransport) Send(netip.AddrPort, []byte) error { return nil }
-func (dropTransport) LocalPort() int                    { return 0 }
-func (dropTransport) Run(func(netip.AddrPort, []byte))  {}
-func (dropTransport) Close() error                      { return nil }
-
-func failedCount(f *fakeWG) int {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	return len(f.failed)
-}
-
-func TestManager_InitialTimeoutFailsImmediately(t *testing.T) {
-	wg := newFakeWG()
-	d := NewManager("bbbb", wg, nil) // bbbb > aaaa -> initiator
-	d.Start(dropTransport{})
-	d.retryInterval = 5 * time.Millisecond
-	d.maxRetries = 3
-	defer d.Stop()
-
-	// Bootstrap offer is produced for signalling; no answer ever comes back -> the
-	// initial exchange fails fast.
-	offer, err := d.SignalOffer("aaaa")
-	require.NoError(t, err)
-	require.NotNil(t, offer)
-
-	require.Eventually(t, func() bool { return failedCount(wg) == 1 }, time.Second, 5*time.Millisecond)
-}
-
-func TestManager_RekeyToleratesKFailures(t *testing.T) {
-	dA, dB, _, wgB, lbB := pair(t)
-	defer dA.Stop()
-	defer dB.Stop()
-
-	// Tighten B's timings before any exchange loop spawns (the loop reads these
-	// fields, so writing them after a loop is running would race).
-	dB.retryInterval = 5 * time.Millisecond
-	dB.maxRetries = 2
-
-	// Establish: bootstrap + data-path-rekeyed so B becomes established and its data
-	// path is usable.
-	bootstrap(t, dA, dB)
-	dA.OnDataPathRekeyed("bbbb")
-	dB.OnDataPathRekeyed("aaaa")
-	require.NotEqual(t, PSK{}, wgB.psk("aaaa"))
-
-	// Drop B's outbound so rekeys can no longer converge.
-	lbB.drop.Store(true)
-
-	// K-1 data-path rekeys must NOT raise OnRekeyFailed.
-	for i := 0; i < DefaultMaxRekeyFailures-1; i++ {
-		_, err := dB.startExchange("aaaa", false, ExchangeID{})
-		require.NoError(t, err)
-		time.Sleep(50 * time.Millisecond)
-	}
-	require.Equal(t, 0, failedCount(wgB), "no failure before K attempts")
-
-	// The K-th failure raises it once.
-	_, err := dB.startExchange("aaaa", false, ExchangeID{})
-	require.NoError(t, err)
-	require.Eventually(t, func() bool { return failedCount(wgB) == 1 }, time.Second, 5*time.Millisecond)
-}

client/internal/pqkem/env.go

@@ -1,59 +0,0 @@
-package pqkem
-
-import (
-	"log/slog"
-	"os"
-	"strconv"
-	"strings"
-
-	log "github.com/sirupsen/logrus"
-)
-
-// EnvEnabled is the environment variable that turns the ML-KEM post-quantum
-// exchange on for this client. Accepts on/off aliases plus anything
-// strconv.ParseBool understands (true/false/1/0).
-const EnvEnabled = "NB_ENABLE_PQ_MLKEM"
-
-// Enabled reports whether the ML-KEM PQ exchange is enabled via the environment.
-// An empty or unrecognized value is treated as disabled.
-func Enabled() bool {
-	raw := strings.ToLower(strings.TrimSpace(os.Getenv(EnvEnabled)))
-	switch raw {
-	case "":
-		return false
-	case "on":
-		return true
-	case "off":
-		return false
-	}
-	enabled, err := strconv.ParseBool(raw)
-	if err != nil {
-		log.Warnf("failed to parse %s value %q: %v", EnvEnabled, raw, err)
-		return false
-	}
-	return enabled
-}
-
-// EnvLogLevel overrides the ML-KEM manager's slog level (debug/info/warn/error).
-// Defaults to info.
-const EnvLogLevel = "NB_PQ_MLKEM_LOG_LEVEL"
-
-// NewLogger builds the slog logger for the ML-KEM manager: a text handler to stdout
-// at the level from EnvLogLevel. Mirrors the Rosenpass manager's logger setup so PQ
-// components log consistently.
-func NewLogger() *slog.Logger {
-	return slog.New(slog.NewTextHandler(os.Stdout, &slog.HandlerOptions{Level: logLevel()}))
-}
-
-func logLevel() slog.Level {
-	switch strings.ToLower(strings.TrimSpace(os.Getenv(EnvLogLevel))) {
-	case "debug":
-		return slog.LevelDebug
-	case "warn":
-		return slog.LevelWarn
-	case "error":
-		return slog.LevelError
-	default:
-		return slog.LevelInfo
-	}
-}

client/internal/pqkem/kem.go

@@ -1,167 +0,0 @@
-// Package pqkem is a spike (NET-1406) for a post-quantum pre-shared-key exchange
-// that could replace Rosenpass. It performs an X25519MLKEM768 hybrid key
-// encapsulation and derives a 32-byte pre-shared key (PSK).
-//
-// The exchange is a single round trip designed to ride the (already
-// authenticated) Signal offer/answer channel:
-//
-//	initiator --Offer(1216B)--> responder
-//	initiator <--Answer(1120B)-- responder
-//
-// Both sides then hold the same PSK, which is bound to the two peers' identities
-// (their peer identity keys) so the derived key cannot be transplanted
-// to a different peer pair even if the transport authentication were bypassed.
-//
-// Combiner note: this follows draft-ietf-tls-ecdhe-mlkem for X25519MLKEM768 — on
-// the wire ML-KEM ‖ X25519 (the draft deliberately reversed the share order for
-// this group), and ML-KEM_ss ‖ X25519_ss fed into the KDF. The spike uses SHA-256
-// (also binding the transcript and peer identities); a production version should
-// use HKDF — see TODO below.
-package pqkem
-
-import (
-	"crypto/ecdh"
-	"crypto/mlkem"
-	"crypto/rand"
-	"crypto/sha256"
-	"fmt"
-)
-
-const (
-	// OfferSize is the initiator message: ML-KEM-768 encapsulation key ‖ X25519 public key
-	// (share order per draft-ietf-tls-ecdhe-mlkem for X25519MLKEM768).
-	OfferSize = mlkem.EncapsulationKeySize768 + 32 // 1216
-	// AnswerSize is the responder message: ML-KEM-768 ciphertext ‖ X25519 public key.
-	AnswerSize = mlkem.CiphertextSize768 + 32 // 1120
-
-	pskLabel = "netbird-pq-psk-v1"
-)
-
-// PSK is the 32-byte derived pre-shared key handed to the consumer to key its channel.
-type PSK [32]byte
-
-// Binding identifies the peer pair the PSK is derived for. Callers set both
-// peer identity keys; the order does not matter (it is canonicalised).
-type Binding struct {
-	LocalID  []byte
-	RemoteID []byte
-}
-
-// Initiator holds the ephemeral secrets between Offer and Finish.
-type Initiator struct {
-	x25519  *ecdh.PrivateKey
-	mlkemDK *mlkem.DecapsulationKey768
-	offer   []byte
-}
-
-// NewInitiator generates the ephemeral X25519 + ML-KEM-768 keypairs.
-func NewInitiator() (*Initiator, error) {
-	x, err := ecdh.X25519().GenerateKey(rand.Reader)
-	if err != nil {
-		return nil, fmt.Errorf("x25519 keygen: %w", err)
-	}
-	dk, err := mlkem.GenerateKey768()
-	if err != nil {
-		return nil, fmt.Errorf("ml-kem keygen: %w", err)
-	}
-
-	offer := make([]byte, 0, OfferSize)
-	offer = append(offer, dk.EncapsulationKey().Bytes()...)
-	offer = append(offer, x.PublicKey().Bytes()...)
-
-	return &Initiator{x25519: x, mlkemDK: dk, offer: offer}, nil
-}
-
-// Offer returns the initiator message to send over Signal.
-func (i *Initiator) Offer() []byte {
-	return i.offer
-}
-
-// Finish consumes the responder's answer and derives the PSK.
-func (i *Initiator) Finish(answer []byte, b Binding) (PSK, error) {
-	if len(answer) != AnswerSize {
-		return PSK{}, fmt.Errorf("answer: got %d bytes, want %d", len(answer), AnswerSize)
-	}
-	ct := answer[:mlkem.CiphertextSize768]
-	peerX := answer[mlkem.CiphertextSize768:]
-
-	ssMLKEM, err := i.mlkemDK.Decapsulate(ct)
-	if err != nil {
-		return PSK{}, fmt.Errorf("ml-kem decapsulate: %w", err)
-	}
-	pub, err := ecdh.X25519().NewPublicKey(peerX)
-	if err != nil {
-		return PSK{}, fmt.Errorf("parse peer x25519: %w", err)
-	}
-	ssX, err := i.x25519.ECDH(pub)
-	if err != nil {
-		return PSK{}, fmt.Errorf("x25519 ecdh: %w", err)
-	}
-
-	return derivePSK(ssMLKEM, ssX, i.offer, answer, b), nil
-}
-
-// Respond consumes an initiator offer, produces the answer, and derives the PSK.
-func Respond(offer []byte, b Binding) (answer []byte, psk PSK, err error) {
-	if len(offer) != OfferSize {
-		return nil, PSK{}, fmt.Errorf("offer: got %d bytes, want %d", len(offer), OfferSize)
-	}
-	peerEK := offer[:mlkem.EncapsulationKeySize768]
-	peerX := offer[mlkem.EncapsulationKeySize768:]
-
-	ek, err := mlkem.NewEncapsulationKey768(peerEK)
-	if err != nil {
-		return nil, PSK{}, fmt.Errorf("parse peer ml-kem key: %w", err)
-	}
-	ssMLKEM, ct := ek.Encapsulate()
-
-	x, err := ecdh.X25519().GenerateKey(rand.Reader)
-	if err != nil {
-		return nil, PSK{}, fmt.Errorf("x25519 keygen: %w", err)
-	}
-	pub, err := ecdh.X25519().NewPublicKey(peerX)
-	if err != nil {
-		return nil, PSK{}, fmt.Errorf("parse peer x25519: %w", err)
-	}
-	ssX, err := x.ECDH(pub)
-	if err != nil {
-		return nil, PSK{}, fmt.Errorf("x25519 ecdh: %w", err)
-	}
-
-	answer = make([]byte, 0, AnswerSize)
-	answer = append(answer, ct...)
-	answer = append(answer, x.PublicKey().Bytes()...)
-
-	// derivePSK uses the same argument order on both sides; the responder's local
-	// binding is the mirror of the initiator's, canonicalised inside derivePSK.
-	return answer, derivePSK(ssMLKEM, ssX, offer, answer, b), nil
-}
-
-// derivePSK combines the two shared secrets and binds the result to the full
-// transcript (offer ‖ answer) and the canonicalised peer identities.
-//
-// TODO(NET-1406): replace the SHA-256 concat with the RFC HKDF combiner
-// (crypto/hkdf, Go 1.24+) and proper labels before this leaves spike status.
-func derivePSK(ssMLKEM, ssX, offer, answer []byte, b Binding) PSK {
-	lo, hi := canonicalPair(b.LocalID, b.RemoteID)
-
-	h := sha256.New()
-	h.Write([]byte(pskLabel))
-	h.Write(ssMLKEM)
-	h.Write(ssX)
-	h.Write(offer)
-	h.Write(answer)
-	h.Write(lo)
-	h.Write(hi)
-
-	var psk PSK
-	copy(psk[:], h.Sum(nil))
-	return psk
-}
-
-func canonicalPair(a, b []byte) (lo, hi []byte) {
-	if string(a) <= string(b) {
-		return a, b
-	}
-	return b, a
-}

