[android] add SSHClient gomobile binding for in-app terminal

Exposes SSHClient + SSHTerminalListener to the Android app. Connect()
auto-detects the server type via banner inspection and selects the auth
path: NetBird-SSH with JWT triggers the device-code OAuth flow via the
existing URLOpener; NetBird-SSH without JWT uses the NetBird private
key; regular SSH falls back to NetBird key then optional password. The
client dials through the running tunnel using a plain net.Dialer and
relies on the gomobile-bound listener for streaming PTY output back to
Java for rendering in an xterm.js WebView.

client/android/ssh_client.go

@@ -0,0 +1,434 @@
+//go:build android
+
+package android
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"net"
+	"strconv"
+	"sync"
+	"time"
+
+	log "github.com/sirupsen/logrus"
+	gossh "golang.org/x/crypto/ssh"
+
+	"github.com/netbirdio/netbird/client/internal"
+	"github.com/netbirdio/netbird/client/internal/auth"
+	"github.com/netbirdio/netbird/client/internal/profilemanager"
+	nbssh "github.com/netbirdio/netbird/client/ssh"
+	"github.com/netbirdio/netbird/client/ssh/detection"
+)
+
+const (
+	sshDialTimeout      = 30 * time.Second
+	sshDetectionTimeout = 5 * time.Second
+)
+
+// SSHTerminalListener receives SSH session events. It is implemented in Java.
+//
+// All callbacks are invoked from goroutines and may run concurrently with each
+// other; the implementation must be safe to call from any thread.
+type SSHTerminalListener interface {
+	OnConnected()
+	OnData(data []byte)
+	OnClose(reason string)
+	OnError(message string)
+}
+
+// SSHClient is a NetBird-aware SSH client exposed to Java via gomobile.
+//
+// It dials through the running NetBird tunnel and runs a standard SSH session
+// on top with PTY enabled. Host-key verification uses the NetBird-provided
+// peer SSH host keys, identical to the desktop client.
+type SSHClient struct {
+	nb        *Client
+	mu        sync.Mutex
+	listener  SSHTerminalListener
+	urlOpener URLOpener
+
+	sshClient *gossh.Client
+	session   *gossh.Session
+	stdin     io.WriteCloser
+	closed    bool
+}
+
+// NewSSHClient creates a new SSH client bound to the running NetBird Client.
+func NewSSHClient(c *Client) *SSHClient {
+	return &SSHClient{nb: c}
+}
+
+// SetListener registers the Java listener. Must be called before Connect to
+// receive any events.
+func (s *SSHClient) SetListener(l SSHTerminalListener) {
+	s.mu.Lock()
+	s.listener = l
+	s.mu.Unlock()
+}
+
+// SetURLOpener registers the Java URL opener used to display the device-code
+// authorization page in a Custom Tabs window when the target peer requires
+// JWT authentication. Must be set before Connect to be effective.
+func (s *SSHClient) SetURLOpener(opener URLOpener) {
+	s.mu.Lock()
+	s.urlOpener = opener
+	s.mu.Unlock()
+}
+
+// Connect dials the SSH server through the NetBird tunnel and performs the
+// SSH handshake. It auto-detects the server type via SSH banner inspection
+// and selects the appropriate authentication path:
+//
+//   - NetBird-SSH server requiring JWT: launches the OAuth 2.0 device-code
+//     flow, opens the verification URL through the registered URLOpener, and
+//     uses the resulting token as the SSH password. Host-key verification
+//     uses the NetBird peer registry.
+//   - NetBird-SSH server without JWT: authenticates with the NetBird SSH
+//     private key. Host-key verification uses the NetBird peer registry.
+//   - Regular SSH server (e.g. OpenSSH): authenticates with the NetBird key
+//     first (so a user-installed NetBird public key works), then falls back
+//     to the supplied password if non-empty. Host-key verification is
+//     disabled (TOFU pending).
+//
+// The password parameter is only consulted for regular SSH servers.
+func (s *SSHClient) Connect(host string, port int, user, password string) error {
+	cfg, _, cc := s.nb.stateSnapshot()
+	if cc == nil {
+		return errors.New("netbird client not running")
+	}
+	if cfg == nil {
+		return errors.New("netbird config not loaded")
+	}
+	engine := cc.Engine()
+	if engine == nil {
+		return errors.New("netbird engine not available")
+	}
+
+	serverType := detectServerType(host, port)
+	log.Infof("SSH server type for %s:%d: %s", host, port, serverType)
+
+	authMethods, hostKeyCallback, err := s.buildAuth(cfg, engine, serverType, password)
+	if err != nil {
+		return err
+	}
+
+	clientConfig := &gossh.ClientConfig{
+		User:            user,
+		Auth:            authMethods,
+		HostKeyCallback: hostKeyCallback,
+		Timeout:         sshDialTimeout,
+	}
+	return s.dialAndHandshake(host, port, clientConfig)
+}
+
+// StartSession requests a PTY and starts an interactive shell. Output from
+// the session is forwarded to the listener via OnData.
+func (s *SSHClient) StartSession(cols, rows int) error {
+	log.Debugf("SSH: starting session %dx%d", cols, rows)
+	s.mu.Lock()
+	sshClient := s.sshClient
+	s.mu.Unlock()
+
+	if sshClient == nil {
+		return errors.New("ssh client not connected")
+	}
+
+	session, err := sshClient.NewSession()
+	if err != nil {
+		return fmt.Errorf("new session: %w", err)
+	}
+
+	modes := gossh.TerminalModes{
+		gossh.ECHO:          1,
+		gossh.TTY_OP_ISPEED: 14400,
+		gossh.TTY_OP_OSPEED: 14400,
+		gossh.VINTR:         3,
+		gossh.VQUIT:         28,
+		gossh.VERASE:        127,
+	}
+	if err := session.RequestPty("xterm-256color", rows, cols, modes); err != nil {
+		closeQuiet(session, "session after pty error")
+		return fmt.Errorf("request pty: %w", err)
+	}
+
+	stdin, err := session.StdinPipe()
+	if err != nil {
+		closeQuiet(session, "session after stdin error")
+		return fmt.Errorf("stdin pipe: %w", err)
+	}
+	stdout, err := session.StdoutPipe()
+	if err != nil {
+		closeQuiet(session, "session after stdout error")
+		return fmt.Errorf("stdout pipe: %w", err)
+	}
+	stderr, err := session.StderrPipe()
+	if err != nil {
+		closeQuiet(session, "session after stderr error")
+		return fmt.Errorf("stderr pipe: %w", err)
+	}
+
+	if err := session.Shell(); err != nil {
+		closeQuiet(session, "session after shell error")
+		return fmt.Errorf("start shell: %w", err)
+	}
+
+	s.mu.Lock()
+	s.session = session
+	s.stdin = stdin
+	s.mu.Unlock()
+
+	go s.readLoop(stdout, "stdout")
+	go s.readLoop(stderr, "stderr")
+	log.Debug("SSH: session started, shell running")
+	return nil
+}
+
+// Write sends data to the SSH session stdin.
+func (s *SSHClient) Write(data []byte) error {
+	s.mu.Lock()
+	stdin := s.stdin
+	s.mu.Unlock()
+	if stdin == nil {
+		return errors.New("ssh session not started")
+	}
+	if _, err := stdin.Write(data); err != nil {
