Revert "[client] Supress ICE signaling (#5820)"

This reverts commit 5a89e6621b.

client/internal/engine.go

@@ -571,7 +571,7 @@ func (e *Engine) Start(netbirdConfig *mgmProto.NetbirdConfig, mgmtURL *url.URL)
 	e.connMgr.Start(e.ctx)
 
 	e.srWatcher = guard.NewSRWatcher(e.signal, e.relayManager, e.mobileDep.IFaceDiscover, iceCfg)
-	e.srWatcher.Start(peer.IsForceRelayed())
+	e.srWatcher.Start()
 
 	e.receiveSignalEvents()
 	e.receiveManagementEvents()

client/internal/peer/conn.go

@@ -185,20 +185,17 @@ func (conn *Conn) Open(engineCtx context.Context) error {
 
 	conn.workerRelay = NewWorkerRelay(conn.ctx, conn.Log, isController(conn.config), conn.config, conn, conn.relayManager)
 
-	forceRelay := IsForceRelayed()
-	if !forceRelay {
-		relayIsSupportedLocally := conn.workerRelay.RelayIsSupportedLocally()
-		workerICE, err := NewWorkerICE(conn.ctx, conn.Log, conn.config, conn, conn.signaler, conn.iFaceDiscover, conn.statusRecorder, relayIsSupportedLocally)
-		if err != nil {
-			return err
-		}
-		conn.workerICE = workerICE
+	relayIsSupportedLocally := conn.workerRelay.RelayIsSupportedLocally()
+	workerICE, err := NewWorkerICE(conn.ctx, conn.Log, conn.config, conn, conn.signaler, conn.iFaceDiscover, conn.statusRecorder, relayIsSupportedLocally)
+	if err != nil {
+		return err
 	}
+	conn.workerICE = workerICE
 
 	conn.handshaker = NewHandshaker(conn.Log, conn.config, conn.signaler, conn.workerICE, conn.workerRelay, conn.metricsStages)
 
 	conn.handshaker.AddRelayListener(conn.workerRelay.OnNewOffer)
-	if !forceRelay {
+	if !isForceRelayed() {
 		conn.handshaker.AddICEListener(conn.workerICE.OnNewOffer)
 	}
 
@@ -254,9 +251,7 @@ func (conn *Conn) Close(signalToRemote bool) {
 		conn.wgWatcherCancel()
 	}
 	conn.workerRelay.CloseConn()
-	if conn.workerICE != nil {
-		conn.workerICE.Close()
-	}
+	conn.workerICE.Close()
 
 	if conn.wgProxyRelay != nil {
 		err := conn.wgProxyRelay.CloseConn()
@@ -299,9 +294,7 @@ func (conn *Conn) OnRemoteAnswer(answer OfferAnswer) {
 // OnRemoteCandidate Handles ICE connection Candidate provided by the remote peer.
 func (conn *Conn) OnRemoteCandidate(candidate ice.Candidate, haRoutes route.HAMap) {
 	conn.dumpState.RemoteCandidate()
-	if conn.workerICE != nil {
-		conn.workerICE.OnRemoteCandidate(candidate, haRoutes)
-	}
+	conn.workerICE.OnRemoteCandidate(candidate, haRoutes)
 }
 
 // SetOnConnected sets a handler function to be triggered by Conn when a new connection to a remote peer established
@@ -719,35 +712,33 @@ func (conn *Conn) evalStatus() ConnStatus {
 	return StatusConnecting
 }
 
-// isConnectedOnAllWay evaluates the overall connection status based on ICE and Relay transports.
-//
-// The result is a tri-state:
-//   - ConnStatusConnected:          all available transports are up
-//   - ConnStatusPartiallyConnected: relay is up but ICE is still pending/reconnecting
-//   - ConnStatusDisconnected:       no working transport
-func (conn *Conn) isConnectedOnAllWay() (status guard.ConnStatus) {
+func (conn *Conn) isConnectedOnAllWay() (connected bool) {
+	// would be better to protect this with a mutex, but it could cause deadlock with Close function
+
 	defer func() {
-		if status == guard.ConnStatusDisconnected {
+		if !connected {
 			conn.logTraceConnState()
 		}
 	}()
 
-	iceWorkerCreated := conn.workerICE != nil
-
-	var iceInProgress bool
-	if iceWorkerCreated {
-		iceInProgress = conn.workerICE.InProgress()
+	// For JS platform: only relay connection is supported
+	if runtime.GOOS == "js" {
+		return conn.statusRelay.Get() == worker.StatusConnected
 	}
 
-	return evalConnStatus(connStatusInputs{
-		forceRelay:          IsForceRelayed(),
-		peerUsesRelay:       conn.workerRelay.IsRelayConnectionSupportedWithPeer(),
-		relayConnected:      conn.statusRelay.Get() == worker.StatusConnected,
-		remoteSupportsICE:   conn.handshaker.RemoteICESupported(),
-		iceWorkerCreated:    iceWorkerCreated,
-		iceStatusConnecting: conn.statusICE.Get() != worker.StatusDisconnected,
-		iceInProgress:       iceInProgress,
-	})
+	// For non-JS platforms: check ICE connection status
+	if conn.statusICE.Get() == worker.StatusDisconnected && !conn.workerICE.InProgress() {
+		return false
+	}
+
+	// If relay is supported with peer, it must also be connected
+	if conn.workerRelay.IsRelayConnectionSupportedWithPeer() {
+		if conn.statusRelay.Get() == worker.StatusDisconnected {
+			return false
+		}
+	}
+
+	return true
 }
 
 func (conn *Conn) enableWgWatcherIfNeeded(enabledTime time.Time) {
@@ -935,43 +926,3 @@ func isController(config ConnConfig) bool {
 func isRosenpassEnabled(remoteRosenpassPubKey []byte) bool {
 	return remoteRosenpassPubKey != nil
 }
-
-func evalConnStatus(in connStatusInputs) guard.ConnStatus {
-	// "Relay up and needed" — the peer uses relay and the transport is connected.
-	relayUsedAndUp := in.peerUsesRelay && in.relayConnected
-
-	// Force-relay mode: ICE never runs. Relay is the only transport and must be up.
-	if in.forceRelay {
-		return boolToConnStatus(relayUsedAndUp)
-	}
-
-	// Remote peer doesn't support ICE, or we haven't created the worker yet:
-	// relay is the only possible transport.
-	if !in.remoteSupportsICE || !in.iceWorkerCreated {
-		return boolToConnStatus(relayUsedAndUp)
-	}
-
-	// ICE counts as "up" when the status is anything other than Disconnected, OR
-	// when a negotiation is currently in progress (so we don't spam offers while one is in flight).
-	iceUp := in.iceStatusConnecting || in.iceInProgress
-
-	// Relay side is acceptable if the peer doesn't rely on relay, or relay is connected.
-	relayOK := !in.peerUsesRelay || in.relayConnected
-
-	switch {
-	case iceUp && relayOK:
-		return guard.ConnStatusConnected
-	case relayUsedAndUp:
-		// Relay is up but ICE is down — partially connected.
-		return guard.ConnStatusPartiallyConnected
-	default:
-		return guard.ConnStatusDisconnected
-	}
-}
-
-func boolToConnStatus(connected bool) guard.ConnStatus {
-	if connected {
-		return guard.ConnStatusConnected
-	}
-	return guard.ConnStatusDisconnected
-}

client/internal/peer/conn_status.go

@@ -13,20 +13,6 @@ const (
 	StatusConnected
 )
 