client/internal/pqkem/kem_test.go

@@ -1,89 +0,0 @@
-package pqkem
-
-import (
-	"testing"
-	"time"
-
-	"github.com/stretchr/testify/require"
-)
-
-var (
-	wgA = []byte("peer-A-wireguard-pubkey-32bytes!")
-	wgB = []byte("peer-B-wireguard-pubkey-32bytes!")
-)
-
-func TestExchange_DerivesMatchingPSK(t *testing.T) {
-	init, err := NewInitiator()
-	require.NoError(t, err)
-
-	require.Len(t, init.Offer(), OfferSize)
-
-	answer, pskB, err := Respond(init.Offer(), Binding{LocalID: wgB, RemoteID: wgA})
-	require.NoError(t, err)
-	require.Len(t, answer, AnswerSize)
-
-	pskA, err := init.Finish(answer, Binding{LocalID: wgA, RemoteID: wgB})
-	require.NoError(t, err)
-
-	require.Equal(t, pskB, pskA, "both sides must derive the same PSK")
-	require.NotEqual(t, PSK{}, pskA, "PSK must not be zero")
-}
-
-func TestExchange_PSKBoundToPeerIdentities(t *testing.T) {
-	init, err := NewInitiator()
-	require.NoError(t, err)
-
-	// responder computes with the honest pair...
-	_, pskHonest, err := Respond(init.Offer(), Binding{LocalID: wgB, RemoteID: wgA})
-	require.NoError(t, err)
-
-	// ...a second responder run with a different peer identity yields a different PSK,
-	// even though the KEM material would otherwise combine identically.
-	wgC := []byte("peer-C-wireguard-pubkey-32bytes!")
-	_, pskWrong, err := Respond(init.Offer(), Binding{LocalID: wgC, RemoteID: wgA})
-	require.NoError(t, err)
-
-	require.NotEqual(t, pskHonest, pskWrong, "PSK must be bound to the peer pair")
-}
-
-func TestExchange_RejectsMalformedMessages(t *testing.T) {
-	init, err := NewInitiator()
-	require.NoError(t, err)
-
-	_, _, err = Respond(init.Offer()[:10], Binding{})
-	require.Error(t, err)
-
-	_, err = init.Finish([]byte("too short"), Binding{})
-	require.Error(t, err)
-}
-
-// TestExchange_ReportSizesAndTiming is a spike measurement, not a pass/fail gate.
-// Run with: go test -run TestExchange_ReportSizesAndTiming -v ./client/internal/pqkem/
-func TestExchange_ReportSizesAndTiming(t *testing.T) {
-	const iters = 200
-
-	var tInit, tResp, tFinish time.Duration
-	for i := 0; i < iters; i++ {
-		s0 := time.Now()
-		init, err := NewInitiator()
-		require.NoError(t, err)
-		tInit += time.Since(s0)
-
-		s1 := time.Now()
-		answer, _, err := Respond(init.Offer(), Binding{LocalID: wgB, RemoteID: wgA})
-		require.NoError(t, err)
-		tResp += time.Since(s1)
-
-		s2 := time.Now()
-		_, err = init.Finish(answer, Binding{LocalID: wgA, RemoteID: wgB})
-		require.NoError(t, err)
-		tFinish += time.Since(s2)
-	}
-
-	t.Logf("wire sizes:   offer=%d B   answer=%d B   (Rosenpass static pubkey ~524160 B)", OfferSize, AnswerSize)
-	t.Logf("total on-wire per handshake: %d B  (~%.0fx smaller than RP static key)", OfferSize+AnswerSize, 524160.0/float64(OfferSize+AnswerSize))
-	t.Logf("avg NewInitiator (keygen): %s", tInit/iters)
-	t.Logf("avg Respond (encaps+dh):   %s", tResp/iters)
-	t.Logf("avg Finish (decaps+dh):    %s", tFinish/iters)
-	t.Logf("avg full handshake CPU:    %s", (tInit+tResp+tFinish)/iters)
-}

client/internal/pqkem/manager.go

@@ -1,372 +0,0 @@
-package pqkem
-
-import (
-	"context"
-	"crypto/rand"
-	"fmt"
-	"log/slog"
-	"net/netip"
-	"sync"
-	"time"
-)
-
-const (
-	// DefaultRetryInterval is how often the initiator retransmits its outstanding
-	// data-path offer while awaiting the answer.
-	DefaultRetryInterval = 2 * time.Second
-	// DefaultMaxRetries bounds how many ticks an exchange may run before it is
-	// declared failed. The convergence deadline is thus MaxRetries * RetryInterval.
-	DefaultMaxRetries = 10
-	// DefaultMaxRekeyFailures is how many consecutive rekey (non-initial) failures
-	// are tolerated before OnRekeyFailed. The initial exchange fails immediately.
-	DefaultMaxRekeyFailures = 3
-)
-
-// LocalID and RemoteID are peer identity keys (e.g. WireGuard public keys). They are
-// distinct types so the local and a remote identity cannot be mixed up.
-type (
-	LocalID  string
-	RemoteID string
-)
-
-// Transport is the data-path socket the Manager drives (the analogue of
-// go-rosenpass's Conn). It is a dumb mover of bytes to/from endpoints: the Manager
-// owns the remoteID<->endpoint routing and hands the transport a resolved endpoint
-// to Send, and reverse-resolves the source of each inbound datagram. Its lifecycle
-// belongs to the Manager (Run at Start, Close at Stop).
-type Transport interface {
-	// Send delivers msg to the given data-path endpoint.
-	Send(endpoint netip.AddrPort, msg []byte) error
-	// LocalPort is the bound local UDP port, announced to peers so they know where
-	// to send data-path messages.
-	LocalPort() int
-	// Run starts delivering inbound datagrams as (source endpoint, msg) to onInbound
-	// and returns immediately; it runs until Close.
-	Run(onInbound func(src netip.AddrPort, msg []byte))
-	// Close stops delivery and releases the socket.
-	Close() error
-}
-
-// exchangeState is the single source of truth for an exchange's role and phase.
-type exchangeState uint8
-
-const (
-	stateReserved       exchangeState = iota // responder: deriving the answer
-	stateAwaitingAnswer                      // initiator: offer sent, awaiting the answer
-	stateAwaitingRekey                       // initiator: PSK derived+set, awaiting OnDataPathRekeyed to chain the next offer
-	stateAwaitingAck                         // responder: answer sent, awaiting the next offer that acks this exchange
-)
-
-// exchangeCtl holds all state for one in-flight exchange with a peer, under the
-// Manager's single lock. state drives every decision. lastSent is the current
-// data-path retransmit payload (the offer, for the initiator). initiator is the
-// ephemeral handle used at Finish; pendingPSK is the responder's derived key.
-// viaSignal records that the offer went to the host for the signalling channel, so
-// the loop does not retransmit it on the data path. Only the initiator runs a
-// retransmit loop, so only it sets cancel.
-type exchangeCtl struct {
-	id         ExchangeID
-	state      exchangeState
-	startedAt  time.Time
-	cancel     context.CancelFunc
-	lastSent   []byte
-	initiator  *Initiator
-	pendingPSK PSK
-	viaSignal  bool
-}
-
-// Manager is the stateful orchestrator — the analogue of go-rosenpass's Server. It
-// drives the X25519MLKEM768 exchange, owns the peer endpoint routing and the data-path
-// transport, and surfaces the derived PSK and convergence to the host via
-// CallbackHandler. It is event-driven: the bootstrap is triggered by the host
-// (SignalOffer) and each rotation is clocked by OnDataPathRekeyed. The cryptography is
-// the pure kem.go primitives; all state lives here under one lock.
-type Manager struct {
-	localID   LocalID
-	cbHandler CallbackHandler
-	logger    *slog.Logger
-
-	retryInterval    time.Duration
-	maxRetries       int
-	maxRekeyFailures int
-
-	rootCtx    context.Context
-	rootCancel context.CancelFunc
-
-	mu          sync.Mutex
-	transport   Transport
-	exchanges   map[RemoteID]*exchangeCtl   // in-flight exchange per peer
-	established map[RemoteID]bool           // peer has completed at least one exchange
-	failures    map[RemoteID]int            // consecutive rekey failures per peer
-	psks        map[RemoteID]PSK            // latest derived PSK per peer (pulled at WG peer-config time)
-	peerAddrs   map[RemoteID]netip.AddrPort // remoteID -> data-path endpoint (send routing)
-	peersByAddr map[netip.AddrPort]RemoteID // reverse: source endpoint -> remoteID (inbound)
-	wait        sync.WaitGroup
-}
-
-// NewManager builds a manager for the local peer identified by its peer identity key
-// (used for the deterministic initiator role and the identity binding). A nil logger
-// falls back to slog.Default(). Install the data-path transport with Start.
-func NewManager(localID LocalID, h CallbackHandler, logger *slog.Logger) *Manager {
-	if logger == nil {
-		logger = slog.Default()
-	}
-	ctx, cancel := context.WithCancel(context.Background())
-	return &Manager{
-		localID:          localID,
-		cbHandler:        h,
-		logger:           logger,
-		retryInterval:    DefaultRetryInterval,
-		maxRetries:       DefaultMaxRetries,
-		maxRekeyFailures: DefaultMaxRekeyFailures,
-		rootCtx:          ctx,
-		rootCancel:       cancel,
-		exchanges:        make(map[RemoteID]*exchangeCtl),
-		established:      make(map[RemoteID]bool),
-		failures:         make(map[RemoteID]int),
-		psks:             make(map[RemoteID]PSK),
-		peerAddrs:        make(map[RemoteID]netip.AddrPort),
-		peersByAddr:      make(map[netip.AddrPort]RemoteID),
-	}
-}
-
-// Start installs the data-path transport and begins its inbound delivery. The Manager
-// owns it from here; Stop closes it. Start/Stop are the transport lifecycle pair.
-func (m *Manager) Start(t Transport) {
-	m.mu.Lock()
-	m.transport = t
-	m.mu.Unlock()
-	if t != nil {
-		t.Run(m.onDataPathInbound)
-	}
-}
-
-// LocalPort is the data-path transport's bound UDP port (0 if no transport), to be
-// announced to peers.
-func (m *Manager) LocalPort() int {
-	m.mu.Lock()
-	t := m.transport
-	m.mu.Unlock()
-	if t == nil {
-		return 0
-	}
-	return t.LocalPort()
-}
-
-// IsInitiator reports whether the local peer drives the exchange for this remote
-// peer. Roles are deterministic (lexicographic identity-key compare) so exactly one
-// side initiates, mirroring how Rosenpass picks its handshake initiator.
-func (m *Manager) IsInitiator(remoteID RemoteID) bool {
-	return string(m.localID) > string(remoteID)
-}
-
-// PSK returns the latest PSK derived for the peer, for the host to program at WG
-// peer-config time (the pull path). ok is false until an exchange has derived one.
-func (m *Manager) PSK(remoteID RemoteID) (PSK, bool) {
-	m.mu.Lock()
-	defer m.mu.Unlock()
-	psk, ok := m.psks[remoteID]
-	return psk, ok
-}
-
-// AddPeer registers where a peer's data-path messages are sent and received: its
-// overlay endpoint (IP:port). Re-adding updates the endpoint.
-func (m *Manager) AddPeer(remoteID RemoteID, endpoint netip.AddrPort) {
-	if !endpoint.IsValid() {
-		return
-	}
-	m.mu.Lock()
-	if old, ok := m.peerAddrs[remoteID]; ok {
-		delete(m.peersByAddr, old)
-	}
-	m.peerAddrs[remoteID] = endpoint
-	m.peersByAddr[endpoint] = remoteID
-	m.mu.Unlock()
-}
-
-// RemovePeer stops any in-flight exchange for a peer and drops its state and routing.
-func (m *Manager) RemovePeer(remoteID RemoteID) {
-	m.mu.Lock()
-	if ex, ok := m.exchanges[remoteID]; ok {
-		if ex.cancel != nil {
-			ex.cancel()
-		}
-		delete(m.exchanges, remoteID)
-	}
-	delete(m.established, remoteID)
-	delete(m.failures, remoteID)
-	delete(m.psks, remoteID)
-	if ep, ok := m.peerAddrs[remoteID]; ok {
-		delete(m.peersByAddr, ep)
-		delete(m.peerAddrs, remoteID)
-	}
-	m.mu.Unlock()
-}
-
-// Stop cancels all in-flight exchanges, closes the transport, and waits for the
-// exchange goroutines to exit.
-func (m *Manager) Stop() {
-	m.rootCancel()
-	m.wait.Wait()
-	m.mu.Lock()
-	t := m.transport
-	m.transport = nil
-	m.exchanges = make(map[RemoteID]*exchangeCtl)
-	m.psks = make(map[RemoteID]PSK)
-	m.mu.Unlock()
-	if t != nil {
-		if err := t.Close(); err != nil {
-			m.logger.Warn("pqkem: closing data-path transport", "err", err)
-		}
-	}
-}
-
-// ---- Signalling channel (host-driven; rides the host's negotiation) ----
-
-// SignalOffer returns the KEM offer for the host to embed in its outgoing offer to
-// remoteID (bootstrap). It returns (nil, nil) when the local peer is not the
-// initiator. It is idempotent for an in-flight bootstrap: a repeat call returns the
-// same offer rather than starting a new exchange.
-func (m *Manager) SignalOffer(remoteID RemoteID) ([]byte, error) {
-	if !m.IsInitiator(remoteID) {
-		return nil, nil
-	}
-	m.mu.Lock()
-	if ex := m.exchanges[remoteID]; ex != nil && ex.viaSignal && ex.state == stateAwaitingAnswer {
-		last := ex.lastSent
-		m.mu.Unlock()
-		return last, nil
-	}
-	m.mu.Unlock()
-	// bootstrap offer acknowledges nothing (zero AckID).
-	return m.startExchange(remoteID, true, ExchangeID{})
-}
-
-// SignalOnOffer processes a KEM offer the host extracted from an incoming offer and
-// returns the KEM answer for the host to embed in its outgoing answer.
-func (m *Manager) SignalOnOffer(remoteID RemoteID, offer []byte) ([]byte, error) {
-	typ, msg, err := Decode(offer)
-	if err != nil {
-		return nil, fmt.Errorf("decode signal offer from %s: %w", remoteID, err)
-	}
-	if typ != MsgOffer {
-		return nil, fmt.Errorf("expected offer from %s, got type %d", remoteID, typ)
-	}
-	return m.processOffer(remoteID, msg.(*OfferMsg))
-}
-
-// SignalOnAnswer processes a KEM answer the host extracted from an incoming answer.
-// There is no reply: the next offer (over the data path) acknowledges this exchange.
-func (m *Manager) SignalOnAnswer(remoteID RemoteID, answer []byte) error {
-	typ, msg, err := Decode(answer)
-	if err != nil {
-		return fmt.Errorf("decode signal answer from %s: %w", remoteID, err)
-	}
-	if typ != MsgAnswer {
-		return fmt.Errorf("expected answer from %s, got type %d", remoteID, typ)
-	}
-	return m.processAnswer(remoteID, msg.(*AnswerMsg))
-}
-
-// ---- Data path ----
-
-// onDataPathInbound is the transport's inbound handler: it reverse-resolves the
-// source endpoint to a peer and dispatches. Unknown sources are dropped.
-func (m *Manager) onDataPathInbound(src netip.AddrPort, msg []byte) {
-	m.mu.Lock()
-	remoteID, ok := m.peersByAddr[src]
-	m.mu.Unlock()
-	if !ok {
-		return
-	}
-	if err := m.OnDataPathMessage(remoteID, msg); err != nil {
-		m.logger.Debug("pqkem: inbound", "peer", remoteID, "err", err)
-	}
-}
-
-// OnDataPathMessage handles a KEM message received over the data path from remoteID
-// and pushes any reply back over the data path.
-func (m *Manager) OnDataPathMessage(remoteID RemoteID, raw []byte) error {
-	typ, msg, err := Decode(raw)
-	if err != nil {
-		return fmt.Errorf("decode data-path msg from %s: %w", remoteID, err)
-	}
-	switch typ {
-	case MsgOffer:
-		answer, err := m.processOffer(remoteID, msg.(*OfferMsg))
-		if err != nil {
-			return err
-		}
-		if answer == nil {
-			return nil
-		}
-		return m.pushDataPath(remoteID, answer)
-	case MsgAnswer:
-		return m.processAnswer(remoteID, msg.(*AnswerMsg))
-	default:
-		return fmt.Errorf("unhandled data-path message type %d from %s", typ, remoteID)
-	}
-}
-
-// OnDataPathRekeyed notifies that the peer's data path is up and freshly keyed with
-// the latest PSK (fired on first establishment AND every rekey). If we are the
-// initiator that just derived a PSK, it chains the next exchange: a fresh offer over
-// the data path that acknowledges the just-completed one (its arrival under the new
-// key proves to the responder that the key works).
-func (m *Manager) OnDataPathRekeyed(remoteID RemoteID) {
-	m.mu.Lock()
-	ex := m.exchanges[remoteID]
-	chain := ex != nil && ex.state == stateAwaitingRekey
-	var ackID ExchangeID
-	if chain {
-		ackID = ex.id
-	}
-	m.mu.Unlock()
-
-	if !chain {
-		return
-	}
-	offer, err := m.startExchange(remoteID, false, ackID)
-	if err != nil {
-		m.logger.Error("pqkem: chain offer failed to start", "peer", remoteID, "err", err)
-		return
-	}
-	if err := m.pushDataPath(remoteID, offer); err != nil {
-		m.logger.Warn("pqkem: send chain offer failed", "peer", remoteID, "err", err)
-	}
-}
-
-// OnDataPathDown notifies that the peer's data path went down. Rotations resume once
-// the host re-bootstraps over signalling on reconnect; in-flight data-path sends will
-// simply fail until then. Reserved as an explicit hook.
-func (m *Manager) OnDataPathDown(remoteID RemoteID) {}
-
-// ---- internals ----
-
-// pushDataPath resolves the peer's endpoint and sends over the data-path transport,
-// erroring if the peer is unknown or no transport is set.
-func (m *Manager) pushDataPath(remoteID RemoteID, msg []byte) error {
-	m.mu.Lock()
-	ep, ok := m.peerAddrs[remoteID]
-	t := m.transport
-	m.mu.Unlock()
-	if !ok {
-		return fmt.Errorf("no data-path endpoint for peer %s", remoteID)
-	}
-	if t == nil {
-		return fmt.Errorf("no data-path transport")
-	}
-	return t.Send(ep, msg)
-}
-
-func (m *Manager) binding(remoteID RemoteID) Binding {
-	return Binding{LocalID: []byte(m.localID), RemoteID: []byte(remoteID)}
-}
-
-func newExchangeID() (ExchangeID, error) {
-	var id ExchangeID
-	if _, err := rand.Read(id[:]); err != nil {
-		return ExchangeID{}, fmt.Errorf("generate exchange id: %w", err)
-	}
-	return id, nil
-}