+		return fmt.Errorf("write stdin: %w", err)
+	}
+	return nil
+}
+
+// Resize updates the PTY window size.
+func (s *SSHClient) Resize(cols, rows int) error {
+	s.mu.Lock()
+	session := s.session
+	s.mu.Unlock()
+	if session == nil {
+		return errors.New("ssh session not started")
+	}
+	return session.WindowChange(rows, cols)
+}
+
+// Close terminates the SSH session and underlying connection. Safe to call
+// multiple times.
+func (s *SSHClient) Close() error {
+	s.mu.Lock()
+	sshClient := s.sshClient
+	session := s.session
+	stdin := s.stdin
+	s.sshClient = nil
+	s.session = nil
+	s.stdin = nil
+	s.mu.Unlock()
+
+	if stdin != nil {
+		if err := stdin.Close(); err != nil {
+			log.Debugf("ssh: stdin close: %v", err)
+		}
+	}
+	if session != nil {
+		if err := session.Close(); err != nil && !errors.Is(err, io.EOF) {
+			log.Debugf("ssh: session close: %v", err)
+		}
+	}
+	var firstErr error
+	if sshClient != nil {
+		if err := sshClient.Close(); err != nil {
+			firstErr = err
+		}
+	}
+	s.notifyClose("closed by client")
+	return firstErr
+}
+
+func (s *SSHClient) buildAuth(cfg *profilemanager.Config, engine *internal.Engine,
+	serverType detection.ServerType, password string) ([]gossh.AuthMethod, gossh.HostKeyCallback, error) {
+
+	switch serverType {
+	case detection.ServerTypeNetBirdJWT:
+		token, err := s.requestJWTToken(cfg)
+		if err != nil {
+			return nil, nil, fmt.Errorf("jwt: %w", err)
+		}
+		auths := []gossh.AuthMethod{gossh.Password(token)}
+		return auths, nbssh.CreateHostKeyCallback(&engineHostKeyVerifier{engine: engine}), nil
+
+	case detection.ServerTypeNetBirdNoJWT:
+		if cfg.SSHKey == "" {
+			return nil, nil, errors.New("no NetBird SSH key available")
+		}
+		signer, err := gossh.ParsePrivateKey([]byte(cfg.SSHKey))
+		if err != nil {
+			return nil, nil, fmt.Errorf("parse netbird ssh key: %w", err)
+		}
+		auths := []gossh.AuthMethod{gossh.PublicKeys(signer)}
+		return auths, nbssh.CreateHostKeyCallback(&engineHostKeyVerifier{engine: engine}), nil
+
+	default: // regular SSH
+		var auths []gossh.AuthMethod
+		if cfg.SSHKey != "" {
+			if signer, err := gossh.ParsePrivateKey([]byte(cfg.SSHKey)); err == nil {
+				auths = append(auths, gossh.PublicKeys(signer))
+			} else {
+				log.Debugf("ssh: parse netbird key for regular auth: %v", err)
+			}
+		}
+		if password != "" {
+			pw := password
+			auths = append(auths, gossh.Password(pw))
+			auths = append(auths, gossh.KeyboardInteractive(func(_, _ string, questions []string, _ []bool) ([]string, error) {
+				answers := make([]string, len(questions))
+				for i := range questions {
+					answers[i] = pw
+				}
+				return answers, nil
+			}))
+		}
+		if len(auths) == 0 {
+			return nil, nil, errors.New("no auth method available: provide a password or configure NetBird SSH key")
+		}
+		return auths, gossh.InsecureIgnoreHostKey(), nil // nolint:gosec // TOFU not yet implemented
+	}
+}
+
+func (s *SSHClient) requestJWTToken(cfg *profilemanager.Config) (string, error) {
+	s.mu.Lock()
+	urlOpener := s.urlOpener
+	s.mu.Unlock()
+	if urlOpener == nil {
+		return "", errors.New("URL opener not configured for JWT auth")
+	}
+
+	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Minute)
+	defer cancel()
+
+	flow, err := auth.NewOAuthFlow(ctx, cfg, false, true, profilemanager.GetLoginHint())
+	if err != nil {
+		return "", fmt.Errorf("create oauth flow: %w", err)
+	}
+
+	flowInfo, err := flow.RequestAuthInfo(ctx)
+	if err != nil {
+		return "", fmt.Errorf("request auth info: %w", err)
+	}
+
+	go urlOpener.Open(flowInfo.VerificationURIComplete, flowInfo.UserCode)
+
+	tokenInfo, err := flow.WaitToken(ctx, flowInfo)
+	if err != nil {
+		return "", fmt.Errorf("wait for token: %w", err)
+	}
+
+	token := tokenInfo.GetTokenToUse()
+	if token == "" {
+		return "", errors.New("empty token returned by IdP")
+	}
+	return token, nil
+}
+
+func (s *SSHClient) dialAndHandshake(host string, port int, clientConfig *gossh.ClientConfig) error {
+	addr := net.JoinHostPort(host, strconv.Itoa(port))
+	log.Infof("SSH: connecting to %s as %s", addr, clientConfig.User)
+
+	ctx, cancel := context.WithTimeout(context.Background(), sshDialTimeout)
+	defer cancel()
+
+	var dialer net.Dialer
+	conn, err := dialer.DialContext(ctx, "tcp", addr)
+	if err != nil {
+		return fmt.Errorf("dial %s: %w", addr, err)
+	}
+
+	sshConn, chans, reqs, err := gossh.NewClientConn(conn, addr, clientConfig)
+	if err != nil {
+		if cerr := conn.Close(); cerr != nil {
+			log.Debugf("ssh: close after handshake error: %v", cerr)
+		}
+		return fmt.Errorf("ssh handshake: %w", err)
+	}
+
+	s.mu.Lock()
+	s.sshClient = gossh.NewClient(sshConn, chans, reqs)
+	listener := s.listener
+	s.mu.Unlock()
+
+	log.Infof("SSH: connected to %s", addr)
+	if listener != nil {
+		listener.OnConnected()
+	}
+	return nil
+}
+
+func (s *SSHClient) readLoop(r io.Reader, name string) {
+	buf := make([]byte, 4096)
+	for {
+		n, err := r.Read(buf)
+		if n > 0 {
+			s.mu.Lock()
+			listener := s.listener
+			s.mu.Unlock()
+			if listener != nil {
+				chunk := make([]byte, n)
+				copy(chunk, buf[:n])
+				listener.OnData(chunk)
+			}
+		}
+		if err != nil {
+			if !errors.Is(err, io.EOF) {
+				log.Debugf("ssh %s read: %v", name, err)
+			}
+			s.notifyClose(err.Error())
+			return
+		}
+	}
+}
+
+func (s *SSHClient) notifyClose(reason string) {
+	s.mu.Lock()
+	if s.closed {
+		s.mu.Unlock()
+		return
+	}
+	s.closed = true
+	listener := s.listener
+	s.mu.Unlock()
+	if listener != nil {
+		listener.OnClose(reason)
+	}
+}
+
+// engineHostKeyVerifier adapts *internal.Engine to nbssh.HostKeyVerifier.
+type engineHostKeyVerifier struct {
+	engine *internal.Engine
+}
+
+func (v *engineHostKeyVerifier) VerifySSHHostKey(peerAddress string, presented []byte) error {
+	storedKey, found := v.engine.GetPeerSSHKey(peerAddress)
+	if !found {
+		return nbssh.ErrPeerNotFound
+	}
+	return nbssh.VerifyHostKey(storedKey, presented, peerAddress)
+}
+
+func closeQuiet(c io.Closer, label string) {
+	if c == nil {
+		return
+	}
+	if err := c.Close(); err != nil && !errors.Is(err, io.EOF) {
+		log.Debugf("ssh: close %s: %v", label, err)
+	}
+}
+
+func detectServerType(host string, port int) detection.ServerType {
+	ctx, cancel := context.WithTimeout(context.Background(), sshDetectionTimeout)
+	defer cancel()
+
+	dialer := &net.Dialer{}
+	serverType, err := detection.DetectSSHServerType(ctx, dialer, host, port)
+	if err != nil {
+		log.Debugf("ssh: server detection for %s:%d failed: %v (assuming regular SSH)", host, port, err)
+		return detection.ServerTypeRegular
+	}
+	return serverType
+}