-// connStatusInputs is the primitive-valued snapshot of the state that drives the
-// tri-state connection classification. Extracted so the decision logic can be unit-tested
-// without constructing full Worker/Handshaker objects.
-type connStatusInputs struct {
-	forceRelay          bool // NB_FORCE_RELAY or JS/WASM
-	peerUsesRelay       bool // remote peer advertises relay support AND local has relay
-	relayConnected      bool // statusRelay reports Connected (independent of whether peer uses relay)
-	remoteSupportsICE   bool // remote peer sent ICE credentials
-	iceWorkerCreated    bool // local WorkerICE exists (false in force-relay mode)
-	iceStatusConnecting bool // statusICE is anything other than Disconnected
-	iceInProgress       bool // a negotiation is currently in flight
-}
-
-
 // ConnStatus describe the status of a peer's connection
 type ConnStatus int32
 

client/internal/peer/conn_status_eval_test.go

@@ -1,201 +0,0 @@
-package peer
-
-import (
-	"testing"
-
-	"github.com/netbirdio/netbird/client/internal/peer/guard"
-)
-
-func TestEvalConnStatus_ForceRelay(t *testing.T) {
-	tests := []struct {
-		name string
-		in   connStatusInputs
-		want guard.ConnStatus
-	}{
-		{
-			name: "force relay, peer uses relay, relay up",
-			in: connStatusInputs{
-				forceRelay:     true,
-				peerUsesRelay:  true,
-				relayConnected: true,
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "force relay, peer uses relay, relay down",
-			in: connStatusInputs{
-				forceRelay:     true,
-				peerUsesRelay:  true,
-				relayConnected: false,
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-		{
-			name: "force relay, peer does NOT use relay - disconnected forever",
-			in: connStatusInputs{
-				forceRelay:     true,
-				peerUsesRelay:  false,
-				relayConnected: true,
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			if got := evalConnStatus(tc.in); got != tc.want {
-				t.Fatalf("evalConnStatus = %v, want %v", got, tc.want)
-			}
-		})
-	}
-}
-
-func TestEvalConnStatus_ICEUnavailable(t *testing.T) {
-	tests := []struct {
-		name string
-		in   connStatusInputs
-		want guard.ConnStatus
-	}{
-		{
-			name: "remote does not support ICE, peer uses relay, relay up",
-			in: connStatusInputs{
-				peerUsesRelay:     true,
-				relayConnected:    true,
-				remoteSupportsICE: false,
-				iceWorkerCreated:  true,
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "remote does not support ICE, peer uses relay, relay down",
-			in: connStatusInputs{
-				peerUsesRelay:     true,
-				relayConnected:    false,
-				remoteSupportsICE: false,
-				iceWorkerCreated:  true,
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-		{
-			name: "ICE worker not yet created, relay up",
-			in: connStatusInputs{
-				peerUsesRelay:     true,
-				relayConnected:    true,
-				remoteSupportsICE: true,
-				iceWorkerCreated:  false,
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "remote does not support ICE, peer does not use relay",
-			in: connStatusInputs{
-				peerUsesRelay:     false,
-				relayConnected:    false,
-				remoteSupportsICE: false,
-				iceWorkerCreated:  true,
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			if got := evalConnStatus(tc.in); got != tc.want {
-				t.Fatalf("evalConnStatus = %v, want %v", got, tc.want)
-			}
-		})
-	}
-}
-
-func TestEvalConnStatus_FullyAvailable(t *testing.T) {
-	base := connStatusInputs{
-		remoteSupportsICE: true,
-		iceWorkerCreated:  true,
-	}
-
-	tests := []struct {
-		name    string
-		mutator func(*connStatusInputs)
-		want    guard.ConnStatus
-	}{
-		{
-			name: "ICE connected, relay connected, peer uses relay",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = true
-				in.relayConnected = true
-				in.iceStatusConnecting = true
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "ICE connected, peer does NOT use relay",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = false
-				in.relayConnected = false
-				in.iceStatusConnecting = true
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "ICE InProgress only, peer does NOT use relay",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = false
-				in.iceStatusConnecting = false
-				in.iceInProgress = true
-			},
-			want: guard.ConnStatusConnected,
-		},
-		{
-			name: "ICE down, relay up, peer uses relay -> partial",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = true
-				in.relayConnected = true
-				in.iceStatusConnecting = false
-				in.iceInProgress = false
-			},
-			want: guard.ConnStatusPartiallyConnected,
-		},
-		{
-			name: "ICE down, peer does NOT use relay -> disconnected",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = false
-				in.relayConnected = false
-				in.iceStatusConnecting = false
-				in.iceInProgress = false
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-		{
-			name: "ICE up, peer uses relay but relay down -> partial (relay required, ICE ignored)",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = true
-				in.relayConnected = false
-				in.iceStatusConnecting = true
-			},
-			// relayOK = false (peer uses relay but it's down), iceUp = true
-			// first switch arm fails (relayOK false), relayUsedAndUp = false (relay down),
-			// falls into default: Disconnected.
-			want: guard.ConnStatusDisconnected,
-		},
-		{
-			name: "ICE down, relay up but peer does not use relay -> disconnected",
-			mutator: func(in *connStatusInputs) {
-				in.peerUsesRelay = false
-				in.relayConnected = true // not actually used since peer doesn't rely on it
-				in.iceStatusConnecting = false
-				in.iceInProgress = false
-			},
-			want: guard.ConnStatusDisconnected,
-		},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			in := base
-			tc.mutator(&in)
-			if got := evalConnStatus(in); got != tc.want {
-				t.Fatalf("evalConnStatus = %v, want %v (inputs: %+v)", got, tc.want, in)
-			}
-		})
-	}
-}

client/internal/peer/env.go

@@ -10,7 +10,7 @@ const (
 	EnvKeyNBForceRelay = "NB_FORCE_RELAY"
 )
 
-func IsForceRelayed() bool {
+func isForceRelayed() bool {
 	if runtime.GOOS == "js" {
 		return true
 	}

client/internal/peer/guard/guard.go

@@ -8,19 +8,7 @@ import (
 	log "github.com/sirupsen/logrus"
 )
 
-// ConnStatus represents the connection state as seen by the guard.
-type ConnStatus int
-
-const (
-	// ConnStatusDisconnected means neither ICE nor Relay is connected.
-	ConnStatusDisconnected ConnStatus = iota
-	// ConnStatusPartiallyConnected means Relay is connected but ICE is not.
-	ConnStatusPartiallyConnected
-	// ConnStatusConnected means all required connections are established.
-	ConnStatusConnected
-)
-
-type connStatusFunc func() ConnStatus
+type isConnectedFunc func() bool
 
 // Guard is responsible for the reconnection logic.
 // It will trigger to send an offer to the peer then has connection issues.
@@ -32,14 +20,14 @@ type connStatusFunc func() ConnStatus
 // - ICE candidate changes
 type Guard struct {
 	log                     *log.Entry
-	isConnectedOnAllWay     connStatusFunc
+	isConnectedOnAllWay     isConnectedFunc
 	timeout                 time.Duration
 	srWatcher               *SRWatcher
 	relayedConnDisconnected chan struct{}
 	iCEConnDisconnected     chan struct{}
 }
 
-func NewGuard(log *log.Entry, isConnectedFn connStatusFunc, timeout time.Duration, srWatcher *SRWatcher) *Guard {
+func NewGuard(log *log.Entry, isConnectedFn isConnectedFunc, timeout time.Duration, srWatcher *SRWatcher) *Guard {
 	return &Guard{
 		log:                     log,
 		isConnectedOnAllWay:     isConnectedFn,
@@ -69,17 +57,8 @@ func (g *Guard) SetICEConnDisconnected() {
 	}
 }
 
-// reconnectLoopWithRetry periodically checks the connection status and sends offers to re-establish connectivity.
-//
-// Behavior depends on the connection state reported by isConnectedOnAllWay:
-//   - Connected: no action, the peer is fully reachable.
-//   - Disconnected (neither ICE nor Relay): retries aggressively with exponential backoff (800ms doubling
-//     up to timeout), never gives up. This ensures rapid recovery when the peer has no connectivity at all.
-//   - PartiallyConnected (Relay up, ICE not): retries up to 3 times with exponential backoff, then switches
-//     to one attempt per hour. This limits signaling traffic when relay already provides connectivity.
-//
-// External events (relay/ICE disconnect, signal/relay reconnect, candidate changes) reset the retry
-// counter and backoff ticker, giving ICE a fresh chance after network conditions change.
+// reconnectLoopWithRetry periodically check the connection status.
+// Try to send offer while the P2P is not established or while the Relay is not connected if is it supported
 func (g *Guard) reconnectLoopWithRetry(ctx context.Context, callback func()) {
 	srReconnectedChan := g.srWatcher.NewListener()
 	defer g.srWatcher.RemoveListener(srReconnectedChan)
@@ -89,47 +68,36 @@ func (g *Guard) reconnectLoopWithRetry(ctx context.Context, callback func()) {
 
 	tickerChannel := ticker.C
 
-	iceState := &iceRetryState{log: g.log}
-	defer iceState.reset()
-
 	for {
 		select {
-		case <-tickerChannel:
-			switch g.isConnectedOnAllWay() {
-			case ConnStatusConnected:
-				// all good, nothing to do
-			case ConnStatusDisconnected:
-				callback()
-			case ConnStatusPartiallyConnected:
-				if iceState.shouldRetry() {
-					callback()
-				} else {
-					iceState.enterHourlyMode()
-					ticker.Stop()
-					tickerChannel = iceState.hourlyC()
-				}
+		case t := <-tickerChannel:
+			if t.IsZero() {
+				g.log.Infof("retry timed out, stop periodic offer sending")
+				// after backoff timeout the ticker.C will be closed. We need to a dummy channel to avoid loop
+				tickerChannel = make(<-chan time.Time)
+				continue
 			}
 