client/internal/pqkem/manager_test.go

@@ -1,169 +0,0 @@
-package pqkem
-
-import (
-	"fmt"
-	"net/netip"
-	"sync"
-	"sync/atomic"
-	"testing"
-
-	"github.com/stretchr/testify/require"
-)
-
-// netSwitch is an in-memory UDP fabric: transports register their endpoint and get
-// datagrams delivered to their inbound handler.
-type netSwitch struct {
-	mu sync.Mutex
-	h  map[netip.AddrPort]func(netip.AddrPort, []byte)
-}
-
-func newSwitch() *netSwitch {
-	return &netSwitch{h: map[netip.AddrPort]func(netip.AddrPort, []byte){}}
-}
-
-func (s *netSwitch) register(ep netip.AddrPort, fn func(netip.AddrPort, []byte)) {
-	s.mu.Lock()
-	s.h[ep] = fn
-	s.mu.Unlock()
-}
-
-func (s *netSwitch) deliver(dst, src netip.AddrPort, msg []byte) error {
-	s.mu.Lock()
-	fn := s.h[dst]
-	s.mu.Unlock()
-	if fn == nil {
-		return fmt.Errorf("no route to %s", dst)
-	}
-	fn(src, msg)
-	return nil
-}
-
-// loopback is an endpoint-based pqkem.Transport over a netSwitch, with a switchable
-// drop flag.
-type loopback struct {
-	ep   netip.AddrPort
-	sw   *netSwitch
-	drop atomic.Bool
-}
-
-func (l *loopback) Send(dst netip.AddrPort, msg []byte) error {
-	if l.drop.Load() {
-		return nil
-	}
-	return l.sw.deliver(dst, l.ep, append([]byte(nil), msg...))
-}
-
-func (l *loopback) LocalPort() int                             { return int(l.ep.Port()) }
-func (l *loopback) Run(onInbound func(netip.AddrPort, []byte)) { l.sw.register(l.ep, onInbound) }
-func (l *loopback) Close() error                               { return nil }
-
-type fakeWG struct {
-	mu     sync.Mutex
-	psks   map[RemoteID]PSK
-	failed []RemoteID
-}
-
-func newFakeWG() *fakeWG { return &fakeWG{psks: map[RemoteID]PSK{}} }
-
-func (f *fakeWG) OnNewPSKReady(remoteID RemoteID, psk PSK) error {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	f.psks[remoteID] = psk
-	return nil
-}
-
-func (f *fakeWG) OnRekeyFailed(remoteID RemoteID) error {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	f.failed = append(f.failed, remoteID)
-	return nil
-}
-
-func (f *fakeWG) psk(peer RemoteID) PSK {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	return f.psks[peer]
-}
-
-var (
-	epA = netip.MustParseAddrPort("100.64.0.1:51833")
-	epB = netip.MustParseAddrPort("100.64.0.2:51833")
-)
-
-// pair builds two wired managers (B is the initiator, "bbbb" > "aaaa") sharing a
-// netSwitch, with each peer's data-path endpoint registered. lbB is B's loopback
-// (for toggling drop).
-func pair(t *testing.T) (dA, dB *Manager, wgA, wgB *fakeWG, lbB *loopback) {
-	t.Helper()
-	sw := newSwitch()
-	wgA = newFakeWG()
-	wgB = newFakeWG()
-	dA = NewManager("aaaa", wgA, nil)
-	dB = NewManager("bbbb", wgB, nil)
-	dA.Start(&loopback{ep: epA, sw: sw})
-	lbB = &loopback{ep: epB, sw: sw}
-	dB.Start(lbB)
-	dA.AddPeer("bbbb", epB)
-	dB.AddPeer("aaaa", epA)
-	return dA, dB, wgA, wgB, lbB
-}
-
-// bootstrap runs the signalling offer/answer (the test plays the host carrying bytes).
-func bootstrap(t *testing.T, dA, dB *Manager) {
-	t.Helper()
-	offer, err := dB.SignalOffer("aaaa")
-	require.NoError(t, err)
-	require.NotNil(t, offer)
-	answer, err := dA.SignalOnOffer("bbbb", offer)
-	require.NoError(t, err)
-	require.NotNil(t, answer)
-	require.NoError(t, dB.SignalOnAnswer("aaaa", answer))
-}
-
-func TestManager_BootstrapDerivesSamePSK(t *testing.T) {
-	dA, dB, wgA, wgB, _ := pair(t)
-	defer dA.Stop()
-	defer dB.Stop()
-
-	bootstrap(t, dA, dB)
-
-	pskA := wgA.psk("bbbb")
-	pskB := wgB.psk("aaaa")
-	require.NotEqual(t, PSK{}, pskA)
-	require.Equal(t, pskB, pskA, "both sides derive the same PSK from the bootstrap exchange")
-}
-
-func TestManager_ChainRotatesAndAcks(t *testing.T) {
-	dA, dB, wgA, wgB, _ := pair(t)
-	defer dA.Stop()
-	defer dB.Stop()
-
-	bootstrap(t, dA, dB)
-	psk1 := wgB.psk("aaaa")
-
-	// Data path up: B (initiator) chains the next offer over the data path, which
-	// rotates both to a fresh PSK and acknowledges A.
-	dA.OnDataPathRekeyed("bbbb")
-	dB.OnDataPathRekeyed("aaaa")
-
-	psk2A := wgA.psk("bbbb")
-	psk2B := wgB.psk("aaaa")
-	require.Equal(t, psk2B, psk2A, "both sides converge on the rotated PSK")
-	require.NotEqual(t, psk1, psk2B, "the chain rotated to a new PSK")
-}
-
-func TestManager_NonInitiatorReturnsNoOffer(t *testing.T) {
-	dA := NewManager("aaaa", newFakeWG(), nil)
-	defer dA.Stop()
-
-	offer, err := dA.SignalOffer("bbbb") // not the initiator vs "bbbb"
-	require.NoError(t, err)
-	require.Nil(t, offer)
-}
-
-func TestManager_StopIsIdempotent(t *testing.T) {
-	dA := NewManager("aaaa", newFakeWG(), nil)
-	dA.Start(&loopback{ep: epA, sw: newSwitch()})
-	dA.Stop()
-	dA.Stop() // must not panic or hang
-}

client/internal/pqkem/message.go

@@ -1,121 +0,0 @@
-package pqkem
-
-import (
-	"crypto/mlkem"
-	"fmt"
-)
-
-// Wire framing for the PQ-KEM exchange. Messages are self-contained, versioned,
-// transport-agnostic byte blobs: the same bytes ride the signalling channel
-// (initial bootstrap) or a data-tunnel packet (rekey). The library only ever sees
-// opaque []byte at the transport seam.
-//
-// Layout (all messages): [type:1][version:1][exchangeID:16][payload...]
-//
-// There is no confirm message: an exchange is acknowledged by the NEXT offer, which
-// carries the acked exchange's id (see OfferMsg.AckID) and — riding the data path
-// under the freshly adopted key — proves that key works.
-
-const (
-	// ProtocolVersion is bumped on any wire-incompatible change; a peer rejects
-	// messages it does not understand rather than misparsing them.
-	ProtocolVersion uint8 = 1
-
-	// ExchangeIDSize identifies one exchange so answers/acks correlate and stale
-	// messages are dropped.
-	ExchangeIDSize = 16
-
-	headerSize = 1 + 1 + ExchangeIDSize
-)
-
-// MsgType tags the two message kinds of the exchange.
-type MsgType uint8
-
-const (
-	MsgOffer MsgType = iota + 1
-	MsgAnswer
-)
-
-// ExchangeID is the per-exchange correlator. The zero value means "none" (an offer
-// that acknowledges nothing, i.e. the first exchange of a connection).
-type ExchangeID [ExchangeIDSize]byte
-
-// OfferMsg carries the initiator's public material (X25519 pub ‖ ML-KEM encap key)
-// and AckID, the id of the previous exchange this offer acknowledges (zero if none).
-type OfferMsg struct {
-	ExchangeID ExchangeID
-	AckID      ExchangeID
-	// KEMOffer is the raw Initiator.Offer() blob (OfferSize bytes).
-	KEMOffer []byte
-}
-
-// AnswerMsg carries the responder's reply (ML-KEM ciphertext ‖ X25519 pub) for the
-// round identified by ExchangeID.
-type AnswerMsg struct {
-	ExchangeID ExchangeID
-	// KEMAnswer is the raw Respond() answer blob (AnswerSize bytes).
-	KEMAnswer []byte
-}
-
-// Encode serialises the offer with its framed header (payload = AckID ‖ KEMOffer).
-func (m *OfferMsg) Encode() ([]byte, error) {
-	if len(m.KEMOffer) != OfferSize {
-		return nil, fmt.Errorf("offer payload: got %d, want %d", len(m.KEMOffer), OfferSize)
-	}
-	payload := make([]byte, 0, ExchangeIDSize+OfferSize)
-	payload = append(payload, m.AckID[:]...)
-	payload = append(payload, m.KEMOffer...)
-	return frame(MsgOffer, m.ExchangeID, payload), nil
-}
-
-// Encode serialises the answer with its framed header.
-func (m *AnswerMsg) Encode() ([]byte, error) {
-	if len(m.KEMAnswer) != AnswerSize {
-		return nil, fmt.Errorf("answer payload: got %d, want %d", len(m.KEMAnswer), AnswerSize)
-	}
-	return frame(MsgAnswer, m.ExchangeID, m.KEMAnswer), nil
-}
-
-// Decode parses a framed message into one of *OfferMsg / *AnswerMsg.
-func Decode(buf []byte) (MsgType, any, error) {
-	if len(buf) < headerSize {
-		return 0, nil, fmt.Errorf("message too short: %d bytes", len(buf))
-	}
-	typ := MsgType(buf[0])
-	if ver := buf[1]; ver != ProtocolVersion {
-		return typ, nil, fmt.Errorf("unsupported protocol version %d (want %d)", ver, ProtocolVersion)
-	}
-
-	var id ExchangeID
-	copy(id[:], buf[2:headerSize])
-	payload := buf[headerSize:]
-
-	switch typ {
-	case MsgOffer:
-		if len(payload) != ExchangeIDSize+OfferSize {
-			return typ, nil, fmt.Errorf("offer payload: got %d, want %d", len(payload), ExchangeIDSize+OfferSize)
-		}
-		var ack ExchangeID
-		copy(ack[:], payload[:ExchangeIDSize])
-		return typ, &OfferMsg{ExchangeID: id, AckID: ack, KEMOffer: payload[ExchangeIDSize:]}, nil
-	case MsgAnswer:
-		if len(payload) != AnswerSize {
-			return typ, nil, fmt.Errorf("answer payload: got %d, want %d", len(payload), AnswerSize)
-		}
-		return typ, &AnswerMsg{ExchangeID: id, KEMAnswer: payload}, nil
-	default:
-		return typ, nil, fmt.Errorf("unknown message type %d", typ)
-	}
-}
-
-func frame(typ MsgType, id ExchangeID, payload []byte) []byte {
-	buf := make([]byte, headerSize+len(payload))
-	buf[0] = byte(typ)
-	buf[1] = ProtocolVersion
-	copy(buf[2:], id[:])
-	copy(buf[headerSize:], payload)
-	return buf
-}
-
-// compile-time assurance the KEM blob sizes referenced here stay in sync with kem.go.
-var _ = [1]struct{}{}[OfferSize-(32+mlkem.EncapsulationKeySize768)]