client/iface/bind/ice_bind_test.go

@@ -239,12 +239,8 @@ func TestICEBind_HandlesConcurrentMixedTraffic(t *testing.T) {
 		ipv6Count++
 	}
 
-	// Allow some UDP packet loss under load (e.g. FreeBSD/QEMU runners). The
-	// routing-correctness checks above are the real assertions; the counts
-	// are a sanity bound to catch a totally silent path.
-	minDelivered := packetsPerFamily * 80 / 100
-	assert.GreaterOrEqual(t, ipv4Count, minDelivered, "IPv4 delivery below threshold")
-	assert.GreaterOrEqual(t, ipv6Count, minDelivered, "IPv6 delivery below threshold")
+	assert.Equal(t, packetsPerFamily, ipv4Count)
+	assert.Equal(t, packetsPerFamily, ipv6Count)
 }
 
 func TestICEBind_DetectsAddressFamilyFromConnection(t *testing.T) {

client/internal/debug/upload_test.go

@@ -3,12 +3,10 @@ package debug
 import (
 	"context"
 	"errors"
-	"net"
 	"net/http"
 	"os"
 	"path/filepath"
 	"testing"
-	"time"
 
 	"github.com/stretchr/testify/require"
 
@@ -21,10 +19,8 @@ func TestUpload(t *testing.T) {
 		t.Skip("Skipping upload test on docker ci")
 	}
 	testDir := t.TempDir()
-	addr := reserveLoopbackPort(t)
-	testURL := "http://" + addr
+	testURL := "http://localhost:8080"
 	t.Setenv("SERVER_URL", testURL)
-	t.Setenv("SERVER_ADDRESS", addr)
 	t.Setenv("STORE_DIR", testDir)
 	srv := server.NewServer()
 	go func() {
@@ -37,7 +33,6 @@ func TestUpload(t *testing.T) {
 			t.Errorf("Failed to stop server: %v", err)
 		}
 	})
-	waitForServer(t, addr)
 
 	file := filepath.Join(t.TempDir(), "tmpfile")
 	fileContent := []byte("test file content")
@@ -52,30 +47,3 @@ func TestUpload(t *testing.T) {
 	require.NoError(t, err)
 	require.Equal(t, fileContent, createdFileContent)
 }
-
-// reserveLoopbackPort binds an ephemeral port on loopback to learn a free
-// address, then releases it so the server under test can rebind. The close/
-// rebind window is racy in theory; on loopback with a kernel-assigned port
-// it's essentially never contended in practice.
-func reserveLoopbackPort(t *testing.T) string {
-	t.Helper()
-	l, err := net.Listen("tcp", "127.0.0.1:0")
-	require.NoError(t, err)
-	addr := l.Addr().String()
-	require.NoError(t, l.Close())
-	return addr
-}
-
-func waitForServer(t *testing.T, addr string) {
-	t.Helper()
-	deadline := time.Now().Add(5 * time.Second)
-	for time.Now().Before(deadline) {
-		c, err := net.DialTimeout("tcp", addr, 100*time.Millisecond)
-		if err == nil {
-			_ = c.Close()
-			return
-		}
-		time.Sleep(20 * time.Millisecond)
-	}
-	t.Fatalf("server did not start listening on %s in time", addr)
-}

client/internal/dns/file_parser_unix.go

@@ -13,7 +13,6 @@ import (
 
 const (
 	defaultResolvConfPath = "/etc/resolv.conf"
-	nsswitchConfPath      = "/etc/nsswitch.conf"
 )
 
 type resolvConf struct {

client/internal/dns/handler_chain.go

@@ -1,10 +1,7 @@
 package dns
 
 import (
-	"context"
 	"fmt"
-	"math"
-	"net"
 	"slices"
 	"strconv"
 	"strings"
@@ -195,12 +192,6 @@ func (c *HandlerChain) logHandlers() {
 }
 
 func (c *HandlerChain) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	c.dispatch(w, r, math.MaxInt)
-}
-
-// dispatch routes a DNS request through the chain, skipping handlers with
-// priority > maxPriority. Shared by ServeDNS and ResolveInternal.
-func (c *HandlerChain) dispatch(w dns.ResponseWriter, r *dns.Msg, maxPriority int) {
 	if len(r.Question) == 0 {
 		return
 	}
@@ -225,9 +216,6 @@ func (c *HandlerChain) dispatch(w dns.ResponseWriter, r *dns.Msg, maxPriority in
 
 	// Try handlers in priority order
 	for _, entry := range handlers {
-		if entry.Priority > maxPriority {
-			continue
-		}
 		if !c.isHandlerMatch(qname, entry) {
 			continue
 		}
@@ -285,55 +273,6 @@ func (c *HandlerChain) logResponse(logger *log.Entry, cw *ResponseWriterChain, q
 		cw.response.Len(), meta, time.Since(startTime))
 }
 