+			if !g.isConnectedOnAllWay() {
+				callback()
+			}
 		case <-g.relayedConnDisconnected:
 			g.log.Debugf("Relay connection changed, reset reconnection ticker")
 			ticker.Stop()
-			ticker = g.newReconnectTicker(ctx)
+			ticker = g.prepareExponentTicker(ctx)
 			tickerChannel = ticker.C
-			iceState.reset()
 
 		case <-g.iCEConnDisconnected:
 			g.log.Debugf("ICE connection changed, reset reconnection ticker")
 			ticker.Stop()
-			ticker = g.newReconnectTicker(ctx)
+			ticker = g.prepareExponentTicker(ctx)
 			tickerChannel = ticker.C
-			iceState.reset()
 
 		case <-srReconnectedChan:
 			g.log.Debugf("has network changes, reset reconnection ticker")
 			ticker.Stop()
-			ticker = g.newReconnectTicker(ctx)
+			ticker = g.prepareExponentTicker(ctx)
 			tickerChannel = ticker.C
-			iceState.reset()
 
 		case <-ctx.Done():
 			g.log.Debugf("context is done, stop reconnect loop")
@@ -152,7 +120,7 @@ func (g *Guard) initialTicker(ctx context.Context) *backoff.Ticker {
 	return backoff.NewTicker(bo)
 }
 
-func (g *Guard) newReconnectTicker(ctx context.Context) *backoff.Ticker {
+func (g *Guard) prepareExponentTicker(ctx context.Context) *backoff.Ticker {
 	bo := backoff.WithContext(&backoff.ExponentialBackOff{
 		InitialInterval:     800 * time.Millisecond,
 		RandomizationFactor: 0.1,

client/internal/peer/guard/ice_retry_state.go

@@ -1,61 +0,0 @@
-package guard
-
-import (
-	"time"
-
-	log "github.com/sirupsen/logrus"
-)
-
-const (
-	// maxICERetries is the maximum number of ICE offer attempts when relay is connected
-	maxICERetries = 3
-	// iceRetryInterval is the periodic retry interval after ICE retries are exhausted
-	iceRetryInterval = 1 * time.Hour
-)
-
-// iceRetryState tracks the limited ICE retry attempts when relay is already connected.
-// After maxICERetries attempts it switches to a periodic hourly retry.
-type iceRetryState struct {
-	log     *log.Entry
-	retries int
-	hourly  *time.Ticker
-}
-
-func (s *iceRetryState) reset() {
-	s.retries = 0
-	if s.hourly != nil {
-		s.hourly.Stop()
-		s.hourly = nil
-	}
-}
-
-// shouldRetry reports whether the caller should send another ICE offer on this tick.
-// Returns false when the per-cycle retry budget is exhausted and the caller must switch
-// to the hourly ticker via enterHourlyMode + hourlyC.
-func (s *iceRetryState) shouldRetry() bool {
-	if s.hourly != nil {
-		s.log.Debugf("hourly ICE retry attempt")
-		return true
-	}
-
-	s.retries++
-	if s.retries <= maxICERetries {
-		s.log.Debugf("ICE retry attempt %d/%d", s.retries, maxICERetries)
-		return true
-	}
-
-	return false
-}
-
-// enterHourlyMode starts the hourly retry ticker. Must be called after shouldRetry returns false.
-func (s *iceRetryState) enterHourlyMode() {
-	s.log.Infof("ICE retries exhausted (%d/%d), switching to hourly retry", maxICERetries, maxICERetries)
-	s.hourly = time.NewTicker(iceRetryInterval)
-}
-
-func (s *iceRetryState) hourlyC() <-chan time.Time {
-	if s.hourly == nil {
-		return nil
-	}
-	return s.hourly.C
-}

client/internal/peer/guard/ice_retry_state_test.go

@@ -1,103 +0,0 @@
-package guard
-
-import (
-	"testing"
-
-	log "github.com/sirupsen/logrus"
-)
-
-func newTestRetryState() *iceRetryState {
-	return &iceRetryState{log: log.NewEntry(log.StandardLogger())}
-}
-
-func TestICERetryState_AllowsInitialBudget(t *testing.T) {
-	s := newTestRetryState()
-
-	for i := 1; i <= maxICERetries; i++ {
-		if !s.shouldRetry() {
-			t.Fatalf("shouldRetry returned false on attempt %d, want true (budget = %d)", i, maxICERetries)
-		}
-	}
-}
-
-func TestICERetryState_ExhaustsAfterBudget(t *testing.T) {
-	s := newTestRetryState()
-
-	for i := 0; i < maxICERetries; i++ {
-		_ = s.shouldRetry()
-	}
-
-	if s.shouldRetry() {
-		t.Fatalf("shouldRetry returned true after budget exhausted, want false")
-	}
-}
-
-func TestICERetryState_HourlyCNilBeforeEnterHourlyMode(t *testing.T) {
-	s := newTestRetryState()
-
-	if s.hourlyC() != nil {
-		t.Fatalf("hourlyC returned non-nil channel before enterHourlyMode")
-	}
-}
-
-func TestICERetryState_EnterHourlyModeArmsTicker(t *testing.T) {
-	s := newTestRetryState()
-	for i := 0; i < maxICERetries+1; i++ {
-		_ = s.shouldRetry()
-	}
-
-	s.enterHourlyMode()
-	defer s.reset()
-
-	if s.hourlyC() == nil {
-		t.Fatalf("hourlyC returned nil after enterHourlyMode")
-	}
-}
-
-func TestICERetryState_ShouldRetryTrueInHourlyMode(t *testing.T) {
-	s := newTestRetryState()
-	s.enterHourlyMode()
-	defer s.reset()
-
-	if !s.shouldRetry() {
-		t.Fatalf("shouldRetry returned false in hourly mode, want true")
-	}
-
-	// Subsequent calls also return true — we keep retrying on each hourly tick.
-	if !s.shouldRetry() {
-		t.Fatalf("second shouldRetry returned false in hourly mode, want true")
-	}
-}
-
-func TestICERetryState_ResetRestoresBudget(t *testing.T) {
-	s := newTestRetryState()
-	for i := 0; i < maxICERetries+1; i++ {
-		_ = s.shouldRetry()
-	}
-	s.enterHourlyMode()
-
-	s.reset()
-
-	if s.hourlyC() != nil {
-		t.Fatalf("hourlyC returned non-nil channel after reset")
-	}
-	if s.retries != 0 {
-		t.Fatalf("retries = %d after reset, want 0", s.retries)
-	}
-
-	for i := 1; i <= maxICERetries; i++ {
-		if !s.shouldRetry() {
-			t.Fatalf("shouldRetry returned false on attempt %d after reset, want true", i)
-		}
-	}
-}
-
-func TestICERetryState_ResetIsIdempotent(t *testing.T) {
-	s := newTestRetryState()
-	s.reset()
-	s.reset() // second call must not panic or re-stop a nil ticker
-
-	if s.hourlyC() != nil {
-		t.Fatalf("hourlyC non-nil after double reset")
-	}
-}

client/internal/peer/guard/sr_watcher.go

@@ -39,7 +39,7 @@ func NewSRWatcher(signalClient chNotifier, relayManager chNotifier, iFaceDiscove
 	return srw
 }
 
-func (w *SRWatcher) Start(disableICEMonitor bool) {
+func (w *SRWatcher) Start() {
 	w.mu.Lock()
 	defer w.mu.Unlock()
 
@@ -50,10 +50,8 @@ func (w *SRWatcher) Start(disableICEMonitor bool) {
 	ctx, cancel := context.WithCancel(context.Background())
 	w.cancelIceMonitor = cancel
 
-	if !disableICEMonitor {
-		iceMonitor := NewICEMonitor(w.iFaceDiscover, w.iceConfig, GetICEMonitorPeriod())
-		go iceMonitor.Start(ctx, w.onICEChanged)
-	}
+	iceMonitor := NewICEMonitor(w.iFaceDiscover, w.iceConfig, GetICEMonitorPeriod())
+	go iceMonitor.Start(ctx, w.onICEChanged)
 	w.signalClient.SetOnReconnectedListener(w.onReconnected)
 	w.relayManager.SetOnReconnectedListener(w.onReconnected)
 

client/internal/peer/handshaker.go

@@ -4,7 +4,6 @@ import (
 	"context"
 	"errors"
 	"sync"
-	"sync/atomic"
 
 	log "github.com/sirupsen/logrus"
 
@@ -44,10 +43,6 @@ type OfferAnswer struct {
 	SessionID *ICESessionID
 }
 
-func (o *OfferAnswer) hasICECredentials() bool {
-	return o.IceCredentials.UFrag != "" && o.IceCredentials.Pwd != ""
-}
-
 type Handshaker struct {
 	mu            sync.Mutex
 	log           *log.Entry
@@ -64,10 +59,6 @@ type Handshaker struct {
 	relayListener *AsyncOfferListener
 	iceListener   func(remoteOfferAnswer *OfferAnswer)
 