client/internal/pqkem/message_test.go

@@ -1,57 +0,0 @@
-package pqkem
-
-import (
-	"testing"
-
-	"github.com/stretchr/testify/require"
-)
-
-func TestMessageRoundTrip(t *testing.T) {
-	init, err := NewInitiator()
-	require.NoError(t, err)
-	answer, _, err := Respond(init.Offer(), Binding{LocalID: wgB, RemoteID: wgA})
-	require.NoError(t, err)
-
-	id := ExchangeID{1, 2, 3, 4}
-	ack := ExchangeID{9, 9, 9}
-
-	offBytes, err := (&OfferMsg{ExchangeID: id, AckID: ack, KEMOffer: init.Offer()}).Encode()
-	require.NoError(t, err)
-	typ, decoded, err := Decode(offBytes)
-	require.NoError(t, err)
-	require.Equal(t, MsgOffer, typ)
-	require.Equal(t, id, decoded.(*OfferMsg).ExchangeID)
-	require.Equal(t, ack, decoded.(*OfferMsg).AckID)
-	require.Equal(t, init.Offer(), decoded.(*OfferMsg).KEMOffer)
-
-	ansBytes, err := (&AnswerMsg{ExchangeID: id, KEMAnswer: answer}).Encode()
-	require.NoError(t, err)
-	typ, decoded, err = Decode(ansBytes)
-	require.NoError(t, err)
-	require.Equal(t, MsgAnswer, typ)
-	require.Equal(t, answer, decoded.(*AnswerMsg).KEMAnswer)
-}
-
-func TestDecodeRejects(t *testing.T) {
-	// too short
-	_, _, err := Decode([]byte{1, 1})
-	require.Error(t, err)
-
-	// wrong version
-	bad := make([]byte, headerSize+ExchangeIDSize+OfferSize)
-	bad[0] = byte(MsgOffer)
-	bad[1] = ProtocolVersion + 1
-	_, _, err = Decode(bad)
-	require.Error(t, err)
-
-	// unknown type
-	bad2 := make([]byte, headerSize)
-	bad2[0] = 99
-	bad2[1] = ProtocolVersion
-	_, _, err = Decode(bad2)
-	require.Error(t, err)
-
-	// offer with wrong payload size
-	_, err = (&OfferMsg{KEMOffer: []byte{1, 2, 3}}).Encode()
-	require.Error(t, err)
-}

client/internal/pqkem_adapter.go

@@ -1,102 +0,0 @@
-package internal
-
-import (
-	"net/netip"
-
-	log "github.com/sirupsen/logrus"
-	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
-
-	"github.com/netbirdio/netbird/client/internal/pqkem"
-)
-
-// pqPresharedKeySetter is the subset of the WireGuard interface the ML-KEM callback
-// needs: programming a peer's preshared key. *iface.WGIface satisfies it.
-type pqPresharedKeySetter interface {
-	SetPresharedKey(peerKey string, psk wgtypes.Key, updateOnly bool) error
-}
-
-// pqCallbackHandler programs the derived PQ PSK onto the WireGuard peer. It is the
-// engine-side implementation of pqkem.CallbackHandler.
-type pqCallbackHandler struct {
-	wg pqPresharedKeySetter
-}
-
-// OnNewPSKReady programs the freshly derived PSK for the peer (updateOnly: a no-op
-// if the peer is not present, mirroring Rosenpass). remoteID is the peer's WG pubkey.
-func (h pqCallbackHandler) OnNewPSKReady(remoteID pqkem.RemoteID, psk pqkem.PSK) error {
-	// updateOnly: applies to an already-configured peer (rotation). At bootstrap the
-	// peer is not configured yet, so this is a no-op there and the PSK is instead
-	// pulled at peer-config time (pqHandshaker.PSK / conn.presharedKey).
-	log.Debugf("pqkem: programming PSK for peer %s", remoteID)
-	return h.wg.SetPresharedKey(string(remoteID), wgtypes.Key(psk), true)
-}
-
-// OnRekeyFailed reports a failed PQ (re)key convergence.
-// TODO(NET-1406): tear the peer connection down / trigger ICE reconnect.
-func (h pqCallbackHandler) OnRekeyFailed(remoteID pqkem.RemoteID) error {
-	log.Warnf("pqkem: post-quantum rekey failed for peer %s", remoteID)
-	return nil
-}
-
-// pqHandshaker adapts the pqkem manager to peer.PQHandshaker (string peer keys),
-// wiring the host's signalling offers/answers to the KEM exchange.
-type pqHandshaker struct {
-	mgr *pqkem.Manager
-}
-
-func (p pqHandshaker) OfferPayload(remoteKey string) ([]byte, int) {
-	payload, err := p.mgr.SignalOffer(pqkem.RemoteID(remoteKey))
-	if err != nil {
-		log.Warnf("pqkem: build offer for %s: %v", remoteKey, err)
-	}
-	return payload, p.mgr.LocalPort()
-}
-
-func (p pqHandshaker) AnswerPayload(remoteKey string, recvOffer []byte) ([]byte, int) {
-	if len(recvOffer) == 0 {
-		return nil, p.mgr.LocalPort()
-	}
-	payload, err := p.mgr.SignalOnOffer(pqkem.RemoteID(remoteKey), recvOffer)
-	if err != nil {
-		log.Warnf("pqkem: build answer for %s: %v", remoteKey, err)
-	}
-	return payload, p.mgr.LocalPort()
-}
-
-func (p pqHandshaker) OnAnswer(remoteKey string, recvAnswer []byte) {
-	if len(recvAnswer) == 0 {
-		return
-	}
-	if err := p.mgr.SignalOnAnswer(pqkem.RemoteID(remoteKey), recvAnswer); err != nil {
-		log.Warnf("pqkem: process answer from %s: %v", remoteKey, err)
-	}
-}
-
-// PSK exposes the peer's derived PSK for the conn to program at WG peer-config time.
-func (p pqHandshaker) PSK(remoteKey string) (wgtypes.Key, bool) {
-	psk, ok := p.mgr.PSK(pqkem.RemoteID(remoteKey))
-	if !ok {
-		return wgtypes.Key{}, false
-	}
-	return wgtypes.Key(psk), true
-}
-
-// SetRemoteAddr registers the peer's data-path endpoint (overlay IP + pq UDP port)
-// learned from signalling. Sends only ever fire once the tunnel is up (clocked by
-// OnDataPathRekeyed), so registering here is safe even before connection-up.
-func (p pqHandshaker) SetRemoteAddr(remoteKey string, addr netip.AddrPort) {
-	if !addr.IsValid() || addr.Port() == 0 {
-		return
-	}
-	p.mgr.AddPeer(pqkem.RemoteID(remoteKey), addr)
-}
-
-// OnDataPathRekeyed clocks the next chained PSK rotation on a fresh WG handshake.
-func (p pqHandshaker) OnDataPathRekeyed(remoteKey string) {
-	p.mgr.OnDataPathRekeyed(pqkem.RemoteID(remoteKey))
-}
-
-// OnDataPathDown signals the peer's tunnel went down.
-func (p pqHandshaker) OnDataPathDown(remoteKey string) {
-	p.mgr.OnDataPathDown(pqkem.RemoteID(remoteKey))
-}

client/internal/pqkem_transport.go

@@ -1,72 +0,0 @@
-package internal
-
-import (
-	"fmt"
-	"net"
-	"net/netip"
-
-	log "github.com/sirupsen/logrus"
-)
-
-// DefaultPort is the preferred UDP port for the ML-KEM data-path service, bound on
-// the WG overlay IP. Since each client owns a distinct overlay IP, this port is
-// almost always free, so it need not be announced (peers assume it). A peer only
-// announces Body.mlkemPort when a collision forced it onto a different port.
-const DefaultPort = 51833
-
-// pqTransport is the ML-KEM data-path transport: a dumb UDP socket bound on the WG
-// overlay IP. It implements pqkem.Transport — the manager owns the remoteID<->endpoint
-// routing and drives this socket's lifecycle (Run / Close).
-type pqTransport struct {
-	conn *net.UDPConn
-	port int
-}
-
-// newPQTransport binds a UDP socket on the WG overlay IP, preferring DefaultPort and
-// falling back to an OS-assigned ephemeral port if it is in use. Call it after the WG
-// interface is up so the overlay IP is assigned; when the bound port is not
-// DefaultPort it must be announced to peers via Body.mlkemPort.
-func newPQTransport(overlayIP netip.Addr) (*pqTransport, error) {
-	if !overlayIP.IsValid() {
-		return nil, fmt.Errorf("invalid overlay IP for pqkem transport")
-	}
-	ip := net.IP(overlayIP.AsSlice())
-	conn, err := net.ListenUDP("udp4", &net.UDPAddr{IP: ip, Port: DefaultPort})
-	if err != nil {
-		log.Debugf("pqkem: default port %d unavailable on %s (%v), using an ephemeral port", DefaultPort, overlayIP, err)
-		conn, err = net.ListenUDP("udp4", &net.UDPAddr{IP: ip, Port: 0})
-		if err != nil {
-			return nil, fmt.Errorf("bind pqkem udp on overlay %s: %w", overlayIP, err)
-		}
-	}
-	return &pqTransport{conn: conn, port: conn.LocalAddr().(*net.UDPAddr).Port}, nil
-}
-
-// Send implements pqkem.Transport.
-func (t *pqTransport) Send(endpoint netip.AddrPort, msg []byte) error {
-	_, err := t.conn.WriteToUDPAddrPort(msg, endpoint)
-	return err
-}
-
-// LocalPort implements pqkem.Transport.
-func (t *pqTransport) LocalPort() int { return t.port }
-
-// Run implements pqkem.Transport: the receive loop, delivering each datagram as
-// (source endpoint, msg). Exits when the socket is closed.
-func (t *pqTransport) Run(onInbound func(src netip.AddrPort, msg []byte)) {
-	go func() {
-		buf := make([]byte, 2048)
-		for {
-			n, src, err := t.conn.ReadFromUDPAddrPort(buf)
-			if err != nil {
-				return
-			}
-			msg := make([]byte, n)
-			copy(msg, buf[:n])
-			onInbound(src, msg)
-		}
-	}()
-}
-
-// Close implements pqkem.Transport.
-func (t *pqTransport) Close() error { return t.conn.Close() }

client/internal/profilemanager/config.go

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

client/internal/profilemanager/service.go

@@ -11,7 +11,6 @@ import (
 	"runtime"
 	"sort"
 	"strings"
-	"syscall"
 
 	log "github.com/sirupsen/logrus"
 
@@ -440,11 +439,7 @@ func (s *ServiceManager) GetStatePath() string {
 
 	activeProf, err := s.GetActiveProfileState()
 	if err != nil {
-		if errors.Is(err, syscall.ENOSYS) {
-			log.Debugf("active profile state unavailable on this platform: %v", err)
-		} else {
-			log.Warnf("failed to get active profile state: %v", err)
-		}
+		log.Warnf("failed to get active profile state: %v", err)
 		return defaultStatePath
 	}
 

client/internal/rosenpass/manager.go

@@ -39,7 +39,6 @@ type rpServer interface {
 
 type Manager struct {
 	ifaceName    string
-	localWgKey   wgtypes.Key
 	spk          []byte
 	ssk          []byte
 	rpKeyHash    string
@@ -52,9 +51,8 @@ type Manager struct {
 	wgIface      PresharedKeySetter
 }
 