-// ResolveInternal runs an in-process DNS query against the chain, skipping any
-// handler with priority > maxPriority. Used by internal callers (e.g. the mgmt
-// cache refresher) that must bypass themselves to avoid loops. Honors ctx
-// cancellation; on ctx.Done the dispatch goroutine is left to drain on its own
-// (bounded by the invoked handler's internal timeout).
-func (c *HandlerChain) ResolveInternal(ctx context.Context, r *dns.Msg, maxPriority int) (*dns.Msg, error) {
-	if len(r.Question) == 0 {
-		return nil, fmt.Errorf("empty question")
-	}
-
-	base := &internalResponseWriter{}
-	done := make(chan struct{})
-	go func() {
-		c.dispatch(base, r, maxPriority)
-		close(done)
-	}()
-
-	select {
-	case <-done:
-	case <-ctx.Done():
-		// Prefer a completed response if dispatch finished concurrently with cancellation.
-		select {
-		case <-done:
-		default:
-			return nil, fmt.Errorf("resolve %s: %w", strings.ToLower(r.Question[0].Name), ctx.Err())
-		}
-	}
-
-	if base.response == nil || base.response.Rcode == dns.RcodeRefused {
-		return nil, fmt.Errorf("no handler resolved %s at priority ≤ %d",
-			strings.ToLower(r.Question[0].Name), maxPriority)
-	}
-	return base.response, nil
-}
-
-// HasRootHandlerAtOrBelow reports whether any "." handler is registered at
-// priority ≤ maxPriority.
-func (c *HandlerChain) HasRootHandlerAtOrBelow(maxPriority int) bool {
-	c.mu.RLock()
-	defer c.mu.RUnlock()
-
-	for _, h := range c.handlers {
-		if h.Pattern == "." && h.Priority <= maxPriority {
-			return true
-		}
-	}
-	return false
-}
-
 func (c *HandlerChain) isHandlerMatch(qname string, entry HandlerEntry) bool {
 	switch {
 	case entry.Pattern == ".":
@@ -352,36 +291,3 @@ func (c *HandlerChain) isHandlerMatch(qname string, entry HandlerEntry) bool {
 		}
 	}
 }
-
-// internalResponseWriter captures a dns.Msg for in-process chain queries.
-type internalResponseWriter struct {
-	response *dns.Msg
-}
-
-func (w *internalResponseWriter) WriteMsg(m *dns.Msg) error { w.response = m; return nil }
-func (w *internalResponseWriter) LocalAddr() net.Addr       { return nil }
-func (w *internalResponseWriter) RemoteAddr() net.Addr      { return nil }
-
-// Write unpacks raw DNS bytes so handlers that call Write instead of WriteMsg
-// still surface their answer to ResolveInternal.
-func (w *internalResponseWriter) Write(p []byte) (int, error) {
-	msg := new(dns.Msg)
-	if err := msg.Unpack(p); err != nil {
-		return 0, err
-	}
-	w.response = msg
-	return len(p), nil
-}
-
-func (w *internalResponseWriter) Close() error      { return nil }
-func (w *internalResponseWriter) TsigStatus() error { return nil }
-
-// TsigTimersOnly is part of dns.ResponseWriter.
-func (w *internalResponseWriter) TsigTimersOnly(bool) {
-	// no-op: in-process queries carry no TSIG state.
-}
-
-// Hijack is part of dns.ResponseWriter.
-func (w *internalResponseWriter) Hijack() {
-	// no-op: in-process queries have no underlying connection to hand off.
-}

client/internal/dns/handler_chain_test.go

@@ -1,15 +1,11 @@
 package dns_test
 
 import (
-	"context"
-	"net"
 	"testing"
-	"time"
 
 	"github.com/miekg/dns"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/mock"
-	"github.com/stretchr/testify/require"
 
 	nbdns "github.com/netbirdio/netbird/client/internal/dns"
 	"github.com/netbirdio/netbird/client/internal/dns/test"
@@ -1046,163 +1042,3 @@ func TestHandlerChain_AddRemoveRoundtrip(t *testing.T) {
 		})
 	}
 }
-
-// answeringHandler writes a fixed A record to ack the query. Used to verify
-// which handler ResolveInternal dispatches to.
-type answeringHandler struct {
-	name string
-	ip   string
-}
-
-func (h *answeringHandler) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	resp := &dns.Msg{}
-	resp.SetReply(r)
-	resp.Answer = []dns.RR{&dns.A{
-		Hdr: dns.RR_Header{Name: r.Question[0].Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-		A:   net.ParseIP(h.ip).To4(),
-	}}
-	_ = w.WriteMsg(resp)
-}
-
-func (h *answeringHandler) String() string { return h.name }
-
-func TestHandlerChain_ResolveInternal_SkipsAboveMaxPriority(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-
-	high := &answeringHandler{name: "high", ip: "10.0.0.1"}
-	low := &answeringHandler{name: "low", ip: "10.0.0.2"}
-
-	chain.AddHandler("example.com.", high, nbdns.PriorityMgmtCache)
-	chain.AddHandler("example.com.", low, nbdns.PriorityUpstream)
-
-	r := new(dns.Msg)
-	r.SetQuestion("example.com.", dns.TypeA)
-
-	resp, err := chain.ResolveInternal(context.Background(), r, nbdns.PriorityUpstream)
-	assert.NoError(t, err)
-	assert.NotNil(t, resp)
-	assert.Equal(t, 1, len(resp.Answer))
-	a, ok := resp.Answer[0].(*dns.A)
-	assert.True(t, ok)
-	assert.Equal(t, "10.0.0.2", a.A.String(), "should skip mgmtCache handler and resolve via upstream")
-}
-
-func TestHandlerChain_ResolveInternal_ErrorWhenNoMatch(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-	high := &answeringHandler{name: "high", ip: "10.0.0.1"}
-	chain.AddHandler("example.com.", high, nbdns.PriorityMgmtCache)
-
-	r := new(dns.Msg)
-	r.SetQuestion("example.com.", dns.TypeA)
-
-	_, err := chain.ResolveInternal(context.Background(), r, nbdns.PriorityUpstream)
-	assert.Error(t, err, "no handler at or below maxPriority should error")
-}
-
-// rawWriteHandler packs a response and calls ResponseWriter.Write directly
-// (instead of WriteMsg), exercising the internalResponseWriter.Write path.
-type rawWriteHandler struct {
-	ip string
-}
-
-func (h *rawWriteHandler) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	resp := &dns.Msg{}
-	resp.SetReply(r)
-	resp.Answer = []dns.RR{&dns.A{
-		Hdr: dns.RR_Header{Name: r.Question[0].Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-		A:   net.ParseIP(h.ip).To4(),
-	}}
-	packed, err := resp.Pack()
-	if err != nil {
-		return
-	}
-	_, _ = w.Write(packed)
-}
-
-func TestHandlerChain_ResolveInternal_CapturesRawWrite(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-	chain.AddHandler("example.com.", &rawWriteHandler{ip: "10.0.0.3"}, nbdns.PriorityUpstream)
-
-	r := new(dns.Msg)
-	r.SetQuestion("example.com.", dns.TypeA)
-
-	resp, err := chain.ResolveInternal(context.Background(), r, nbdns.PriorityUpstream)
-	assert.NoError(t, err)
-	require.NotNil(t, resp)
-	require.Len(t, resp.Answer, 1)
-	a, ok := resp.Answer[0].(*dns.A)
-	require.True(t, ok)
-	assert.Equal(t, "10.0.0.3", a.A.String(), "handlers calling Write(packed) must still surface their answer")
-}
-
-func TestHandlerChain_ResolveInternal_EmptyQuestion(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-	_, err := chain.ResolveInternal(context.Background(), new(dns.Msg), nbdns.PriorityUpstream)
-	assert.Error(t, err)
-}
-
-// hangingHandler blocks indefinitely until closed, simulating a wedged upstream.
-type hangingHandler struct {
-	block chan struct{}
-}
-
-func (h *hangingHandler) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	<-h.block
-	resp := &dns.Msg{}
-	resp.SetReply(r)
-	_ = w.WriteMsg(resp)
-}
-
-func (h *hangingHandler) String() string { return "hangingHandler" }
-
-func TestHandlerChain_ResolveInternal_HonorsContextTimeout(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-	h := &hangingHandler{block: make(chan struct{})}
-	defer close(h.block)
-
-	chain.AddHandler("example.com.", h, nbdns.PriorityUpstream)
-
-	r := new(dns.Msg)
-	r.SetQuestion("example.com.", dns.TypeA)
-
-	ctx, cancel := context.WithTimeout(context.Background(), 100*time.Millisecond)
-	defer cancel()
-
-	start := time.Now()
-	_, err := chain.ResolveInternal(ctx, r, nbdns.PriorityUpstream)
-	elapsed := time.Since(start)
-
-	assert.Error(t, err)
-	assert.ErrorIs(t, err, context.DeadlineExceeded)
-	assert.Less(t, elapsed, 500*time.Millisecond, "ResolveInternal must return shortly after ctx deadline")
-}
-
-func TestHandlerChain_HasRootHandlerAtOrBelow(t *testing.T) {
-	chain := nbdns.NewHandlerChain()
-	h := &answeringHandler{name: "h", ip: "10.0.0.1"}
-
-	assert.False(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "empty chain")
-
-	chain.AddHandler("example.com.", h, nbdns.PriorityUpstream)
-	assert.False(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "non-root handler does not count")
-
-	chain.AddHandler(".", h, nbdns.PriorityMgmtCache)
-	assert.False(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "root handler above threshold excluded")
-
-	chain.AddHandler(".", h, nbdns.PriorityDefault)
-	assert.True(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "root handler at PriorityDefault included")
-
-	chain.RemoveHandler(".", nbdns.PriorityDefault)
-	assert.False(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream))
-
-	// Primary nsgroup case: root handler lands at PriorityUpstream.
-	chain.AddHandler(".", h, nbdns.PriorityUpstream)
-	assert.True(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "root at PriorityUpstream included")
-	chain.RemoveHandler(".", nbdns.PriorityUpstream)
-
-	// Fallback case: original /etc/resolv.conf entries land at PriorityFallback.
-	chain.AddHandler(".", h, nbdns.PriorityFallback)
-	assert.True(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream), "root at PriorityFallback included")
-	chain.RemoveHandler(".", nbdns.PriorityFallback)
-	assert.False(t, chain.HasRootHandlerAtOrBelow(nbdns.PriorityUpstream))
-}