-	// remoteICESupported tracks whether the remote peer includes ICE credentials in its offers/answers.
-	// When false, the local side skips ICE listener dispatch and suppresses ICE credentials in responses.
-	remoteICESupported atomic.Bool
-
 	// remoteOffersCh is a channel used to wait for remote credentials to proceed with the connection
 	remoteOffersCh chan OfferAnswer
 	// remoteAnswerCh is a channel used to wait for remote credentials answer (confirmation of our offer) to proceed with the connection
@@ -75,7 +66,7 @@ type Handshaker struct {
 }
 
 func NewHandshaker(log *log.Entry, config ConnConfig, signaler *Signaler, ice *WorkerICE, relay *WorkerRelay, metricsStages *MetricsStages) *Handshaker {
-	h := &Handshaker{
+	return &Handshaker{
 		log:            log,
 		config:         config,
 		signaler:       signaler,
@@ -85,13 +76,6 @@ func NewHandshaker(log *log.Entry, config ConnConfig, signaler *Signaler, ice *W
 		remoteOffersCh: make(chan OfferAnswer),
 		remoteAnswerCh: make(chan OfferAnswer),
 	}
-	// assume remote supports ICE until we learn otherwise from received offers
-	h.remoteICESupported.Store(ice != nil)
-	return h
-}
-
-func (h *Handshaker) RemoteICESupported() bool {
-	return h.remoteICESupported.Load()
 }
 
 func (h *Handshaker) AddRelayListener(offer func(remoteOfferAnswer *OfferAnswer)) {
@@ -106,20 +90,18 @@ func (h *Handshaker) Listen(ctx context.Context) {
 	for {
 		select {
 		case remoteOfferAnswer := <-h.remoteOffersCh:
-			h.log.Infof("received offer, running version %s, remote WireGuard listen port %d, session id: %s, remote ICE supported: %t", remoteOfferAnswer.Version, remoteOfferAnswer.WgListenPort, remoteOfferAnswer.SessionIDString(), remoteOfferAnswer.hasICECredentials())
+			h.log.Infof("received offer, running version %s, remote WireGuard listen port %d, session id: %s", remoteOfferAnswer.Version, remoteOfferAnswer.WgListenPort, remoteOfferAnswer.SessionIDString())
 
 			// Record signaling received for reconnection attempts
 			if h.metricsStages != nil {
 				h.metricsStages.RecordSignalingReceived()
 			}
 
-			h.updateRemoteICEState(&remoteOfferAnswer)
-
 			if h.relayListener != nil {
 				h.relayListener.Notify(&remoteOfferAnswer)
 			}
 
-			if h.iceListener != nil && h.RemoteICESupported() {
+			if h.iceListener != nil {
 				h.iceListener(&remoteOfferAnswer)
 			}
 
@@ -128,20 +110,18 @@ func (h *Handshaker) Listen(ctx context.Context) {
 				continue
 			}
 		case remoteOfferAnswer := <-h.remoteAnswerCh:
-			h.log.Infof("received answer, running version %s, remote WireGuard listen port %d, session id: %s, remote ICE supported: %t", remoteOfferAnswer.Version, remoteOfferAnswer.WgListenPort, remoteOfferAnswer.SessionIDString(), remoteOfferAnswer.hasICECredentials())
+			h.log.Infof("received answer, running version %s, remote WireGuard listen port %d, session id: %s", remoteOfferAnswer.Version, remoteOfferAnswer.WgListenPort, remoteOfferAnswer.SessionIDString())
 
 			// Record signaling received for reconnection attempts
 			if h.metricsStages != nil {
 				h.metricsStages.RecordSignalingReceived()
 			}
 
-			h.updateRemoteICEState(&remoteOfferAnswer)
-
 			if h.relayListener != nil {
 				h.relayListener.Notify(&remoteOfferAnswer)
 			}
 
-			if h.iceListener != nil && h.RemoteICESupported() {
+			if h.iceListener != nil {
 				h.iceListener(&remoteOfferAnswer)
 			}
 		case <-ctx.Done():
@@ -203,18 +183,15 @@ func (h *Handshaker) sendAnswer() error {
 }
 
 func (h *Handshaker) buildOfferAnswer() OfferAnswer {
+	uFrag, pwd := h.ice.GetLocalUserCredentials()
+	sid := h.ice.SessionID()
 	answer := OfferAnswer{
+		IceCredentials:  IceCredentials{uFrag, pwd},
 		WgListenPort:    h.config.LocalWgPort,
 		Version:         version.NetbirdVersion(),
 		RosenpassPubKey: h.config.RosenpassConfig.PubKey,
 		RosenpassAddr:   h.config.RosenpassConfig.Addr,
-	}
-
-	if h.ice != nil && h.RemoteICESupported() {
-		uFrag, pwd := h.ice.GetLocalUserCredentials()
-		sid := h.ice.SessionID()
-		answer.IceCredentials = IceCredentials{uFrag, pwd}
-		answer.SessionID = &sid
+		SessionID:       &sid,
 	}
 
 	if addr, err := h.relay.RelayInstanceAddress(); err == nil {
@@ -223,18 +200,3 @@ func (h *Handshaker) buildOfferAnswer() OfferAnswer {
 
 	return answer
 }
-
-func (h *Handshaker) updateRemoteICEState(offer *OfferAnswer) {
-	hasICE := offer.hasICECredentials()
-	prev := h.remoteICESupported.Swap(hasICE)
-	if prev != hasICE {
-		if hasICE {
-			h.log.Infof("remote peer started sending ICE credentials")
-		} else {
-			h.log.Infof("remote peer stopped sending ICE credentials")
-			if h.ice != nil {
-				h.ice.Close()
-			}
-		}
-	}
-}

client/internal/peer/signaler.go

@@ -46,13 +46,9 @@ func (s *Signaler) Ready() bool {
 
 // SignalOfferAnswer signals either an offer or an answer to remote peer
 func (s *Signaler) signalOfferAnswer(offerAnswer OfferAnswer, remoteKey string, bodyType sProto.Body_Type) error {
-	var sessionIDBytes []byte
-	if offerAnswer.SessionID != nil {
-		var err error
-		sessionIDBytes, err = offerAnswer.SessionID.Bytes()
-		if err != nil {
-			log.Warnf("failed to get session ID bytes: %v", err)
-		}
+	sessionIDBytes, err := offerAnswer.SessionID.Bytes()
+	if err != nil {
+		log.Warnf("failed to get session ID bytes: %v", err)
 	}
 	msg, err := signal.MarshalCredential(
 		s.wgPrivateKey,

management/internals/server/boot.go

@@ -19,7 +19,6 @@ import (
 	"google.golang.org/grpc/keepalive"
 
 	cachestore "github.com/eko/gocache/lib/v4/store"
-
 	"github.com/netbirdio/management-integrations/integrations"
 
 	"github.com/netbirdio/netbird/encryption"
@@ -35,7 +34,6 @@ import (
 	"github.com/netbirdio/netbird/management/server/store"
 	"github.com/netbirdio/netbird/management/server/telemetry"
 	mgmtProto "github.com/netbirdio/netbird/shared/management/proto"
-	"github.com/netbirdio/netbird/shared/settingoverrider"
 	"github.com/netbirdio/netbird/util/crypt"
 )
 
@@ -75,23 +73,6 @@ func (s *BaseServer) CacheStore() cachestore.StoreInterface {
 	})
 }
 
-// SettingOverrider returns a shared setting overrider backed by Redis.
-// Returns a no-op overrider if no Redis address is configured.
-func (s *BaseServer) SettingOverrider() *settingoverrider.Overrider {
-	return Create(s, func() *settingoverrider.Overrider {
-		redisAddr := nbcache.GetAddrFromEnv()
-		if redisAddr == "" {
-			return settingoverrider.NewNoop()
-		}
-
-		o, err := settingoverrider.New(context.Background(), redisAddr)
-		if err != nil {
-			log.Fatalf("failed to create setting overrider: %v", err)
-		}
-		return o
-	})
-}
-
 func (s *BaseServer) Store() store.Store {
 	return Create(s, func() store.Store {
 		store, err := store.NewStore(context.Background(), s.Config.StoreConfig.Engine, s.Config.Datadir, s.Metrics(), false)

management/internals/server/server.go

@@ -23,7 +23,6 @@ import (
 	"github.com/netbirdio/netbird/management/server/idp"
 	"github.com/netbirdio/netbird/management/server/metrics"
 	"github.com/netbirdio/netbird/management/server/store"
-	"github.com/netbirdio/netbird/shared/settingoverrider"
 	"github.com/netbirdio/netbird/util/wsproxy"
 	wsproxyserver "github.com/netbirdio/netbird/util/wsproxy/server"
 	"github.com/netbirdio/netbird/version"
@@ -124,15 +123,6 @@ func (s *BaseServer) Start(ctx context.Context) error {
 	s.PeersManager()
 	s.GeoLocationManager()
 