-// NewManager creates a new Rosenpass manager. localWgKey is the local
-// WireGuard public key, used to derive the per-peer rendezvous key.
-func NewManager(preSharedKey *wgtypes.Key, wgIfaceName string, localWgKey wgtypes.Key) (*Manager, error) {
+// NewManager creates a new Rosenpass manager
+func NewManager(preSharedKey *wgtypes.Key, wgIfaceName string) (*Manager, error) {
 	public, secret, err := rp.GenerateKeyPair()
 	if err != nil {
 		return nil, err
@@ -64,7 +62,6 @@ func NewManager(preSharedKey *wgtypes.Key, wgIfaceName string, localWgKey wgtype
 	log.Tracef("generated new rosenpass key pair with public key %s", rpKeyHash)
 	return &Manager{
 		ifaceName:    wgIfaceName,
-		localWgKey:   localWgKey,
 		rpKeyHash:    rpKeyHash,
 		spk:          public,
 		ssk:          secret,
@@ -76,7 +73,7 @@ func NewManager(preSharedKey *wgtypes.Key, wgIfaceName string, localWgKey wgtype
 		// nil receiver in addPeer -> m.rpWgHandler.AddPeer. generateConfig will
 		// replace it with a fresh handler on each Run() to clear stale peer
 		// state from previous engine sessions.
-		rpWgHandler: NewNetbirdHandler((*[32]byte)(preSharedKey), localWgKey),
+		rpWgHandler: NewNetbirdHandler(),
 		lock:        sync.Mutex{},
 	}, nil
 }
@@ -164,7 +161,7 @@ func (m *Manager) generateConfig() (rp.Config, error) {
 	cfg.Peers = []rp.PeerConfig{}
 
 	m.lock.Lock()
-	m.rpWgHandler = NewNetbirdHandler(m.preSharedKey, m.localWgKey)
+	m.rpWgHandler = NewNetbirdHandler()
 	if m.wgIface != nil {
 		m.rpWgHandler.SetInterface(m.wgIface)
 	}

client/internal/rosenpass/manager_test.go

@@ -85,7 +85,7 @@ func newTestManager(spkFirstByte byte, mock *mockServer) *Manager {
 		ssk:         make([]byte, 32),
 		rpKeyHash:   "test-hash",
 		rpPeerIDs:   make(map[string]*rp.PeerID),
-		rpWgHandler: NewNetbirdHandler(nil, wgtypes.Key{0x01}),
+		rpWgHandler: NewNetbirdHandler(),
 		server:      mock,
 	}
 }
@@ -255,7 +255,7 @@ func TestAddPeer_NilServer_ReturnsErrorNoCrash(t *testing.T) {
 // issue #4341 cannot occur in the window between NewManager and Run().
 func TestNewManager_PreInitializesHandler(t *testing.T) {
 	psk := wgtypes.Key{}
-	m, err := NewManager(&psk, "wt0", wgtypes.Key{0x01})
+	m, err := NewManager(&psk, "wt0")
 	require.NoError(t, err)
 	require.NotNil(t, m.rpWgHandler, "rpWgHandler must be initialized in NewManager")
 }
@@ -329,10 +329,10 @@ func TestIsPresharedKeyInitialized_AddedButNotHandshaken_ReturnsFalse(t *testing
 	require.False(t, m.IsPresharedKeyInitialized(wgKey))
 }
 
-// --- NetbirdHandler.applyKey ----------------------------------------------
+// --- NetbirdHandler.outputKey ----------------------------------------------
 
-func TestHandler_ApplyKey_FirstCallUsesUpdateOnlyFalse(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+func TestHandler_OutputKey_FirstCallUsesUpdateOnlyFalse(t *testing.T) {
+	h := NewNetbirdHandler()
 	iface := &mockIface{}
 	h.SetInterface(iface)
 
@@ -348,8 +348,8 @@ func TestHandler_ApplyKey_FirstCallUsesUpdateOnlyFalse(t *testing.T) {
 	require.Equal(t, wgKey.String(), iface.calls[0].peerKey)
 }
 
-func TestHandler_ApplyKey_SubsequentCallsUseUpdateOnlyTrue(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+func TestHandler_OutputKey_SubsequentCallsUseUpdateOnlyTrue(t *testing.T) {
+	h := NewNetbirdHandler()
 	iface := &mockIface{}
 	h.SetInterface(iface)
 
@@ -364,8 +364,8 @@ func TestHandler_ApplyKey_SubsequentCallsUseUpdateOnlyTrue(t *testing.T) {
 	require.True(t, iface.calls[1].updateOnly, "subsequent rotations must use updateOnly=true")
 }
 
-func TestHandler_ApplyKey_NilInterface_NoCrashNoCall(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+func TestHandler_OutputKey_NilInterface_NoCrashNoCall(t *testing.T) {
+	h := NewNetbirdHandler()
 	// no SetInterface — iface remains nil
 	pid := rp.PeerID{0x03}
 	h.AddPeer(pid, "wt0", rp.Key(wgtypes.Key{}))
@@ -374,8 +374,8 @@ func TestHandler_ApplyKey_NilInterface_NoCrashNoCall(t *testing.T) {
 	h.HandshakeCompleted(pid, rp.Key{})
 }
 
-func TestHandler_ApplyKey_UnknownPeer_NoCall(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+func TestHandler_OutputKey_UnknownPeer_NoCall(t *testing.T) {
+	h := NewNetbirdHandler()
 	iface := &mockIface{}
 	h.SetInterface(iface)
 
@@ -384,7 +384,7 @@ func TestHandler_ApplyKey_UnknownPeer_NoCall(t *testing.T) {
 }
 
 func TestHandler_RemovePeer_ClearsInitializedState(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+	h := NewNetbirdHandler()
 	iface := &mockIface{}
 	h.SetInterface(iface)
 
@@ -398,7 +398,7 @@ func TestHandler_RemovePeer_ClearsInitializedState(t *testing.T) {
 }
 
 func TestHandler_SetInterfaceAfterAddPeer_StillReceivesKey(t *testing.T) {
-	h := NewNetbirdHandler(nil, wgtypes.Key{0x01})
+	h := NewNetbirdHandler()
 	pid := rp.PeerID{0x05}
 	wgKey := wgtypes.Key{0xEE}
 	h.AddPeer(pid, "wt0", rp.Key(wgKey))

client/internal/rosenpass/netbird_handler.go

@@ -18,34 +18,19 @@ type PresharedKeySetter interface {
 type wireGuardPeer struct {
 	Interface string
 	PublicKey rp.Key
-	// initialized is true once a completed exchange has set a
-	// Rosenpass-managed PSK for this peer.
-	initialized bool
-	// chainKey is the key output by the last completed exchange, advanced by
-	// one ratchet step on expiry. Nil until the first exchange completes and
-	// after the peer has fallen back to the rendezvous key.
-	chainKey *wgtypes.Key
-	// expiries counts failed renewals since the last completed exchange.
-	expiries int
 }
 
 type NetbirdHandler struct {
-	mu    sync.Mutex
-	iface PresharedKeySetter
-	// preSharedKey is the account-level preshared key, used as the rendezvous
-	// key when set. Nil means the deterministic seed key is used instead.
-	preSharedKey *[32]byte
-	// localWgKey is the local WireGuard public key, one of the two inputs to
-	// the deterministic seed key.
-	localWgKey wgtypes.Key
-	peers      map[rp.PeerID]*wireGuardPeer
+	mu               sync.Mutex
+	iface            PresharedKeySetter
+	peers            map[rp.PeerID]wireGuardPeer
+	initializedPeers map[rp.PeerID]bool
 }
 
-func NewNetbirdHandler(preSharedKey *[32]byte, localWgKey wgtypes.Key) *NetbirdHandler {
+func NewNetbirdHandler() *NetbirdHandler {
 	return &NetbirdHandler{
-		preSharedKey: preSharedKey,
-		localWgKey:   localWgKey,
-		peers:        map[rp.PeerID]*wireGuardPeer{},
+		peers:            map[rp.PeerID]wireGuardPeer{},
+		initializedPeers: map[rp.PeerID]bool{},
 	}
 }
 
@@ -57,16 +42,10 @@ func (h *NetbirdHandler) SetInterface(iface PresharedKeySetter) {
 	h.iface = iface
 }
 
-// AddPeer registers a peer with the handler. Re-adding a known peer (every
-// reconnection does) keeps its key recovery state.
 func (h *NetbirdHandler) AddPeer(pid rp.PeerID, intf string, pk rp.Key) {
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	if existing, ok := h.peers[pid]; ok && existing.PublicKey == pk {
-		existing.Interface = intf
-		return
-	}
-	h.peers[pid] = &wireGuardPeer{
+	h.peers[pid] = wireGuardPeer{
 		Interface: intf,
 		PublicKey: pk,
 	}
@@ -76,6 +55,7 @@ func (h *NetbirdHandler) RemovePeer(pid rp.PeerID) {
 	h.mu.Lock()
 	defer h.mu.Unlock()
 	delete(h.peers, pid)
+	delete(h.initializedPeers, pid)
 }
 
 // IsPeerInitialized returns true if Rosenpass has completed a handshake
@@ -83,120 +63,50 @@ func (h *NetbirdHandler) RemovePeer(pid rp.PeerID) {
 func (h *NetbirdHandler) IsPeerInitialized(pid rp.PeerID) bool {
 	h.mu.Lock()
 	defer h.mu.Unlock()
-	peer, ok := h.peers[pid]
-	return ok && peer.initialized
+	return h.initializedPeers[pid]
 }
 
-// HandshakeCompleted programs the freshly exchanged output key and resets the
-// peer's key recovery state.
 func (h *NetbirdHandler) HandshakeCompleted(pid rp.PeerID, key rp.Key) {
-	psk := wgtypes.Key(key)
-
-	h.mu.Lock()
-	defer h.mu.Unlock()
-
-	peer, ok := h.peers[pid]
-	if !ok {
-		return
-	}
-	if peer.expiries > 0 {
-		log.Infof("rosenpass exchange completed for peer %s after %d expired renewals", wgtypes.Key(peer.PublicKey), peer.expiries)
-	}
-	// chainKey tracks the shared exchange output regardless of the local write
-	// outcome, so both ends still converge on the next expiry.
-	peer.chainKey = &psk
-	peer.expiries = 0
-	if !h.applyKeyLocked(pid, psk, peer.initialized) {
-		return
-	}
-	peer.initialized = true
+	h.outputKey(rp.KeyOutputReasonStale, pid, key)
 }
 
-// HandshakeExpired replaces the expired key. The renewal exchange runs over
-// the tunnel keyed by the PSK itself, so the replacement must be derivable on
-// both ends without communication: the first expiry ratchets the last shared
-// key forward, repeated expiries (and expiries without a completed exchange)
-// fall back to the rendezvous key and drop the peer out of the initialized
-// state so connection reconfigurations reprogram the rendezvous key as well.
 func (h *NetbirdHandler) HandshakeExpired(pid rp.PeerID) {
+	key, _ := rp.GeneratePresharedKey()
+	h.outputKey(rp.KeyOutputReasonStale, pid, key)
+}
+
+func (h *NetbirdHandler) outputKey(_ rp.KeyOutputReason, pid rp.PeerID, psk rp.Key) {
 	h.mu.Lock()
-	defer h.mu.Unlock()
+	iface := h.iface
+	wg, ok := h.peers[pid]
+	isInitialized := h.initializedPeers[pid]
+	h.mu.Unlock()
 