client/internal/dns/host_unix.go

@@ -46,12 +46,12 @@ type restoreHostManager interface {
 }
 
 func newHostManager(wgInterface string) (hostManager, error) {
-	osManager, reason, err := getOSDNSManagerType()
+	osManager, err := getOSDNSManagerType()
 	if err != nil {
 		return nil, fmt.Errorf("get os dns manager type: %w", err)
 	}
 
-	log.Infof("System DNS manager discovered: %s (%s)", osManager, reason)
+	log.Infof("System DNS manager discovered: %s", osManager)
 	mgr, err := newHostManagerFromType(wgInterface, osManager)
 	// need to explicitly return nil mgr on error to avoid returning a non-nil interface containing a nil value
 	if err != nil {
@@ -74,49 +74,17 @@ func newHostManagerFromType(wgInterface string, osManager osManagerType) (restor
 	}
 }
 
-func getOSDNSManagerType() (osManagerType, string, error) {
-	resolved := isSystemdResolvedRunning()
-	nss := isLibnssResolveUsed()
-	stub := checkStub()
-
-	// Prefer systemd-resolved whenever it owns libc resolution, regardless of
-	// who wrote /etc/resolv.conf. File-mode rewrites do not affect lookups
-	// that go through nss-resolve, and in foreign mode they can loop back
-	// through resolved as an upstream.
-	if resolved && (nss || stub) {
-		return systemdManager, fmt.Sprintf("systemd-resolved active (nss-resolve=%t, stub=%t)", nss, stub), nil
-	}
-
-	mgr, reason, rejected, err := scanResolvConfHeader()
-	if err != nil {
-		return 0, "", err
-	}
-	if reason != "" {
-		return mgr, reason, nil
-	}
-
-	fallback := fmt.Sprintf("no manager matched (resolved=%t, nss-resolve=%t, stub=%t)", resolved, nss, stub)
-	if len(rejected) > 0 {
-		fallback += "; rejected: " + strings.Join(rejected, ", ")
-	}
-	return fileManager, fallback, nil
-}
-
-// scanResolvConfHeader walks /etc/resolv.conf header comments and returns the
-// matching manager. If reason is empty the caller should pick file mode and
-// use rejected for diagnostics.
-func scanResolvConfHeader() (osManagerType, string, []string, error) {
+func getOSDNSManagerType() (osManagerType, error) {
 	file, err := os.Open(defaultResolvConfPath)
 	if err != nil {
-		return 0, "", nil, fmt.Errorf("unable to open %s for checking owner, got error: %w", defaultResolvConfPath, err)
+		return 0, fmt.Errorf("unable to open %s for checking owner, got error: %w", defaultResolvConfPath, err)
 	}
 	defer func() {
-		if cerr := file.Close(); cerr != nil {
-			log.Errorf("close file %s: %s", defaultResolvConfPath, cerr)
+		if err := file.Close(); err != nil {
+			log.Errorf("close file %s: %s", defaultResolvConfPath, err)
 		}
 	}()
 
-	var rejected []string
 	scanner := bufio.NewScanner(file)
 	for scanner.Scan() {
 		text := scanner.Text()
@@ -124,48 +92,41 @@ func scanResolvConfHeader() (osManagerType, string, []string, error) {
 			continue
 		}
 		if text[0] != '#' {
-			break
+			return fileManager, nil
 		}
-		if mgr, reason, rej := matchResolvConfHeader(text); reason != "" {
-			return mgr, reason, nil, nil
-		} else if rej != "" {
-			rejected = append(rejected, rej)
+		if strings.Contains(text, fileGeneratedResolvConfContentHeader) {
+			return netbirdManager, nil
+		}
+		if strings.Contains(text, "NetworkManager") && isDbusListenerRunning(networkManagerDest, networkManagerDbusObjectNode) && isNetworkManagerSupported() {
+			return networkManager, nil
+		}
+		if strings.Contains(text, "systemd-resolved") && isSystemdResolvedRunning() {
+			if checkStub() {
+				return systemdManager, nil
+			} else {
+				return fileManager, nil
+			}
+		}
+		if strings.Contains(text, "resolvconf") {
+			if isSystemdResolveConfMode() {
+				return systemdManager, nil
+			}
+
+			return resolvConfManager, nil
 		}
 	}
 	if err := scanner.Err(); err != nil && err != io.EOF {
-		return 0, "", nil, fmt.Errorf("scan: %w", err)
+		return 0, fmt.Errorf("scan: %w", err)
 	}
-	return 0, "", rejected, nil
+
+	return fileManager, nil
 }
 