-	s.SettingOverrider().Poll(settingoverrider.DefaultInterval, "managementLogLevel", func(value string) error {
-		level, err := log.ParseLevel(value)
-		if err != nil {
-			return fmt.Errorf("parsing log level %q: %w", value, err)
-		}
-		log.SetLevel(level)
-		return nil
-	})
-
 	err := s.Metrics().Expose(srvCtx, s.mgmtMetricsPort, "/metrics")
 	if err != nil {
 		return fmt.Errorf("failed to expose metrics: %v", err)
@@ -245,7 +235,6 @@ func (s *BaseServer) Stop() error {
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 	defer cancel()
 
-	_ = s.SettingOverrider().Close()
 	s.IntegratedValidator().Stop(ctx)
 	if s.GeoLocationManager() != nil {
 		_ = s.GeoLocationManager().Stop()

management/server/account_test.go

@@ -2311,6 +2311,29 @@ func TestAccount_GetExpiredPeers(t *testing.T) {
 	}
 }
 
+func TestGetExpiredPeers_SkipsAlreadyExpired(t *testing.T) {
+	ctx := context.Background()
+
+	testStore, cleanUp, err := store.NewTestStoreFromSQL(ctx, "testdata/store_with_expired_peers.sql", t.TempDir())
+	t.Cleanup(cleanUp)
+	require.NoError(t, err)
+
+	accountID := "bf1c8084-ba50-4ce7-9439-34653001fc3b"
+
+	// Verify the already-expired peer is excluded at the store level
+	peers, err := testStore.GetAccountPeersWithExpiration(ctx, store.LockingStrengthNone, accountID)
+	require.NoError(t, err)
+
+	for _, peer := range peers {
+		assert.NotEqual(t, "cg05lnblo1hkg2j514p0", peer.ID, "already expired peer should be excluded by the store query")
+		assert.False(t, peer.Status.LoginExpired, "returned peers should not already be marked as login expired")
+	}
+
+	// Only the non-expired peer with expiration enabled should be returned
+	require.Len(t, peers, 1)
+	assert.Equal(t, "notexpired01", peers[0].ID)
+}
+
 func TestAccount_GetInactivePeers(t *testing.T) {
 	type test struct {
 		name          string
@@ -3230,6 +3253,13 @@ func setupNetworkMapTest(t *testing.T) (*DefaultAccountManager, *update_channel.
 	return manager, updateManager, account, peer1, peer2, peer3
 }
 
+// peerUpdateTimeout bounds how long peerShouldReceiveUpdate and its outer
+// wrappers wait for an expected update message. Sized for slow CI runners
+// (MySQL, FreeBSD, loaded sqlite) where the channel publish can take
+// seconds. Only runs down on failure; passing tests return immediately
+// when the channel delivers.
+const peerUpdateTimeout = 5 * time.Second
+
 func peerShouldNotReceiveUpdate(t *testing.T, updateMessage <-chan *network_map.UpdateMessage) {
 	t.Helper()
 	select {
@@ -3248,7 +3278,7 @@ func peerShouldReceiveUpdate(t *testing.T, updateMessage <-chan *network_map.Upd
 		if msg == nil {
 			t.Errorf("Received nil update message, expected valid message")
 		}
-	case <-time.After(500 * time.Millisecond):
+	case <-time.After(peerUpdateTimeout):
 		t.Error("Timed out waiting for update message")
 	}
 }

management/server/dns_test.go

@@ -458,7 +458,7 @@ func TestDNSAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -478,7 +478,7 @@ func TestDNSAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -518,7 +518,7 @@ func TestDNSAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/group_test.go

@@ -620,7 +620,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -638,7 +638,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -656,7 +656,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -689,7 +689,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -730,7 +730,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -757,7 +757,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -804,7 +804,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/management_proto_test.go

@@ -267,8 +267,8 @@ func Test_SyncProtocol(t *testing.T) {
 	}
 
 	// expired peers come separately.
-	if len(networkMap.GetOfflinePeers()) != 1 {
-		t.Fatal("expecting SyncResponse to have NetworkMap with 1 offline peer")
+	if len(networkMap.GetOfflinePeers()) != 2 {
+		t.Fatal("expecting SyncResponse to have NetworkMap with 2 offline peer")
 	}
 
 	expiredPeerPubKey := "RlSy2vzoG2HyMBTUImXOiVhCBiiBa5qD5xzMxkiFDW4="

management/server/nameserver_test.go

@@ -1087,7 +1087,7 @@ func TestNameServerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1105,7 +1105,7 @@ func TestNameServerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/peer.go

@@ -1405,6 +1405,10 @@ func (am *DefaultAccountManager) getExpiredPeers(ctx context.Context, accountID
 
 	var peers []*nbpeer.Peer
 	for _, peer := range peersWithExpiry {
+		if peer.Status.LoginExpired {
+			continue
+		}
+
 		expired, _ := peer.LoginExpired(settings.PeerLoginExpiration)
 		if expired {
 			peers = append(peers, peer)

management/server/peer_test.go

@@ -1907,7 +1907,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1929,7 +1929,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1994,7 +1994,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2012,7 +2012,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2058,7 +2058,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2076,7 +2076,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2113,7 +2113,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2131,7 +2131,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/policy_test.go

@@ -1231,7 +1231,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1263,7 +1263,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1294,7 +1294,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1314,7 +1314,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1355,7 +1355,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1373,7 +1373,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 
@@ -1393,7 +1393,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/posture_checks_test.go

@@ -244,7 +244,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -273,7 +273,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -292,7 +292,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -395,7 +395,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -438,7 +438,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/route_test.go

@@ -2070,7 +2070,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 
@@ -2107,7 +2107,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2127,7 +2127,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2145,7 +2145,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2185,7 +2185,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2225,7 +2225,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/store/sql_store.go

@@ -3310,7 +3310,7 @@ func (s *SqlStore) GetAccountPeersWithExpiration(ctx context.Context, lockStreng
 
 	var peers []*nbpeer.Peer
 	result := tx.
-		Where("login_expiration_enabled = ? AND user_id IS NOT NULL AND user_id != ''", true).
+		Where("login_expiration_enabled = ? AND peer_status_login_expired != ? AND user_id IS NOT NULL AND user_id != ''", true, true).
 		Find(&peers, accountIDCondition, accountID)
 	if err := result.Error; err != nil {
 		log.WithContext(ctx).Errorf("failed to get peers with expiration from the store: %s", result.Error)

management/server/store/sql_store_test.go

@@ -2729,7 +2729,7 @@ func TestSqlStore_GetAccountPeers(t *testing.T) {
 		{
 			name:          "should retrieve peers for an existing account ID",
 			accountID:     "bf1c8084-ba50-4ce7-9439-34653001fc3b",
-			expectedCount: 4,
+			expectedCount: 5,
 		},
 		{
 			name:          "should return no peers for a non-existing account ID",
@@ -2751,7 +2751,7 @@ func TestSqlStore_GetAccountPeers(t *testing.T) {
 			name:          "should filter peers by partial name",
 			accountID:     "bf1c8084-ba50-4ce7-9439-34653001fc3b",
 			nameFilter:    "host",
-			expectedCount: 3,
+			expectedCount: 4,
 		},
 		{
 			name:          "should filter peers by ip",
@@ -2777,14 +2777,16 @@ func TestSqlStore_GetAccountPeersWithExpiration(t *testing.T) {
 	require.NoError(t, err)
 
 	tests := []struct {
-		name          string
-		accountID     string
-		expectedCount int
+		name            string
+		accountID       string
+		expectedCount   int
+		expectedPeerIDs []string
 	}{
 		{
-			name:          "should retrieve peers with expiration for an existing account ID",
-			accountID:     "bf1c8084-ba50-4ce7-9439-34653001fc3b",
-			expectedCount: 1,
+			name:            "should retrieve only non-expired peers with expiration enabled",
+			accountID:       "bf1c8084-ba50-4ce7-9439-34653001fc3b",
+			expectedCount:   1,
+			expectedPeerIDs: []string{"notexpired01"},
 		},
 		{
 			name:          "should return no peers with expiration for a non-existing account ID",
@@ -2803,10 +2805,30 @@ func TestSqlStore_GetAccountPeersWithExpiration(t *testing.T) {
 			peers, err := store.GetAccountPeersWithExpiration(context.Background(), LockingStrengthNone, tt.accountID)
 			require.NoError(t, err)
 			require.Len(t, peers, tt.expectedCount)
+			for i, peer := range peers {
+				assert.Equal(t, tt.expectedPeerIDs[i], peer.ID)
+			}
 		})
 	}
 }
 