-	peer, ok := h.peers[pid]
-	if !ok {
-		return
-	}
-
-	peer.expiries++
-
-	var psk wgtypes.Key
-	if peer.chainKey != nil && peer.expiries == 1 {
-		log.Infof("rosenpass key for peer %s expired without renewal, advancing to ratcheted key", wgtypes.Key(peer.PublicKey))
-		psk = RatchetKey(*peer.chainKey)
-		peer.chainKey = &psk
-	} else {
-		rendezvous, err := h.rendezvousKey(peer)
-		if err != nil {
-			// Fail closed: without a rendezvous key the expired key must
-			// still be rotated out, even if the replacement is unusable.
-			log.Errorf("failed to derive rendezvous key, replacing expired key with a random one: %v", err)
-			h.applyRandomKeyLocked(pid)
-			return
-		}
-		log.Warnf("rosenpass key for peer %s expired %d times without renewal, falling back to the rendezvous key", wgtypes.Key(peer.PublicKey), peer.expiries)
-		psk = rendezvous
-		peer.chainKey = nil
-		peer.initialized = false
-	}
-
-	h.applyKeyLocked(pid, psk, true)
-}
-
-// rendezvousKey returns the key both ends converge on without communication:
-// the account-level preshared key when configured, the deterministic seed key
-// otherwise. It mirrors the key that peer connections program when Rosenpass
-// does not manage the peer yet.
-func (h *NetbirdHandler) rendezvousKey(peer *wireGuardPeer) (wgtypes.Key, error) {
-	if h.preSharedKey != nil {
-		return *h.preSharedKey, nil
-	}
-
-	seed, err := DeterministicSeedKey(h.localWgKey.String(), wgtypes.Key(peer.PublicKey).String())
-	if err != nil {
-		return wgtypes.Key{}, err
-	}
-	return *seed, nil
-}
-
-// applyKeyLocked writes the preshared key for the peer to the WireGuard
-// interface and reports whether the write succeeded. Callers must hold h.mu
-// for the whole state-mutation-plus-write so that a concurrent completion and
-// expiry cannot reorder their writes relative to the in-memory chain key.
-func (h *NetbirdHandler) applyKeyLocked(pid rp.PeerID, psk wgtypes.Key, updateOnly bool) bool {
-	peer, ok := h.peers[pid]
-	if !ok {
-		return false
-	}
-
-	if h.iface == nil {
+	if iface == nil {
 		log.Warn("rosenpass: interface not set, cannot update preshared key")
-		return false
-	}
-
-	peerKey := wgtypes.Key(peer.PublicKey).String()
-	if err := h.iface.SetPresharedKey(peerKey, psk, updateOnly); err != nil {
-		log.Errorf("Failed to apply rosenpass key: %v", err)
-		return false
-	}
-
-	return true
-}
-
-func (h *NetbirdHandler) applyRandomKeyLocked(pid rp.PeerID) {
-	key, err := rp.GeneratePresharedKey()
-	if err != nil {
-		log.Errorf("failed to generate random preshared key: %v", err)
 		return
 	}
-	h.applyKeyLocked(pid, wgtypes.Key(key), true)
+
+	if !ok {
+		return
+	}
+
+	peerKey := wgtypes.Key(wg.PublicKey).String()
+	pskKey := wgtypes.Key(psk)
+
+	// Use updateOnly=true for later rotations (peer already has Rosenpass PSK)
+	// Use updateOnly=false for first rotation (peer has original/empty PSK)
+	if err := iface.SetPresharedKey(peerKey, pskKey, isInitialized); err != nil {
+		log.Errorf("Failed to apply rosenpass key: %v", err)
+		return
+	}
+
+	// Mark peer as isInitialized after the successful first rotation
+	if !isInitialized {
+		h.mu.Lock()
+		if _, exists := h.peers[pid]; exists {
+			h.initializedPeers[pid] = true
+		}
+		h.mu.Unlock()
+	}
 }

client/internal/rosenpass/netbird_handler_test.go

@@ -1,250 +0,0 @@
-package rosenpass
-
-import (
-	"testing"
-
-	rp "cunicu.li/go-rosenpass"
-	"github.com/stretchr/testify/require"
-	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
-)
-
-// handlerTestLink wires two NetbirdHandlers as the two ends of a single
-// tunnel: handler A manages the rosenpass peer B and vice versa, the way two
-// NetBird clients see each other.
-type handlerTestLink struct {
-	handlerA, handlerB *NetbirdHandler
-	ifaceA, ifaceB     *mockIface
-	pidA, pidB         rp.PeerID
-	wgKeyA, wgKeyB     wgtypes.Key
-}
-
-func newHandlerTestLink(t *testing.T, preSharedKey *[32]byte) *handlerTestLink {
-	t.Helper()
-
-	link := &handlerTestLink{
-		ifaceA: &mockIface{},
-		ifaceB: &mockIface{},
-	}
-	link.pidA[0] = 0xaa
-	link.pidB[0] = 0xbb
-	link.wgKeyA[31] = 1
-	link.wgKeyB[31] = 2
-
-	link.handlerA = NewNetbirdHandler(preSharedKey, link.wgKeyA)
-	link.handlerB = NewNetbirdHandler(preSharedKey, link.wgKeyB)
-
-	link.handlerA.SetInterface(link.ifaceA)
-	link.handlerB.SetInterface(link.ifaceB)
-
-	link.handlerA.AddPeer(link.pidB, "wt0", rp.Key(link.wgKeyB))
-	link.handlerB.AddPeer(link.pidA, "wt0", rp.Key(link.wgKeyA))
-
-	return link
-}
-
-// complete simulates a completed rosenpass exchange: both ends derive the
-// same output key.
-func (l *handlerTestLink) complete(osk rp.Key) {
-	l.handlerA.HandshakeCompleted(l.pidB, osk)
-	l.handlerB.HandshakeCompleted(l.pidA, osk)
-}
-
-// expire simulates a failed key renewal on both ends.
-func (l *handlerTestLink) expire() {
-	l.handlerA.HandshakeExpired(l.pidB)
-	l.handlerB.HandshakeExpired(l.pidA)
-}
-
-func lastPSK(t *testing.T, m *mockIface) wgtypes.Key {
-	t.Helper()
-	m.mu.Lock()
-	defer m.mu.Unlock()
-	require.NotEmpty(t, m.calls, "expected at least one SetPresharedKey call")
-	return m.calls[len(m.calls)-1].psk
-}
-
-func TestHandshakeCompleted_SetsKeyAndInitializes(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	require.Equal(t, wgtypes.Key(osk), lastPSK(t, link.ifaceA), "completed exchange must program the osk")
-	require.False(t, link.ifaceA.calls[0].updateOnly, "first rotation must not be update-only")
-	require.True(t, link.handlerA.IsPeerInitialized(link.pidB), "peer must be initialized after first completed exchange")
-
-	link.complete(osk)
-	require.True(t, link.ifaceA.calls[1].updateOnly, "later rotations must be update-only")
-}
-
-// TestHandshakeExpired_BothSidesConverge encodes the core recovery invariant:
-// rosenpass renewals run over the tunnel that the PSK itself keys, so when a
-// renewal fails on both ends, both ends must fall back to the same key or the
-// tunnel can never handshake again.
-func TestHandshakeExpired_BothSidesConverge(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	keyA := lastPSK(t, link.ifaceA)
-	keyB := lastPSK(t, link.ifaceB)
-	require.NotEqual(t, wgtypes.Key(osk), keyA, "expired key must be rotated out")
-	require.Equal(t, keyA, keyB, "both ends must converge on the same key after expiry")
-
-	link.expire()
-	require.Equal(t, lastPSK(t, link.ifaceA), lastPSK(t, link.ifaceB),
-		"both ends must still converge after repeated expiries")
-}
-
-// TestHandshakeExpired_ExpiryWithoutCompletionConverges covers the bootstrap
-// case: the initial exchange never completed (the tunnel ran on the rendezvous
-// key), so an expiry must not replace the working key with an unrecoverable
-// one on either end.
-func TestHandshakeExpired_ExpiryWithoutCompletionConverges(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	link.expire()
-	require.Equal(t, lastPSK(t, link.ifaceA), lastPSK(t, link.ifaceB),
-		"both ends must converge when the exchange never completed")
-}
-
-// TestHandshakeExpired_RepeatedExpiryClearsInitialized: once renewals keep
-// failing, the peer must drop out of the initialized state so the next
-// connection reconfiguration reprograms the rendezvous key instead of
-// preserving a poisoned rosenpass-managed key.
-func TestHandshakeExpired_RepeatedExpiryClearsInitialized(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	link.expire()
-
-	require.False(t, link.handlerA.IsPeerInitialized(link.pidB),
-		"repeated expiries must clear the initialized state")
-	require.False(t, link.handlerB.IsPeerInitialized(link.pidA),
-		"repeated expiries must clear the initialized state")
-}
-
-// TestHandshakeCompleted_AfterExpiryRecovers: a completed exchange after a
-// desync must fully reset the recovery state.
-func TestHandshakeCompleted_AfterExpiryRecovers(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk1, osk2 rp.Key
-	osk1[0] = 1
-	osk2[0] = 2
-
-	link.complete(osk1)
-	link.expire()
-	link.expire()
-
-	link.complete(osk2)
-	require.Equal(t, wgtypes.Key(osk2), lastPSK(t, link.ifaceA), "new exchange must program the fresh osk")
-	require.True(t, link.handlerA.IsPeerInitialized(link.pidB), "peer must be initialized again after recovery")
-
-	link.expire()
-	require.Equal(t, lastPSK(t, link.ifaceA), lastPSK(t, link.ifaceB),
-		"recovered link must converge again on the next expiry")
-	require.NotEqual(t, wgtypes.Key(osk2), lastPSK(t, link.ifaceA), "expired key must be rotated out")
-}
-
-// TestHandshakeExpired_FirstExpiryRatchetsLastKey: the first expiry must
-// derive the replacement from the last shared key, so an attacker who only
-// blocks the renewal exchange gains nothing over the previous key.
-func TestHandshakeExpired_FirstExpiryRatchetsLastKey(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	require.Equal(t, RatchetKey(wgtypes.Key(osk)), lastPSK(t, link.ifaceA),
-		"first expiry must program the ratcheted key")
-	require.True(t, link.handlerA.IsPeerInitialized(link.pidB),
-		"ratchet step must keep the peer initialized so reconfigurations preserve the key")
-}
-
-// TestHandshakeExpired_RepeatedExpiryFallsBackToSeed: once the ratchet key
-// also fails, both ends must land on the same key that peer connections
-// program for uninitialized peers, so a reconnect completes the recovery.
-func TestHandshakeExpired_RepeatedExpiryFallsBackToSeed(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	link.expire()
-
-	seed, err := DeterministicSeedKey(link.wgKeyA.String(), link.wgKeyB.String())
-	require.NoError(t, err)
-	require.Equal(t, *seed, lastPSK(t, link.ifaceA), "repeated expiry must fall back to the seed key")
-	require.Equal(t, *seed, lastPSK(t, link.ifaceB), "repeated expiry must fall back to the seed key")
-}
-
-// TestHandshakeExpired_ConfiguredPSKUsedAsRendezvous: with an account-level
-// preshared key configured, the fallback must be that key, matching what peer
-// connections program for uninitialized peers.
-func TestHandshakeExpired_ConfiguredPSKUsedAsRendezvous(t *testing.T) {
-	psk := &[32]byte{0x77}
-	link := newHandlerTestLink(t, psk)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	link.expire()
-
-	require.Equal(t, wgtypes.Key(*psk), lastPSK(t, link.ifaceA),
-		"fallback must be the configured preshared key")
-	require.Equal(t, wgtypes.Key(*psk), lastPSK(t, link.ifaceB),
-		"fallback must be the configured preshared key on both ends")
-}
-
-// TestHandshakeExpired_ExpiryWritesAreUpdateOnly: expiry replacements must
-// never create a WireGuard peer that connection management has removed.
-func TestHandshakeExpired_ExpiryWritesAreUpdateOnly(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-
-	link.expire()
-	link.expire()
-
-	for _, call := range link.ifaceA.calls[1:] {
-		require.True(t, call.updateOnly, "expiry writes must be update-only")
-	}
-}
-
-// TestAddPeer_ReAddKeepsRecoveryState: reconnections re-add the peer on every
-// OnConnected; that must not reset the expiry chain state.
-func TestAddPeer_ReAddKeepsRecoveryState(t *testing.T) {
-	link := newHandlerTestLink(t, nil)
-
-	var osk rp.Key
-	osk[0] = 0x42
-	link.complete(osk)
-	link.expire()
-
-	link.handlerA.AddPeer(link.pidB, "wt0", rp.Key(link.wgKeyB))
-	require.True(t, link.handlerA.IsPeerInitialized(link.pidB),
-		"re-adding a known peer must keep its state")
-
-	link.expire()
-	seed, err := DeterministicSeedKey(link.wgKeyA.String(), link.wgKeyB.String())
-	require.NoError(t, err)
-	require.Equal(t, *seed, lastPSK(t, link.ifaceA),
-		"second expiry after re-add must continue to the seed fallback")
-}

client/internal/rosenpass/seed.go

@@ -1,28 +1,11 @@
 package rosenpass
 
 import (
-	"crypto/sha256"
 	"fmt"
 
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
 )
 
-// ratchetLabel domain-separates the expiry ratchet from other uses of the
-// rosenpass output key.
-const ratchetLabel = "netbird-rosenpass-expiry-ratchet"
-
-// RatchetKey derives the successor preshared key from the previous Rosenpass
-// output key. When a key expires without a completed renewal, both peers
-// advance their last shared key by one ratchet step: the expired key is
-// rotated out while both ends still converge on an identical, non-public
-// replacement without communicating.
-func RatchetKey(prev wgtypes.Key) wgtypes.Key {
-	input := make([]byte, 0, len(ratchetLabel)+len(prev))
-	input = append(input, ratchetLabel...)
-	input = append(input, prev[:]...)
-	return sha256.Sum256(input)
-}
-
 // DeterministicSeedKey derives a 32-byte WireGuard preshared key from a pair
 // of peer public keys. Both peers, given the same key pair, produce the same
 // output regardless of which side runs the function: the inputs are ordered

client/internal/routemanager/dynamic/route.go

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

client/internal/routemanager/dynamic/route_generic.go

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

client/internal/routemanager/dynamic/route_ios.go

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

client/internal/routemanager/manager.go

@@ -12,7 +12,6 @@ import (
 	"strings"
 	"sync"
 	"sync/atomic"
-	"syscall"
 	"time"
 
 	"github.com/google/uuid"
@@ -61,7 +60,6 @@ type Manager interface {
 	InitialRouteRange() []string
 	SetFirewall(firewall.Manager) error
 	SetDNSForwarderPort(port uint16)
-	ReconcilePeerAllowedIPs(peerKey string) error
 	Stop(stateManager *statemanager.Manager)
 }
 
@@ -233,30 +231,6 @@ func (m *DefaultManager) setupRefCounters(useNoop bool) {
 	)
 }
 