-// matchResolvConfHeader inspects a single comment line. Returns either a
-// definitive (manager, reason) or a non-empty rejected diagnostic.
-func matchResolvConfHeader(text string) (osManagerType, string, string) {
-	if strings.Contains(text, fileGeneratedResolvConfContentHeader) {
-		return netbirdManager, "netbird-managed resolv.conf header detected", ""
-	}
-	if strings.Contains(text, "NetworkManager") {
-		if isDbusListenerRunning(networkManagerDest, networkManagerDbusObjectNode) && isNetworkManagerSupported() {
-			return networkManager, "NetworkManager header + supported version on dbus", ""
-		}
-		return 0, "", "NetworkManager header (no dbus or unsupported version)"
-	}
-	if strings.Contains(text, "resolvconf") {
-		if isSystemdResolveConfMode() {
-			return systemdManager, "resolvconf header in systemd-resolved compatibility mode", ""
-		}
-		return resolvConfManager, "resolvconf header detected", ""
-	}
-	return 0, "", ""
-}
-
-// checkStub reports whether systemd-resolved's stub (127.0.0.53) is listed
-// in /etc/resolv.conf. On parse failure we assume it is, to avoid dropping
-// into file mode while resolved is active.
+// checkStub checks if the stub resolver is disabled in systemd-resolved. If it is disabled, we fall back to file manager.
 func checkStub() bool {
 	rConf, err := parseDefaultResolvConf()
 	if err != nil {
-		log.Warnf("failed to parse resolv conf, assuming stub is active: %s", err)
+		log.Warnf("failed to parse resolv conf: %s", err)
 		return true
 	}
 
@@ -178,36 +139,3 @@ func checkStub() bool {
 
 	return false
 }
-
-// isLibnssResolveUsed reports whether nss-resolve is listed before dns on
-// the hosts: line of /etc/nsswitch.conf. When it is, libc lookups are
-// delegated to systemd-resolved regardless of /etc/resolv.conf.
-func isLibnssResolveUsed() bool {
-	bs, err := os.ReadFile(nsswitchConfPath)
-	if err != nil {
-		log.Debugf("read %s: %v", nsswitchConfPath, err)
-		return false
-	}
-	return parseNsswitchResolveAhead(bs)
-}
-
-func parseNsswitchResolveAhead(data []byte) bool {
-	for _, line := range strings.Split(string(data), "\n") {
-		if i := strings.IndexByte(line, '#'); i >= 0 {
-			line = line[:i]
-		}
-		fields := strings.Fields(line)
-		if len(fields) < 2 || fields[0] != "hosts:" {
-			continue
-		}
-		for _, module := range fields[1:] {
-			switch module {
-			case "dns":
-				return false
-			case "resolve":
-				return true
-			}
-		}
-	}
-	return false
-}

client/internal/dns/host_unix_test.go

@@ -1,76 +0,0 @@
-//go:build (linux && !android) || freebsd
-
-package dns
-
-import "testing"
-
-func TestParseNsswitchResolveAhead(t *testing.T) {
-	tests := []struct {
-		name string
-		in   string
-		want bool
-	}{
-		{
-			name: "resolve before dns with action token",
-			in:   "hosts: mymachines resolve [!UNAVAIL=return] files myhostname dns\n",
-			want: true,
-		},
-		{
-			name: "dns before resolve",
-			in:   "hosts: files mdns4_minimal [NOTFOUND=return] dns resolve\n",
-			want: false,
-		},
-		{
-			name: "debian default with only dns",
-			in:   "hosts: files mdns4_minimal [NOTFOUND=return] dns mymachines\n",
-			want: false,
-		},
-		{
-			name: "neither resolve nor dns",
-			in:   "hosts: files myhostname\n",
-			want: false,
-		},
-		{
-			name: "no hosts line",
-			in:   "passwd: files systemd\ngroup: files systemd\n",
-			want: false,
-		},
-		{
-			name: "empty",
-			in:   "",
-			want: false,
-		},
-		{
-			name: "comments and blank lines ignored",
-			in:   "# comment\n\n# another\nhosts: resolve dns\n",
-			want: true,
-		},
-		{
-			name: "trailing inline comment",
-			in:   "hosts: resolve [!UNAVAIL=return] dns # fallback\n",
-			want: true,
-		},
-		{
-			name: "hosts token must be the first field",
-			in:   "  hosts: resolve dns\n",
-			want: true,
-		},
-		{
-			name: "other db line mentioning resolve is ignored",
-			in:   "networks: resolve\nhosts: dns\n",
-			want: false,
-		},
-		{
-			name: "only resolve, no dns",
-			in:   "hosts: files resolve\n",
-			want: true,
-		},
-	}
-	for _, tt := range tests {
-		t.Run(tt.name, func(t *testing.T) {
-			if got := parseNsswitchResolveAhead([]byte(tt.in)); got != tt.want {
-				t.Errorf("parseNsswitchResolveAhead() = %v, want %v", got, tt.want)
-			}
-		})
-	}
-}

client/internal/dns/mgmt/mgmt.go

@@ -2,83 +2,40 @@ package mgmt
 
 import (
 	"context"
-	"errors"
 	"fmt"
 	"net"
+	"net/netip"
 	"net/url"
-	"os"
-	"slices"
 	"strings"
 	"sync"
-	"sync/atomic"
 	"time"
 
 	"github.com/miekg/dns"
 	log "github.com/sirupsen/logrus"
-	"golang.org/x/sync/singleflight"
 
 	dnsconfig "github.com/netbirdio/netbird/client/internal/dns/config"
-	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/shared/management/domain"
 )
 
-const (
-	dnsTimeout     = 5 * time.Second
-	defaultTTL     = 300 * time.Second
-	refreshBackoff = 30 * time.Second
+const dnsTimeout = 5 * time.Second
 
-	// envMgmtCacheTTL overrides defaultTTL for integration/dev testing.
-	envMgmtCacheTTL = "NB_MGMT_CACHE_TTL"
-)
-
-// ChainResolver lets the cache refresh stale entries through the DNS handler
-// chain instead of net.DefaultResolver, avoiding loopback when NetBird is the
-// system resolver.
-type ChainResolver interface {
-	ResolveInternal(ctx context.Context, msg *dns.Msg, maxPriority int) (*dns.Msg, error)
-	HasRootHandlerAtOrBelow(maxPriority int) bool
-}
-
-// cachedRecord holds DNS records plus timestamps used for TTL refresh.
-// records and cachedAt are set at construction and treated as immutable;
-// lastFailedRefresh and consecFailures are mutable and must be accessed under
-// Resolver.mutex.
-type cachedRecord struct {
-	records           []dns.RR
-	cachedAt          time.Time
-	lastFailedRefresh time.Time
-	consecFailures    int
-}
-
-// Resolver caches critical NetBird infrastructure domains.
-// records, refreshing, mgmtDomain and serverDomains are all guarded by mutex.
+// Resolver caches critical NetBird infrastructure domains
 type Resolver struct {
-	records       map[dns.Question]*cachedRecord
+	records       map[dns.Question][]dns.RR
 	mgmtDomain    *domain.Domain
 	serverDomains *dnsconfig.ServerDomains
 	mutex         sync.RWMutex
+}
 