+func TestSqlStore_GetAccountPeersWithExpiration_ExcludesAlreadyExpired(t *testing.T) {
+	store, cleanup, err := NewTestStoreFromSQL(context.Background(), "../testdata/store_with_expired_peers.sql", t.TempDir())
+	t.Cleanup(cleanup)
+	require.NoError(t, err)
+
+	accountID := "bf1c8084-ba50-4ce7-9439-34653001fc3b"
+
+	peers, err := store.GetAccountPeersWithExpiration(context.Background(), LockingStrengthNone, accountID)
+	require.NoError(t, err)
+
+	// Verify the already-expired peer (cg05lnblo1hkg2j514p0) is not returned
+	for _, peer := range peers {
+		assert.NotEqual(t, "cg05lnblo1hkg2j514p0", peer.ID, "already expired peer should not be returned")
+		assert.False(t, peer.Status.LoginExpired, "returned peers should not have LoginExpired set")
+	}
+}
+
 func TestSqlStore_GetAccountPeersWithInactivity(t *testing.T) {
 	store, cleanup, err := NewTestStoreFromSQL(context.Background(), "../testdata/store_with_expired_peers.sql", t.TempDir())
 	t.Cleanup(cleanup)
@@ -2887,7 +2909,7 @@ func TestSqlStore_GetUserPeers(t *testing.T) {
 			name:          "should retrieve peers for another valid account ID and user ID",
 			accountID:     "bf1c8084-ba50-4ce7-9439-34653001fc3b",
 			userID:        "edafee4e-63fb-11ec-90d6-0242ac120003",
-			expectedCount: 2,
+			expectedCount: 3,
 		},
 		{
 			name:          "should return no peers for existing account ID with empty user ID",

management/server/testdata/store_with_expired_peers.sql

@@ -31,6 +31,7 @@ INSERT INTO peers VALUES('cfvprsrlo1hqoo49ohog','bf1c8084-ba50-4ce7-9439-3465300
 INSERT INTO peers VALUES('cg05lnblo1hkg2j514p0','bf1c8084-ba50-4ce7-9439-34653001fc3b','RlSy2vzoG2HyMBTUImXOiVhCBiiBa5qD5xzMxkiFDW4=','','"100.64.39.54"','expiredhost','linux','Linux','22.04','x86_64','Ubuntu','','development','','',NULL,'','','','{"Cloud":"","Platform":""}',NULL,'expiredhost','expiredhost','2023-03-02 09:19:57.276717255+01:00',0,1,0,'edafee4e-63fb-11ec-90d6-0242ac120003','ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMbK5ZXJsGOOWoBT4OmkPtgdPZe2Q7bDuS/zjn2CZxhK',0,1,0,'2023-03-02 09:14:21.791679181+01:00','2024-10-02 17:00:32.527947+02:00',0,'""','','',0);
 INSERT INTO peers VALUES('cg3161rlo1hs9cq94gdg','bf1c8084-ba50-4ce7-9439-34653001fc3b','mVABSKj28gv+JRsf7e0NEGKgSOGTfU/nPB2cpuG56HU=','','"100.64.117.96"','testhost','linux','Linux','22.04','x86_64','Ubuntu','','development','','',NULL,'','','','{"Cloud":"","Platform":""}',NULL,'testhost','testhost','2023-03-06 18:21:27.252010027+01:00',0,0,0,'edafee4e-63fb-11ec-90d6-0242ac120003','ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINWvvUkFFcrj48CWTkNUb/do/n52i1L5dH4DhGu+4ZuM',0,0,0,'2023-03-07 09:02:47.442857106+01:00','2024-10-02 17:00:32.527947+02:00',0,'""','','',0);
 INSERT INTO peers VALUES('csrnkiq7qv9d8aitqd50','bf1c8084-ba50-4ce7-9439-34653001fc3b','nVABSKj28gv+JRsf7e0NEGKgSOGTfU/nPB2cpuG56HX=','','"100.64.117.97"','testhost','linux','Linux','22.04','x86_64','Ubuntu','','development','','',NULL,'','','','{"Cloud":"","Platform":""}',NULL,'testhost','testhost-1','2023-03-06 18:21:27.252010027+01:00',0,0,0,'f4f6d672-63fb-11ec-90d6-0242ac120003','ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINWvvUkFFcrj48CWTkNUb/do/n52i1L5dH4DhGu+4ZuM',0,0,1,'2023-03-07 09:02:47.442857106+01:00','2024-10-02 17:00:32.527947+02:00',0,'""','','',0);
+INSERT INTO peers VALUES('notexpired01','bf1c8084-ba50-4ce7-9439-34653001fc3b','oVABSKj28gv+JRsf7e0NEGKgSOGTfU/nPB2cpuG56HY=','','"100.64.117.98"','activehost','linux','Linux','22.04','x86_64','Ubuntu','','development','','',NULL,'','','','{"Cloud":"","Platform":""}',NULL,'activehost','activehost','2023-03-06 18:21:27.252010027+01:00',0,0,0,'edafee4e-63fb-11ec-90d6-0242ac120003','ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAINWvvUkFFcrj48CWTkNUb/do/n52i1L5dH4DhGu+4ZuM',0,1,0,'2023-03-07 09:02:47.442857106+01:00','2024-10-02 17:00:32.527947+02:00',0,'""','','',0);
 INSERT INTO users VALUES('f4f6d672-63fb-11ec-90d6-0242ac120003','bf1c8084-ba50-4ce7-9439-34653001fc3b','user',0,0,'','[]',0,NULL,'2024-10-02 17:00:32.528196+02:00','api',0,'');
 INSERT INTO users VALUES('edafee4e-63fb-11ec-90d6-0242ac120003','bf1c8084-ba50-4ce7-9439-34653001fc3b','admin',0,0,'','[]',0,NULL,'2024-10-02 17:00:32.528196+02:00','api',0,'');
 INSERT INTO installations VALUES(1,'');

management/server/user_test.go

@@ -1586,7 +1586,7 @@ func TestUserAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1609,7 +1609,7 @@ func TestUserAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(time.Second):
+		case <-time.After(peerUpdateTimeout):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

proxy/internal/auth/middleware.go

@@ -433,6 +433,7 @@ func setSessionCookie(w http.ResponseWriter, token string, expiration time.Durat
 	http.SetCookie(w, &http.Cookie{
 		Name:     auth.SessionCookieName,
 		Value:    token,
+		Path:     "/",
 		HttpOnly: true,
 		Secure:   true,
 		SameSite: http.SameSiteLaxMode,

proxy/internal/auth/middleware_test.go

@@ -391,6 +391,15 @@ func TestProtect_SchemeAuthRedirectsWithCookie(t *testing.T) {
 	assert.Equal(t, http.SameSiteLaxMode, sessionCookie.SameSite)
 }
 
+func TestSetSessionCookieHasRootPath(t *testing.T) {
+	w := httptest.NewRecorder()
+	setSessionCookie(w, "test-token", time.Hour)
+
+	cookies := w.Result().Cookies()
+	require.Len(t, cookies, 1)
+	assert.Equal(t, "/", cookies[0].Path, "session cookie must be scoped to root so it applies to all paths")
+}
+
 func TestProtect_FailedAuthDoesNotSetCookie(t *testing.T) {
 	mw := NewMiddleware(log.StandardLogger(), nil, nil)
 	kp := generateTestKeyPair(t)