-// ReconcilePeerAllowedIPs re-applies every routed allowed IP currently tracked for the peer
-// onto the WireGuard device. The allowed-IP refcounter only calls its AddFunc (which pushes to
-// the device) on a prefix's 0->1 transition, so a peer whose device entry was rebuilt without a
-// matching refcounter change — e.g. a lazy connection cycling through idle->wake, which recreates
-// the WireGuard peer with the overlay /32 only — ends up missing routed prefixes the refcounter
-// still considers installed, and nothing retries. Calling this when the peer's WireGuard entry is
-// (re)created restores convergence. It is add-only and idempotent: AddAllowedIP is update-only, so
-// prefixes are re-added to an existing peer and an absent peer is left untouched.
-func (m *DefaultManager) ReconcilePeerAllowedIPs(peerKey string) error {
-	if m.allowedIPsRefCounter == nil {
-		return nil
-	}
-
-	return m.allowedIPsRefCounter.ReapplyMatching(
-		func(out string) bool { return out == peerKey },
-		func(prefix netip.Prefix) error {
-			if err := m.wgInterface.AddAllowedIP(peerKey, prefix); err != nil {
-				return fmt.Errorf("add allowed IP %s for peer %s: %w", prefix, peerKey, err)
-			}
-			return nil
-		},
-	)
-}
-
 // Init sets up the routing
 func (m *DefaultManager) Init() error {
 	m.routeSelector = m.initSelector()
@@ -290,11 +264,7 @@ func (m *DefaultManager) initSelector() *routeselector.RouteSelector {
 
 	// restore selector state if it exists
 	if err := m.stateManager.LoadState(state); err != nil {
-		if errors.Is(err, syscall.ENOSYS) {
-			log.Debugf("route selector state unavailable on this platform: %v", err)
-		} else {
-			log.Warnf("failed to load state: %v", err)
-		}
+		log.Warnf("failed to load state: %v", err)
 		return routeselector.NewRouteSelector()
 	}
 

client/internal/routemanager/mock.go

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

client/internal/routemanager/reconcile_test.go

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

client/internal/routemanager/refcounter/refcounter.go

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

client/internal/routemanager/refcounter/refcounter_test.go

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

client/internal/statemanager/manager.go

@@ -1,7 +1,6 @@
 package statemanager
 
 import (
-	"bytes"
 	"context"
 	"encoding/json"
 	"errors"
@@ -306,11 +305,6 @@ func (m *Manager) loadStateFile(deleteCorrupt bool) (map[string]json.RawMessage,
 
 	var rawStates map[string]json.RawMessage
 	if err := json.Unmarshal(data, &rawStates); err != nil {
-		if len(bytes.TrimSpace(data)) == 0 {
-			log.Warnf("state file %s is empty (%d bytes)", m.filePath, len(data))
-		} else {
-			log.Warnf("state file %s has malformed content (%d bytes)", m.filePath, len(data))
-		}
 		m.handleCorruptedState(deleteCorrupt)
 		return nil, fmt.Errorf("unmarshal states: %w", err)
 	}

client/ios/NetBirdSDK/client.go

@@ -233,9 +233,6 @@ func (c *Client) DebugBundle(anonymize bool) (string, error) {
 		deps.SyncResponse = resp
 
 		if e := cc.Engine(); e != nil {
-			deps.RefreshStatus = func() {
-				e.RunHealthProbes(context.Background(), true)
-			}
 			if cm := e.GetClientMetrics(); cm != nil {
 				deps.ClientMetrics = cm
 			}

client/ios/NetBirdSDK/login.go

@@ -44,25 +44,10 @@ type Auth struct {
 // NewAuth instantiate Auth struct and validate the management URL
 func NewAuth(cfgPath string, mgmURL string) (*Auth, error) {
 	inputCfg := profilemanager.ConfigInput{
-		ConfigPath:    cfgPath,
 		ManagementURL: mgmURL,
 	}
 
-	// Load the existing config when a config file is already present so an
-	// interactive re-login reuses the peer's persisted WireGuard private key
-	// (and thus its identity) instead of generating a fresh one. Generating a
-	// new key registers a brand-new peer on the management server on every
-	// re-auth (named after the fallback hostname). Only fall back to a fresh
-	// in-memory config for the first-time login when no config file exists yet.
-	// DirectUpdateOrCreateConfig uses non-atomic writes so it also works inside
-	// the tvOS App Group sandbox where atomic temp-file+rename is blocked.
-	var cfg *profilemanager.Config
-	var err error
-	if cfgPath != "" {
-		cfg, err = profilemanager.DirectUpdateOrCreateConfig(inputCfg)
-	} else {
-		cfg, err = profilemanager.CreateInMemoryConfig(inputCfg)
-	}
+	cfg, err := profilemanager.CreateInMemoryConfig(inputCfg)
 	if err != nil {
 		return nil, err
 	}

client/ios/NetBirdSDK/version.go

@@ -1,12 +0,0 @@
-//go:build ios
-
-package NetBirdSDK
-
-import "github.com/netbirdio/netbird/version"
-
-// GoClientVersion returns the NetBird Go client version that was baked into
-// the framework at compile time via
-// -ldflags "-X github.com/netbirdio/netbird/version.version=<version>".
-func GoClientVersion() string {
-	return version.NetbirdVersion()
-}

client/mdm/canonical_loaders.go

@@ -22,7 +22,6 @@ var allKeys = []string{
 	KeyDisableMetricsCollection,
 	KeyAllowServerSSH,
 	KeyDisableAutoConnect,
-	KeyDisableAutostart,
 	KeyPreSharedKey,
 	KeyRosenpassEnabled,
 	KeyRosenpassPermissive,

client/mdm/policy.go

@@ -20,10 +20,10 @@ import (
 // names (lowerCamelCase) so the daemon can map a Policy key directly to a
 // configuration field.
 const (
-	KeyManagementURL         = "managementURL"
-	KeyDisableUpdateSettings = "disableUpdateSettings"
-	KeyDisableProfiles       = "disableProfiles"
-	KeyDisableNetworks       = "disableNetworks"
+	KeyManagementURL            = "managementURL"
+	KeyDisableUpdateSettings    = "disableUpdateSettings"
+	KeyDisableProfiles          = "disableProfiles"
+	KeyDisableNetworks          = "disableNetworks"
 	// KeyDisableAdvancedView gates the advanced-view section in the
 	// upcoming UI revision. UI-only: NOT stored on Config, not
 	// applied by applyMDMPolicy, not rejectable via SetConfig. The
@@ -37,16 +37,10 @@ const (
 	KeyDisableMetricsCollection = "disableMetricsCollection"
 	KeyAllowServerSSH           = "allowServerSSH"
 	KeyDisableAutoConnect       = "disableAutoConnect"
-	// KeyDisableAutostart suppresses the GUI's fresh-install
-	// launch-on-login default and marks the Settings toggle as
-	// MDM-managed. UI-only: NOT stored on Config and not applied by
-	// applyMDMPolicy; the GUI reads it directly and it appears in
-	// GetConfigResponse.mDMManagedFields when set.
-	KeyDisableAutostart    = "disableAutostart"
-	KeyPreSharedKey        = "preSharedKey"
-	KeyRosenpassEnabled    = "rosenpassEnabled"
-	KeyRosenpassPermissive = "rosenpassPermissive"
-	KeyWireguardPort       = "wireguardPort"
+	KeyPreSharedKey             = "preSharedKey"
+	KeyRosenpassEnabled         = "rosenpassEnabled"
+	KeyRosenpassPermissive      = "rosenpassPermissive"
+	KeyWireguardPort            = "wireguardPort"
 
 	// Split tunnel is modeled as a single conceptual policy with two
 	// registry/plist values. KeySplitTunnelMode is the discriminator

client/server/server.go

@@ -181,7 +181,7 @@ func (s *Server) Start() error {
 		log.Warnf("failed to redirect stderr: %v", err)
 	}
 
-	if err := RestoreResidualState(s.rootCtx, s.profileManager.GetStatePath()); err != nil {
+	if err := restoreResidualState(s.rootCtx, s.profileManager.GetStatePath()); err != nil {
 		log.Warnf(errRestoreResidualState, err)
 	}
 
@@ -551,7 +551,7 @@ func (s *Server) Login(callerCtx context.Context, msg *proto.LoginRequest) (*pro
 	s.actCancel = cancel
 	s.mutex.Unlock()
 
-	if err := RestoreResidualState(s.rootCtx, s.profileManager.GetStatePath()); err != nil {
+	if err := restoreResidualState(s.rootCtx, s.profileManager.GetStatePath()); err != nil {
 		log.Warnf(errRestoreResidualState, err)
 	}
 
@@ -828,7 +828,6 @@ func (s *Server) WaitSSOLogin(callerCtx context.Context, msg *proto.WaitSSOLogin
 		return nil, err
 	}
 
-	log.Infof("SSO login flow finished, returning success to caller")
 	return &proto.WaitSSOLoginResponse{
 		Email: tokenInfo.Email,
 	}, nil
@@ -836,7 +835,6 @@ func (s *Server) WaitSSOLogin(callerCtx context.Context, msg *proto.WaitSSOLogin
 
 // Up starts engine work in the daemon.
 func (s *Server) Up(callerCtx context.Context, msg *proto.UpRequest) (*proto.UpResponse, error) {
-	log.Infof("up request received")
 	s.mutex.Lock()
 	// clientRunning is the daemon-intent flag (set by previous Up/Start, cleared
 	// by Down). connectionGoroutineRunning() reports whether the previous retry-loop
@@ -858,7 +856,7 @@ func (s *Server) Up(callerCtx context.Context, msg *proto.UpRequest) (*proto.UpR
 
 		return s.waitForUp(callerCtx)
 	}
-	if err := RestoreResidualState(callerCtx, s.profileManager.GetStatePath()); err != nil {
+	if err := restoreResidualState(callerCtx, s.profileManager.GetStatePath()); err != nil {
 		log.Warnf(errRestoreResidualState, err)
 	}
 
@@ -1081,10 +1079,7 @@ func (s *Server) Down(ctx context.Context, _ *proto.DownRequest) (*proto.DownRes
 
 	if err := s.cleanupConnection(); err != nil {
 		s.mutex.Unlock()
-		if errors.Is(err, ErrServiceNotUp) {
-			log.Debugf("Down called while service not up: %v", err)
-			return nil, err
-		}
+		// todo review to update the status in case any type of error
 		log.Errorf("failed to shut down properly: %v", err)
 		return nil, err
 	}
@@ -1157,7 +1152,7 @@ func (s *Server) cleanupConnection() error {
 	// making the run loop the sole owner of engine shutdown.
 	if engine != nil {
 		if err := engine.Stop(); err != nil {
-			log.Errorf("failed to stop engine during cleanup: %v", err)
+			return err
 		}
 	}
 

client/server/state.go

@@ -46,7 +46,7 @@ func (s *Server) CleanState(ctx context.Context, req *proto.CleanStateRequest) (
 
 	if req.All {
 		// Reuse existing cleanup logic for all states
-		if err := RestoreResidualState(ctx, statePath); err != nil {
+		if err := restoreResidualState(ctx, statePath); err != nil {
 			return nil, status.Errorf(codes.Internal, "failed to clean all states: %v", err)
 		}
 
@@ -113,9 +113,9 @@ func (s *Server) DeleteState(ctx context.Context, req *proto.DeleteStateRequest)
 	}, nil
 }
 
-// RestoreResidualState checks if the client was not shut down in a clean way and restores residual if required.
+// restoreResidualState checks if the client was not shut down in a clean way and restores residual if required.
 // Otherwise, we might not be able to connect to the management server to retrieve new config.
-func RestoreResidualState(ctx context.Context, statePath string) error {
+func restoreResidualState(ctx context.Context, statePath string) error {
 	if statePath == "" {
 		return nil
 	}

client/ssh/config/manager.go

@@ -14,7 +14,6 @@ import (
 	log "github.com/sirupsen/logrus"
 
 	nbssh "github.com/netbirdio/netbird/client/ssh"
-	"github.com/netbirdio/netbird/shared/management/domain"
 )
 
 const (
@@ -219,20 +218,11 @@ func (m *Manager) buildHostPatterns(peer PeerSSHInfo) []string {
 	if peer.IPv6.IsValid() {
 		hostPatterns = append(hostPatterns, peer.IPv6.String())
 	}
-	// Peer FQDNs and hostnames originate from remote peers, so they must be
-	// validated as plain DNS names before being embedded in the ssh_config
-	// "Match host" pattern list. This prevents injection of arbitrary
-	// ssh_config directives via embedded quotes, whitespace, newlines, the
-	// comma pattern separator, or the "*"/"?" pattern metacharacters.
-	if domain.IsValidDomainNoWildcard(peer.FQDN) {
+	if peer.FQDN != "" {
 		hostPatterns = append(hostPatterns, peer.FQDN)
-	} else if peer.FQDN != "" {
-		log.Warnf("skipping peer FQDN with invalid characters in SSH config: %q", peer.FQDN)
 	}
-	if peer.Hostname != peer.FQDN && domain.IsValidDomainNoWildcard(peer.Hostname) {
+	if peer.Hostname != "" && peer.Hostname != peer.FQDN {
 		hostPatterns = append(hostPatterns, peer.Hostname)
-	} else if peer.Hostname != "" && peer.Hostname != peer.FQDN {
-		log.Warnf("skipping peer hostname with invalid characters in SSH config: %q", peer.Hostname)
 	}
 	return hostPatterns
 }

client/ssh/config/manager_test.go

@@ -148,45 +148,6 @@ func TestManager_MatchHostFormat(t *testing.T) {
 		"should use Match host with comma-separated patterns")
 }
 