-	chain            ChainResolver
-	chainMaxPriority int
-	refreshGroup     singleflight.Group
-
-	// refreshing tracks questions whose refresh is running via the OS
-	// fallback path. A ServeDNS hit for a question in this map indicates
-	// the OS resolver routed the recursive query back to us (loop). Only
-	// the OS path arms this so chain-path refreshes don't produce false
-	// positives. The atomic bool is CAS-flipped once per refresh to
-	// throttle the warning log.
-	refreshing map[dns.Question]*atomic.Bool
-
-	cacheTTL time.Duration
+type ipsResponse struct {
+	ips []netip.Addr
+	err error
 }
 
 // NewResolver creates a new management domains cache resolver.
 func NewResolver() *Resolver {
 	return &Resolver{
-		records:    make(map[dns.Question]*cachedRecord),
-		refreshing: make(map[dns.Question]*atomic.Bool),
-		cacheTTL:   resolveCacheTTL(),
+		records: make(map[dns.Question][]dns.RR),
 	}
 }
 
@@ -87,19 +44,7 @@ func (m *Resolver) String() string {
 	return "MgmtCacheResolver"
 }
 
-// SetChainResolver wires the handler chain used to refresh stale cache entries.
-// maxPriority caps which handlers may answer refresh queries (typically
-// PriorityUpstream, so upstream/default/fallback handlers are consulted and
-// mgmt/route/local handlers are skipped).
-func (m *Resolver) SetChainResolver(chain ChainResolver, maxPriority int) {
-	m.mutex.Lock()
-	m.chain = chain
-	m.chainMaxPriority = maxPriority
-	m.mutex.Unlock()
-}
-
-// ServeDNS serves cached A/AAAA records. Stale entries are returned
-// immediately and refreshed asynchronously (stale-while-revalidate).
+// ServeDNS implements dns.Handler interface.
 func (m *Resolver) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 	if len(r.Question) == 0 {
 		m.continueToNext(w, r)
@@ -115,14 +60,7 @@ func (m *Resolver) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 	}
 
 	m.mutex.RLock()
-	cached, found := m.records[question]
-	inflight := m.refreshing[question]
-	var shouldRefresh bool
-	if found {
-		stale := time.Since(cached.cachedAt) > m.cacheTTL
-		inBackoff := !cached.lastFailedRefresh.IsZero() && time.Since(cached.lastFailedRefresh) < refreshBackoff
-		shouldRefresh = stale && !inBackoff
-	}
+	records, found := m.records[question]
 	m.mutex.RUnlock()
 
 	if !found {
@@ -130,23 +68,12 @@ func (m *Resolver) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 		return
 	}
 
-	if inflight != nil && inflight.CompareAndSwap(false, true) {
-		log.Warnf("mgmt cache: possible resolver loop for domain=%s: served stale while an OS-fallback refresh was inflight (if NetBird is the system resolver, the OS-path predicate is wrong)",
-			question.Name)
-	}
-
-	// Skip scheduling a refresh goroutine if one is already inflight for
-	// this question; singleflight would dedup anyway but skipping avoids
-	// a parked goroutine per stale hit under bursty load.
-	if shouldRefresh && inflight == nil {
-		m.scheduleRefresh(question, cached)
-	}
-
 	resp := &dns.Msg{}
 	resp.SetReply(r)
 	resp.Authoritative = false
 	resp.RecursionAvailable = true
-	resp.Answer = cloneRecordsWithTTL(cached.records, m.responseTTL(cached.cachedAt))
+
+	resp.Answer = append(resp.Answer, records...)
 
 	log.Debugf("serving %d cached records for domain=%s", len(resp.Answer), question.Name)
 
@@ -171,260 +98,101 @@ func (m *Resolver) continueToNext(w dns.ResponseWriter, r *dns.Msg) {
 	}
 }
 
-// AddDomain resolves a domain and stores its A/AAAA records in the cache.
-// A family that resolves NODATA (nil err, zero records) evicts any stale
-// entry for that qtype.
+// AddDomain manually adds a domain to cache by resolving it.
 func (m *Resolver) AddDomain(ctx context.Context, d domain.Domain) error {
 	dnsName := strings.ToLower(dns.Fqdn(d.PunycodeString()))
 
 	ctx, cancel := context.WithTimeout(ctx, dnsTimeout)
 	defer cancel()
 
-	aRecords, aaaaRecords, errA, errAAAA := m.lookupBoth(ctx, d, dnsName)
-
-	if errA != nil && errAAAA != nil {
-		return fmt.Errorf("resolve %s: %w", d.SafeString(), errors.Join(errA, errAAAA))
+	ips, err := lookupIPWithExtraTimeout(ctx, d)
+	if err != nil {
+		return err
 	}
 
-	if len(aRecords) == 0 && len(aaaaRecords) == 0 {
-		if err := errors.Join(errA, errAAAA); err != nil {
-			return fmt.Errorf("resolve %s: no A/AAAA records: %w", d.SafeString(), err)
+	var aRecords, aaaaRecords []dns.RR
+	for _, ip := range ips {
+		if ip.Is4() {
+			rr := &dns.A{
+				Hdr: dns.RR_Header{
+					Name:   dnsName,
+					Rrtype: dns.TypeA,
+					Class:  dns.ClassINET,
+					Ttl:    300,
+				},
+				A: ip.AsSlice(),
+			}
+			aRecords = append(aRecords, rr)
+		} else if ip.Is6() {
+			rr := &dns.AAAA{
+				Hdr: dns.RR_Header{
+					Name:   dnsName,
+					Rrtype: dns.TypeAAAA,
+					Class:  dns.ClassINET,
+					Ttl:    300,
+				},
+				AAAA: ip.AsSlice(),
+			}
+			aaaaRecords = append(aaaaRecords, rr)
 		}
-		return fmt.Errorf("resolve %s: no A/AAAA records", d.SafeString())
 	}
 
-	now := time.Now()
 	m.mutex.Lock()
-	defer m.mutex.Unlock()
 
-	m.applyFamilyRecords(dnsName, dns.TypeA, aRecords, errA, now)
-	m.applyFamilyRecords(dnsName, dns.TypeAAAA, aaaaRecords, errAAAA, now)
+	if len(aRecords) > 0 {
+		aQuestion := dns.Question{
+			Name:   dnsName,
+			Qtype:  dns.TypeA,
+			Qclass: dns.ClassINET,
+		}
+		m.records[aQuestion] = aRecords
+	}
 
-	log.Debugf("added/updated domain=%s with %d A records and %d AAAA records",
+	if len(aaaaRecords) > 0 {
+		aaaaQuestion := dns.Question{
+			Name:   dnsName,
+			Qtype:  dns.TypeAAAA,
+			Qclass: dns.ClassINET,
+		}
+		m.records[aaaaQuestion] = aaaaRecords
+	}
+
+	m.mutex.Unlock()
+
+	log.Debugf("added domain=%s with %d A records and %d AAAA records",
 		d.SafeString(), len(aRecords), len(aaaaRecords))
 
 	return nil
 }
 
-// applyFamilyRecords writes records, evicts on NODATA, leaves the cache
-// untouched on error. Caller holds m.mutex.
-func (m *Resolver) applyFamilyRecords(dnsName string, qtype uint16, records []dns.RR, err error, now time.Time) {
-	q := dns.Question{Name: dnsName, Qtype: qtype, Qclass: dns.ClassINET}
-	switch {
-	case len(records) > 0:
-		m.records[q] = &cachedRecord{records: records, cachedAt: now}
-	case err == nil:
-		delete(m.records, q)
-	}
-}
+func lookupIPWithExtraTimeout(ctx context.Context, d domain.Domain) ([]netip.Addr, error) {
+	log.Infof("looking up IP for mgmt domain=%s", d.SafeString())
+	defer log.Infof("done looking up IP for mgmt domain=%s", d.SafeString())
+	resultChan := make(chan *ipsResponse, 1)
 