shared/settingoverrider/overrider.go

@@ -1,120 +0,0 @@
-package settingoverrider
-
-import (
-	"context"
-	"errors"
-	"fmt"
-	"time"
-
-	"github.com/redis/go-redis/v9"
-	log "github.com/sirupsen/logrus"
-)
-
-const (
-	DefaultInterval = 5 * time.Minute
-)
-
-// ApplyFunc is called with the raw Redis string value whenever it changes.
-// The function is responsible for parsing and applying the value.
-// Return an error to log a warning without stopping the polling loop.
-type ApplyFunc func(value string) error
-
-// Overrider holds a shared Redis connection and allows registering
-// individual settings that are polled independently.
-type Overrider struct {
-	client *redis.Client
-	cancel context.CancelFunc
-	ctx    context.Context
-	noop   bool
-}
-
-// New creates an Overrider by connecting to Redis at the given address.
-// The address should follow the Redis URL format (e.g. "redis://localhost:6379").
-func New(ctx context.Context, redisAddr string) (*Overrider, error) {
-	if redisAddr == "" {
-		return nil, fmt.Errorf("redis address is empty")
-	}
-
-	options, err := redis.ParseURL(redisAddr)
-	if err != nil {
-		return nil, fmt.Errorf("parsing redis address: %w", err)
-	}
-
-	client := redis.NewClient(options)
-
-	pingCtx, cancel := context.WithTimeout(ctx, 2*time.Second)
-	defer cancel()
-
-	if _, err := client.Ping(pingCtx).Result(); err != nil {
-		_ = client.Close()
-		return nil, fmt.Errorf("connecting to redis: %w", err)
-	}
-
-	oCtx, oCancel := context.WithCancel(ctx)
-
-	return &Overrider{client: client, cancel: oCancel, ctx: oCtx}, nil
-}
-
-// NewNoop returns an Overrider that does nothing.
-// Poll calls are silently ignored and Close is a no-op.
-func NewNoop() *Overrider {
-	return &Overrider{noop: true}
-}
-
-// Close stops all polling goroutines and closes the underlying Redis client.
-func (o *Overrider) Close() error {
-	if o.noop {
-		return nil
-	}
-	o.cancel()
-	return o.client.Close()
-}
-
-// Poll starts a background goroutine that polls a single Redis key at the given interval
-// and calls apply whenever the value changes. The goroutine stops when the Overrider is closed.
-func (o *Overrider) Poll(interval time.Duration, redisKey string, apply ApplyFunc) {
-	if o.noop {
-		return
-	}
-
-	go func() {
-		ticker := time.NewTicker(interval)
-		defer ticker.Stop()
-
-		var lastSeen *string
-
-		for {
-			select {
-			case <-o.ctx.Done():
-				log.WithContext(o.ctx).Infof("Stopping settings overrider for key %q", redisKey)
-				return
-			case <-ticker.C:
-				getCtx, cancel := context.WithTimeout(o.ctx, 5*time.Second)
-				val, err := o.client.Get(getCtx, redisKey).Result()
-				cancel()
-
-				if errors.Is(err, redis.Nil) || val == "" {
-					continue
-				}
-				if err != nil {
-					if o.ctx.Err() != nil {
-						return
-					}
-					log.WithContext(o.ctx).Errorf("Unable to get setting %q from Redis: %v", redisKey, err)
-					continue
-				}
-
-				if lastSeen != nil && *lastSeen == val {
-					continue
-				}
-
-				if err := apply(val); err != nil {
-					log.WithContext(o.ctx).Warnf("Failed to apply setting %q with value %q: %v", redisKey, val, err)
-					continue
-				}
-
-				lastSeen = &val
-			}
-		}
-	}()
-}

shared/settingoverrider/overrider_test.go

@@ -1,111 +0,0 @@
-package settingoverrider
-
-import (
-	"context"
-	"sync/atomic"
-	"testing"
-	"time"
-
-	"github.com/redis/go-redis/v9"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-	"github.com/testcontainers/testcontainers-go"
-	testcontainersredis "github.com/testcontainers/testcontainers-go/modules/redis"
-	"github.com/testcontainers/testcontainers-go/wait"
-)
-
-func TestPoll_AppliesSettingFromRedis(t *testing.T) {
-	o, client := setupOverrider(t)
-
-	key := "test-setting-key"
-	require.NoError(t, client.Set(context.Background(), key, "hello", 0).Err())
-
-	var applied atomic.Value
-
-	o.Poll(100*time.Millisecond, key, func(value string) error {
-		applied.Store(value)
-		return nil
-	})
-
-	assert.Eventually(t, func() bool {
-		v := applied.Load()
-		return v != nil && v.(string) == "hello"
-	}, 5*time.Second, 50*time.Millisecond)
-}
-
-func TestPoll_IndependentSettings(t *testing.T) {
-	o, client := setupOverrider(t)
-
-	require.NoError(t, client.Set(context.Background(), "key-a", "val-a", 0).Err())
-	require.NoError(t, client.Set(context.Background(), "key-b", "val-b", 0).Err())
-
-	var gotA, gotB atomic.Value
-
-	o.Poll(100*time.Millisecond, "key-a", func(v string) error { gotA.Store(v); return nil })
-	o.Poll(100*time.Millisecond, "key-b", func(v string) error { gotB.Store(v); return nil })
-
-	assert.Eventually(t, func() bool {
-		a, b := gotA.Load(), gotB.Load()
-		return a != nil && a.(string) == "val-a" && b != nil && b.(string) == "val-b"
-	}, 5*time.Second, 50*time.Millisecond)
-}
-
-func TestPoll_SkipsDuplicateValues(t *testing.T) {
-	o, client := setupOverrider(t)
-
-	key := "test-dedup"
-	require.NoError(t, client.Set(context.Background(), key, "same", 0).Err())
-
-	var count atomic.Int32
-
-	o.Poll(100*time.Millisecond, key, func(string) error {
-		count.Add(1)
-		return nil
-	})
-
-	// wait for a few ticks
-	time.Sleep(600 * time.Millisecond)
-	assert.Equal(t, int32(1), count.Load(), "Apply should be called only once for unchanged value")
-}
-
-func setupOverrider(t *testing.T) (*Overrider, *redis.Client) {
-	t.Helper()
-
-	ctx := context.Background()
-	redisContainer, err := testcontainersredis.RunContainer(ctx,
-		testcontainers.WithImage("redis:7"),
-		testcontainers.WithWaitStrategy(
-			wait.ForListeningPort("6379/tcp"),
-		),
-	)
-	require.NoError(t, err, "Failed to create redis test container")
-
-	t.Cleanup(func() {
-		if err := redisContainer.Terminate(ctx); err != nil {
-			t.Logf("failed to terminate redis container: %s", err)
-		}
-	})
-
-	redisURL, err := redisContainer.ConnectionString(ctx)
-	require.NoError(t, err)
-
-	o, err := New(ctx, redisURL)
-	require.NoError(t, err)
-	t.Cleanup(func() {
-		if err := o.Close(); err != nil {
-			t.Logf("failed to close overrider: %s", err)
-		}
-	})
-
-	// separate client for test setup (setting keys)
-	options, err := redis.ParseURL(redisURL)
-	require.NoError(t, err)
-	client := redis.NewClient(options)
-	t.Cleanup(func() {
-		if err := client.Close(); err != nil {
-			t.Logf("failed to close redis client: %s", err)
-		}
-	})
-
-	return o, client
-}

signal/cmd/run.go

@@ -18,9 +18,7 @@ import (
 	"golang.org/x/net/http2"
 	"golang.org/x/net/http2/h2c"
 
-	"github.com/netbirdio/netbird/management/server/cache"
 	"github.com/netbirdio/netbird/shared/metrics"
-	"github.com/netbirdio/netbird/shared/settingoverrider"
 
 	"github.com/netbirdio/netbird/encryption"
 	"github.com/netbirdio/netbird/shared/signal/proto"
@@ -116,24 +114,7 @@ var (
 				}
 			}()
 
-			overrider := settingoverrider.NewNoop()
-			if redisAddr := cache.GetAddrFromEnv(); redisAddr != "" {
-				overrider, err = settingoverrider.New(cmd.Context(), redisAddr)
-				if err != nil {
-					return fmt.Errorf("failed to create setting overrider: %w", err)
-				}
-				defer func() { _ = overrider.Close() }()
-			}
-			overrider.Poll(settingoverrider.DefaultInterval, "signalLogLevel", func(value string) error {
-				level, err := log.ParseLevel(value)
-				if err != nil {
-					return fmt.Errorf("parsing log level %q: %w", value, err)
-				}
-				log.SetLevel(level)
-				return nil
-			})
-
-			srv, err := server.NewServer(cmd.Context(), metricsServer.Meter, overrider)
+			srv, err := server.NewServer(cmd.Context(), metricsServer.Meter)
 			if err != nil {
 				return fmt.Errorf("creating signal server: %v", err)
 			}