-func TestManager_HostPatternInjection(t *testing.T) {
-	tempDir, err := os.MkdirTemp("", "netbird-ssh-config-test")
-	require.NoError(t, err)
-	defer func() { assert.NoError(t, os.RemoveAll(tempDir)) }()
-
-	manager := &Manager{
-		sshConfigDir:  filepath.Join(tempDir, "ssh_config.d"),
-		sshConfigFile: "99-netbird.conf",
-	}
-
-	// A malicious peer FQDN/hostname attempts to break out of the Match host
-	// directive and inject arbitrary ssh_config (a ProxyCommand executing a
-	// command). It must be rejected, not written to the config.
-	peers := []PeerSSHInfo{
-		{
-			Hostname: "evil\"\n    ProxyCommand touch /tmp/pwned\nHost x",
-			IP:       netip.MustParseAddr("100.125.1.1"),
-			FQDN:     "evil\"\n    ProxyCommand touch /tmp/pwned\nHost x.nb.internal",
-		},
-		{Hostname: "peer2", IP: netip.MustParseAddr("100.125.1.2"), FQDN: "peer2.nb.internal"},
-	}
-
-	err = manager.SetupSSHClientConfig(peers)
-	require.NoError(t, err)
-
-	configPath := filepath.Join(manager.sshConfigDir, manager.sshConfigFile)
-	content, err := os.ReadFile(configPath)
-	require.NoError(t, err)
-	configStr := string(content)
-
-	assert.NotContains(t, configStr, "ProxyCommand touch /tmp/pwned",
-		"injected directive must not appear in generated config")
-	assert.NotContains(t, configStr, "evil",
-		"malicious pattern must be dropped entirely")
-	// The valid peer must still be present, on a single Match host line.
-	assert.Contains(t, configStr, "Match host \"100.125.1.1,100.125.1.2,peer2.nb.internal,peer2\"",
-		"valid peers must survive, injected patterns dropped")
-}
-
 func TestManager_ForcedSSHConfig(t *testing.T) {
 	// Set force environment variable
 	t.Setenv(EnvForceSSHConfig, "true")

client/ssh/server/getent_unix.go

@@ -69,8 +69,7 @@ func parseGetentPasswd(output string) (*user.User, string, error) {
 
 // validateGetentInput checks that the input is safe to pass to getent or id.
 // Allows POSIX usernames, numeric UIDs, and common NSS extensions
-// (@ for Kerberos, $ for Samba, + for NIS compat). A leading hyphen is
-// rejected so the input can never be parsed as a command-line flag.
+// (@ for Kerberos, $ for Samba, + for NIS compat).
 func validateGetentInput(input string) bool {
 	maxLen := 32
 	if runtime.GOOS == "linux" {
@@ -81,10 +80,6 @@ func validateGetentInput(input string) bool {
 		return false
 	}
 
-	if input[0] == '-' {
-		return false
-	}
-
 	for _, r := range input {
 		if isAllowedGetentChar(r) {
 			continue

client/ssh/server/getent_unix_test.go

@@ -157,9 +157,6 @@ func TestValidateGetentInput(t *testing.T) {
 		{"numeric UID", "1001", true},
 		{"dots and underscores", "alice.bob_test", true},
 		{"hyphen", "alice-bob", true},
-		{"leading hyphen rejected", "-i", false},
-		{"leading double hyphen rejected", "--no-idn", false},
-		{"lone hyphen rejected", "-", false},
 		{"kerberos principal", "user@REALM", true},
 		{"samba machine account", "MACHINE$", true},
 		{"NIS compat", "+user", true},

client/status/status.go

@@ -746,8 +746,6 @@ func ToProtoFullStatus(fullStatus peer.FullStatus) *proto.FullStatus {
 		pbFullStatus.DnsServers = append(pbFullStatus.DnsServers, pbDnsState)
 	}
 
-	pbFullStatus.Events = fullStatus.Events
-
 	return &pbFullStatus
 }
 

client/system/info.go

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

client/system/process.go

@@ -7,7 +7,7 @@ import (
 	"os"
 	"slices"
 
-	"github.com/shirou/gopsutil/v4/process"
+	"github.com/shirou/gopsutil/v3/process"
 )
 
 // getRunningProcesses returns a list of running process paths. The context bounds the work:

client/system/process_test.go

@@ -4,7 +4,7 @@ import (
 	"context"
 	"testing"
 
-	"github.com/shirou/gopsutil/v4/process"
+	"github.com/shirou/gopsutil/v3/process"
 )
 
 func Benchmark_getRunningProcesses(b *testing.B) {

client/ui/autostart_default.go

@@ -1,121 +0,0 @@
-//go:build !android && !ios && !freebsd && !js
-
-package main
-
-import (
-	"context"
-	"os"
-	"path/filepath"
-
-	log "github.com/sirupsen/logrus"
-
-	"github.com/netbirdio/netbird/client/internal/profilemanager"
-	"github.com/netbirdio/netbird/client/mdm"
-	"github.com/netbirdio/netbird/client/ui/preferences"
-	"github.com/netbirdio/netbird/client/ui/services"
-)
-
-// autostartDefaultState carries the guard inputs of the one-time autostart
-// default decision so the decision itself stays a pure, testable function.
-type autostartDefaultState struct {
-	supported    bool
-	mdmDisabled  bool
-	priorInstall bool
-}
-
-// shouldEnableAutostartDefault applies the first-run guards in order and
-// returns whether autostart may be enabled, plus the reason when it may not.
-func shouldEnableAutostartDefault(s autostartDefaultState) (bool, string) {
-	switch {
-	case !s.supported:
-		return false, "autostart not supported on this platform"
-	case s.mdmDisabled:
-		return false, "autostart disabled by MDM policy"
-	case s.priorInstall:
-		return false, "existing NetBird installation"
-	}
-	return true, ""
-}
-
-// autostartDisabledByMDM reports whether the MDM policy manages the
-// disableAutostart key in a way that must suppress the default. An
-// unparseable managed value is treated as disabled to stay on the safe side.
-func autostartDisabledByMDM(policy *mdm.Policy) bool {
-	if !policy.HasKey(mdm.KeyDisableAutostart) {
-		return false
-	}
-	disabled, ok := policy.GetBool(mdm.KeyDisableAutostart)
-	return !ok || disabled
-}
-
-// netbirdFootprintExists reports whether the machine already carries NetBird
-// daemon config or state, meaning this is not a genuinely fresh install. It is
-// the update-safety gate for the autostart default: upgrading users always
-// have a footprint, so an update can never trigger a autostart entry write.
-func netbirdFootprintExists() bool {
-	candidates := []string{
-		profilemanager.DefaultConfigPath,
-		filepath.Join(profilemanager.DefaultConfigPathDir, "config.json"),
-		filepath.Join(profilemanager.DefaultConfigPathDir, "state.json"),
-	}
-	for _, path := range candidates {
-		if path != "" && fileExists(path) {
-			return true
-		}
-	}
-	return false
-}
-
-// applyAutostartDefault runs the one-time launch-on-login default for genuinely
-// fresh installs. The autostartInitialized marker is persisted before any
-// enable attempt so a crash mid-flow degrades to "never enabled" instead of
-// retrying autostart entry writes on every launch. A user's later disable in
-// Settings is never overridden: the marker guarantees at-most-once, ever.
-func applyAutostartDefault(ctx context.Context, autostart *services.Autostart, prefs *preferences.Store, prefsFileExisted bool) {
-	mdmDisabled := autostartDisabledByMDM(mdm.LoadPolicy())
-
-	if mdmDisabled {
-		if enabled, err := autostart.IsEnabled(ctx); err != nil {
-			log.Warnf("MDM disableAutostart: read autostart state: %v", err)
-		} else if enabled {
-			if err := autostart.SetEnabled(ctx, false); err != nil {
-				log.Warnf("MDM disableAutostart: force off failed: %v", err)
-			} else {
-				log.Info("MDM disableAutostart enforced: autostart turned off")
-			}
-		}
-	}
-
-	priorFootprint := netbirdFootprintExists() || prefsFileExisted
-
-	if prefs.Get().AutostartInitialized {
-		return
-	}
-	if err := prefs.SetAutostartInitialized(true); err != nil {
-		log.Warnf("persist autostart marker, skipping autostart default: %v", err)
-		return
-	}
-
-	state := autostartDefaultState{
-		supported:    autostart.Supported(ctx),
-		mdmDisabled:  mdmDisabled,
-		priorInstall: priorFootprint,
-	}
-	enable, reason := shouldEnableAutostartDefault(state)
-	if !enable {
-		log.Debugf("skipping autostart default: %s", reason)
-		return
-	}
-
-	if err := autostart.SetEnabled(ctx, true); err != nil {
-		log.Warnf("enable autostart on fresh install: %v", err)
-		return
-	}
-	log.Info("autostart enabled by default on fresh install")
-}
-
-// fileExists reports whether path exists.
-func fileExists(path string) bool {
-	_, err := os.Stat(path)
-	return err == nil
-}

client/ui/autostart_default_test.go

@@ -1,125 +0,0 @@
-//go:build !android && !ios && !freebsd && !js
-
-package main
-
-import (
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-
-	"github.com/netbirdio/netbird/client/mdm"
-)
-
-func TestShouldEnableAutostartDefault(t *testing.T) {
-	allPass := autostartDefaultState{
-		supported:    true,
-		mdmDisabled:  false,
-		priorInstall: false,
-	}
-
-	tests := []struct {
-		name       string
-		mutate     func(*autostartDefaultState)
-		wantEnable bool
-		wantReason string
-	}{
-		{
-			name:       "fresh install with all guards passing enables",
-			mutate:     func(*autostartDefaultState) {},
-			wantEnable: true,
-		},
-		{
-			name:       "unsupported platform skips",
-			mutate:     func(s *autostartDefaultState) { s.supported = false },
-			wantReason: "autostart not supported on this platform",
-		},
-		{
-			name:       "MDM disable skips",
-			mutate:     func(s *autostartDefaultState) { s.mdmDisabled = true },
-			wantReason: "autostart disabled by MDM policy",
-		},
-		{
-			name:       "existing installation (upgrade) skips",
-			mutate:     func(s *autostartDefaultState) { s.priorInstall = true },
-			wantReason: "existing NetBird installation",
-		},
-		{
-			name: "unsupported wins over every other guard",
-			mutate: func(s *autostartDefaultState) {
-				s.supported = false
-				s.mdmDisabled = true
-				s.priorInstall = true
-			},
-			wantReason: "autostart not supported on this platform",
-		},
-		{
-			name: "MDM disable wins over prior install",
-			mutate: func(s *autostartDefaultState) {
-				s.mdmDisabled = true
-				s.priorInstall = true
-			},
-			wantReason: "autostart disabled by MDM policy",
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			state := allPass
-			tc.mutate(&state)
-			enable, reason := shouldEnableAutostartDefault(state)
-			assert.Equal(t, tc.wantEnable, enable, "enable decision should match for state %+v", state)
-			assert.Equal(t, tc.wantReason, reason, "skip reason should identify the failing guard")
-		})
-	}
-}
-
-func TestAutostartDisabledByMDM(t *testing.T) {
-	tests := []struct {
-		name   string
-		values map[string]any
-		want   bool
-	}{
-		{
-			name:   "empty policy does not disable",
-			values: nil,
-			want:   false,
-		},
-		{
-			name:   "unrelated managed keys do not disable",
-			values: map[string]any{mdm.KeyDisableAutoConnect: true},
-			want:   false,
-		},
-		{
-			name:   "disableAutostart true disables",
-			values: map[string]any{mdm.KeyDisableAutostart: true},
-			want:   true,
-		},
-		{
-			name:   "disableAutostart registry DWORD 1 disables",
-			values: map[string]any{mdm.KeyDisableAutostart: int64(1)},
-			want:   true,
-		},
-		{
-			name:   "disableAutostart string true disables",
-			values: map[string]any{mdm.KeyDisableAutostart: "true"},
-			want:   true,
-		},
-		{
-			name:   "disableAutostart explicit false allows",
-			values: map[string]any{mdm.KeyDisableAutostart: false},
-			want:   false,
-		},
-		{
-			name:   "unparseable managed value is treated as disabled",
-			values: map[string]any{mdm.KeyDisableAutostart: "not-a-bool"},
-			want:   true,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			got := autostartDisabledByMDM(mdm.NewPolicy(tc.values))
-			assert.Equal(t, tc.want, got, "MDM disable decision should match for values %v", tc.values)
-		})
-	}
-}

client/ui/frontend/src/app.tsx

@@ -16,13 +16,10 @@ import LoginWaitingForBrowserDialog from "@/modules/login/LoginWaitingForBrowser
 import { initI18n } from "@/lib/i18n";
 import { initPlatform } from "@/lib/platform";
 import { initLogForwarding } from "@/lib/logs";
-import { initStallWatch } from "@/lib/stallwatch";
 
 // Must run first so even init-time logs reach the Go log pipeline.
 initLogForwarding();
 
-initStallWatch();
-
 welcome();
 
 Promise.all([