signal/server/send_tracker.go

@@ -1,135 +0,0 @@
-package server
-
-import (
-	"context"
-	"math"
-	"os"
-	"strconv"
-	"sync"
-	"sync/atomic"
-	"time"
-
-	log "github.com/sirupsen/logrus"
-)
-
-const (
-	defaultSendRateLogInterval = 5 * time.Minute
-	defaultSendRateTopPercent  = 0.95
-	envSendRateLogInterval     = "NB_SIGNAL_SEND_RATE_LOG_INTERVAL"
-	envSendRateTopPercent      = "NB_SIGNAL_SEND_RATE_LOG_TOP_PERCENT"
-)
-
-// sendRateTracker tracks per-key message counts and logs the busiest peers periodically.
-type sendRateTracker struct {
-	mu     sync.Mutex
-	counts map[string]int64
-
-	// atomic so they can be updated by the setting overrider without locking
-	intervalNs atomic.Int64
-	// topPercent stored as float64 bits for atomic access
-	topPercentBits atomic.Uint64
-}
-
-func newSendRateTracker() *sendRateTracker {
-	interval := defaultSendRateLogInterval
-	if v := os.Getenv(envSendRateLogInterval); v != "" {
-		if parsed, err := time.ParseDuration(v); err == nil && parsed > 0 {
-			interval = parsed
-		}
-	}
-
-	topPercent := defaultSendRateTopPercent
-	if v := os.Getenv(envSendRateTopPercent); v != "" {
-		if parsed, err := strconv.ParseFloat(v, 64); err == nil && parsed > 0 && parsed <= 1 {
-			topPercent = parsed
-		}
-	}
-
-	log.Debugf("send rate tracker: interval=%s, top_percent=%.2f", interval, topPercent)
-
-	t := &sendRateTracker{
-		counts: make(map[string]int64),
-	}
-	t.intervalNs.Store(int64(interval))
-	t.topPercentBits.Store(math.Float64bits(topPercent))
-	return t
-}
-
-func (t *sendRateTracker) getInterval() time.Duration {
-	return time.Duration(t.intervalNs.Load())
-}
-
-func (t *sendRateTracker) setInterval(d time.Duration) {
-	t.intervalNs.Store(int64(d))
-}
-
-func (t *sendRateTracker) getTopPercent() float64 {
-	return math.Float64frombits(t.topPercentBits.Load())
-}
-
-func (t *sendRateTracker) setTopPercent(p float64) {
-	t.topPercentBits.Store(math.Float64bits(p))
-}
-
-func (t *sendRateTracker) increment(key string) {
-	t.mu.Lock()
-	t.counts[key]++
-	t.mu.Unlock()
-}
-
-// resetAndSnapshot atomically returns current counts and resets the tracker.
-func (t *sendRateTracker) resetAndSnapshot() map[string]int64 {
-	t.mu.Lock()
-	snap := t.counts
-	t.counts = make(map[string]int64, len(snap))
-	t.mu.Unlock()
-	return snap
-}
-
-// logSendRates periodically logs peers in the top percentile of the busiest peer.
-func (t *sendRateTracker) logSendRates(ctx context.Context) {
-	currentInterval := t.getInterval()
-	ticker := time.NewTicker(currentInterval)
-	defer ticker.Stop()
-
-	for {
-		select {
-		case <-ctx.Done():
-			return
-		case <-ticker.C:
-			if newInterval := t.getInterval(); newInterval != currentInterval {
-				currentInterval = newInterval
-				ticker.Reset(currentInterval)
-			}
-
-			snap := t.resetAndSnapshot()
-			if len(snap) == 0 {
-				continue
-			}
-
-			var maxCount int64
-			for _, count := range snap {
-				if count > maxCount {
-					maxCount = count
-				}
-			}
-
-			topPercent := t.getTopPercent()
-			threshold := int64(float64(maxCount) * topPercent)
-			intervalMin := currentInterval.Minutes()
-
-			log.Debugf("send rate stats: %d unique peers in last %.0fs, max rate %.1f msg/min",
-				len(snap), currentInterval.Seconds(), float64(maxCount)/intervalMin)
-			logged := 0
-			for key, count := range snap {
-				if count >= threshold {
-					log.Debugf("peer [%s] %.1f msg/min", key, float64(count)/intervalMin)
-					logged++
-					if logged >= 100 {
-						break
-					}
-				}
-			}
-		}
-	}
-}

signal/server/send_tracker_test.go

@@ -1,56 +0,0 @@
-package server
-
-import (
-	"sync"
-	"testing"
-)
-
-func TestSendRateTracker_Increment(t *testing.T) {
-	tracker := newSendRateTracker()
-
-	tracker.increment("peer-a")
-	tracker.increment("peer-a")
-	tracker.increment("peer-b")
-
-	snap := tracker.resetAndSnapshot()
-	if snap["peer-a"] != 2 {
-		t.Errorf("expected peer-a count 2, got %d", snap["peer-a"])
-	}
-	if snap["peer-b"] != 1 {
-		t.Errorf("expected peer-b count 1, got %d", snap["peer-b"])
-	}
-}
-
-func TestSendRateTracker_ResetAndSnapshot_Resets(t *testing.T) {
-	tracker := newSendRateTracker()
-	tracker.increment("peer-a")
-
-	snap1 := tracker.resetAndSnapshot()
-	if snap1["peer-a"] != 1 {
-		t.Fatalf("expected 1, got %d", snap1["peer-a"])
-	}
-
-	snap2 := tracker.resetAndSnapshot()
-	if len(snap2) != 0 {
-		t.Errorf("expected empty snapshot after reset, got %v", snap2)
-	}
-}
-
-func TestSendRateTracker_ConcurrentIncrement(t *testing.T) {
-	tracker := newSendRateTracker()
-
-	var wg sync.WaitGroup
-	for i := 0; i < 100; i++ {
-		wg.Add(1)
-		go func() {
-			defer wg.Done()
-			tracker.increment("peer-x")
-		}()
-	}
-	wg.Wait()
-
-	snap := tracker.resetAndSnapshot()
-	if snap["peer-x"] != 100 {
-		t.Errorf("expected 100, got %d", snap["peer-x"])
-	}
-}

signal/server/signal.go

@@ -5,7 +5,6 @@ import (
 	"errors"
 	"fmt"
 	"os"
-	"strconv"
 	"time"
 
 	log "github.com/sirupsen/logrus"
@@ -18,7 +17,6 @@ import (
 
 	"github.com/netbirdio/signal-dispatcher/dispatcher"
 
-	"github.com/netbirdio/netbird/shared/settingoverrider"
 	"github.com/netbirdio/netbird/shared/signal/proto"
 	"github.com/netbirdio/netbird/signal/metrics"
 	"github.com/netbirdio/netbird/signal/peer"
@@ -61,12 +59,10 @@ type Server struct {
 	successHeader metadata.MD
 
 	sendTimeout time.Duration
-
-	sendTracker *sendRateTracker
 }
 
 // NewServer creates a new Signal server
-func NewServer(ctx context.Context, meter metric.Meter, overrider *settingoverrider.Overrider, metricsPrefix ...string) (*Server, error) {
+func NewServer(ctx context.Context, meter metric.Meter, metricsPrefix ...string) (*Server, error) {
 	appMetrics, err := metrics.NewAppMetrics(meter, metricsPrefix...)
 	if err != nil {
 		return nil, fmt.Errorf("creating app metrics: %v", err)
@@ -84,36 +80,14 @@ func NewServer(ctx context.Context, meter metric.Meter, overrider *settingoverri
 		sTimeout = parsed
 	}
 
-	tracker := newSendRateTracker()
-
 	s := &Server{
 		dispatcher:    d,
 		registry:      peer.NewRegistry(appMetrics),
 		metrics:       appMetrics,
 		successHeader: metadata.Pairs(proto.HeaderRegistered, "1"),
 		sendTimeout:   sTimeout,
-		sendTracker:   tracker,
 	}
 
-	overrider.Poll(settingoverrider.DefaultInterval, "signalSendRateLogInterval", func(value string) error {
-		parsed, err := time.ParseDuration(value)
-		if err != nil || parsed <= 0 {
-			return fmt.Errorf("invalid send rate log interval %q: %w", value, err)
-		}
-		tracker.setInterval(parsed)
-		return nil
-	})
-	overrider.Poll(settingoverrider.DefaultInterval, "signalSendRateTopPercent", func(value string) error {
-		parsed, err := strconv.ParseFloat(value, 64)
-		if err != nil || parsed <= 0 || parsed > 1 {
-			return fmt.Errorf("invalid send rate top percent %q: %w", value, err)
-		}
-		tracker.setTopPercent(parsed)
-		return nil
-	})
-
-	go tracker.logSendRates(ctx)
-
 	return s, nil
 }
 
@@ -121,8 +95,6 @@ func NewServer(ctx context.Context, meter metric.Meter, overrider *settingoverri
 func (s *Server) Send(ctx context.Context, msg *proto.EncryptedMessage) (*proto.EncryptedMessage, error) {
 	log.Tracef("received a new message to send from peer [%s] to peer [%s]", msg.Key, msg.RemoteKey)
 
-	s.sendTracker.increment(msg.Key)
-
 	if _, found := s.registry.Get(msg.RemoteKey); found {
 		s.forwardMessageToPeer(ctx, msg)
 		return &proto.EncryptedMessage{}, nil