-// scheduleRefresh kicks off an async refresh. DoChan spawns one goroutine per
-// unique in-flight key; bursty stale hits share its channel. expected is the
-// cachedRecord pointer observed by the caller; the refresh only mutates the
-// cache if that pointer is still the one stored, so a stale in-flight refresh
-// can't clobber a newer entry written by AddDomain or a competing refresh.
-func (m *Resolver) scheduleRefresh(question dns.Question, expected *cachedRecord) {
-	key := question.Name + "|" + dns.TypeToString[question.Qtype]
-	_ = m.refreshGroup.DoChan(key, func() (any, error) {
-		return nil, m.refreshQuestion(question, expected)
-	})
-}
-
-// refreshQuestion replaces the cached records on success, or marks the entry
-// failed (arming the backoff) on failure. While this runs, ServeDNS can detect
-// a resolver loop by spotting a query for this same question arriving on us.
-// expected pins the cache entry observed at schedule time; mutations only apply
-// if m.records[question] still points at it.
-func (m *Resolver) refreshQuestion(question dns.Question, expected *cachedRecord) error {
-	ctx, cancel := context.WithTimeout(context.Background(), dnsTimeout)
-	defer cancel()
-
-	d, err := domain.FromString(strings.TrimSuffix(question.Name, "."))
-	if err != nil {
-		m.markRefreshFailed(question, expected)
-		return fmt.Errorf("parse domain: %w", err)
-	}
-
-	records, err := m.lookupRecords(ctx, d, question)
-	if err != nil {
-		fails := m.markRefreshFailed(question, expected)
-		logf := log.Warnf
-		if fails == 0 || fails > 1 {
-			logf = log.Debugf
+	go func() {
+		ips, err := net.DefaultResolver.LookupNetIP(ctx, "ip", d.PunycodeString())
+		resultChan <- &ipsResponse{
+			err: err,
+			ips: ips,
 		}
-		logf("refresh mgmt cache domain=%s type=%s: %v (consecutive failures=%d)",
-			d.SafeString(), dns.TypeToString[question.Qtype], err, fails)
-		return err
+	}()
+
+	var resp *ipsResponse
+
+	select {
+	case <-time.After(dnsTimeout + time.Millisecond*500):
+		log.Warnf("timed out waiting for IP for mgmt domain=%s", d.SafeString())
+		return nil, fmt.Errorf("timed out waiting for ips to be available for domain %s", d.SafeString())
+	case <-ctx.Done():
+		return nil, ctx.Err()
+	case resp = <-resultChan:
 	}
 
-	// NOERROR/NODATA: family gone upstream, evict so we stop serving stale.
-	if len(records) == 0 {
-		m.mutex.Lock()
-		if m.records[question] == expected {
-			delete(m.records, question)
-			m.mutex.Unlock()
-			log.Infof("removed mgmt cache domain=%s type=%s: no records returned",
-				d.SafeString(), dns.TypeToString[question.Qtype])
-			return nil
-		}
-		m.mutex.Unlock()
-		log.Debugf("skipping refresh evict for domain=%s type=%s: entry changed during refresh",
-			d.SafeString(), dns.TypeToString[question.Qtype])
-		return nil
+	if resp.err != nil {
+		return nil, fmt.Errorf("resolve domain %s: %w", d.SafeString(), resp.err)
 	}
-
-	now := time.Now()
-	m.mutex.Lock()
-	if m.records[question] != expected {
-		m.mutex.Unlock()
-		log.Debugf("skipping refresh write for domain=%s type=%s: entry changed during refresh",
-			d.SafeString(), dns.TypeToString[question.Qtype])
-		return nil
-	}
-	m.records[question] = &cachedRecord{records: records, cachedAt: now}
-	m.mutex.Unlock()
-
-	log.Infof("refreshed mgmt cache domain=%s type=%s",
-		d.SafeString(), dns.TypeToString[question.Qtype])
-	return nil
-}
-
-func (m *Resolver) markRefreshing(question dns.Question) {
-	m.mutex.Lock()
-	m.refreshing[question] = &atomic.Bool{}
-	m.mutex.Unlock()
-}
-
-func (m *Resolver) clearRefreshing(question dns.Question) {
-	m.mutex.Lock()
-	delete(m.refreshing, question)
-	m.mutex.Unlock()
-}
-
-// markRefreshFailed arms the backoff and returns the new consecutive-failure
-// count so callers can downgrade subsequent failure logs to debug.
-func (m *Resolver) markRefreshFailed(question dns.Question, expected *cachedRecord) int {
-	m.mutex.Lock()
-	defer m.mutex.Unlock()
-	c, ok := m.records[question]
-	if !ok || c != expected {
-		return 0
-	}
-	c.lastFailedRefresh = time.Now()
-	c.consecFailures++
-	return c.consecFailures
-}
-
-// lookupBoth resolves A and AAAA via chain or OS. Per-family errors let
-// callers tell records, NODATA (nil err, no records), and failure apart.
-func (m *Resolver) lookupBoth(ctx context.Context, d domain.Domain, dnsName string) (aRecords, aaaaRecords []dns.RR, errA, errAAAA error) {
-	m.mutex.RLock()
-	chain := m.chain
-	maxPriority := m.chainMaxPriority
-	m.mutex.RUnlock()
-
-	if chain != nil && chain.HasRootHandlerAtOrBelow(maxPriority) {
-		aRecords, errA = m.lookupViaChain(ctx, chain, maxPriority, dnsName, dns.TypeA)
-		aaaaRecords, errAAAA = m.lookupViaChain(ctx, chain, maxPriority, dnsName, dns.TypeAAAA)
-		return
-	}
-
-	// TODO: drop once every supported OS registers a fallback resolver. Safe
-	// today: no root handler at priority ≤ PriorityUpstream means NetBird is
-	// not the system resolver, so net.DefaultResolver will not loop back.
-	aRecords, errA = m.osLookup(ctx, d, dnsName, dns.TypeA)
-	aaaaRecords, errAAAA = m.osLookup(ctx, d, dnsName, dns.TypeAAAA)
-	return
-}
-
-// lookupRecords resolves a single record type via chain or OS. The OS branch
-// arms the loop detector for the duration of its call so that ServeDNS can
-// spot the OS resolver routing the recursive query back to us.
-func (m *Resolver) lookupRecords(ctx context.Context, d domain.Domain, q dns.Question) ([]dns.RR, error) {
-	m.mutex.RLock()
-	chain := m.chain
-	maxPriority := m.chainMaxPriority
-	m.mutex.RUnlock()
-
-	if chain != nil && chain.HasRootHandlerAtOrBelow(maxPriority) {
-		return m.lookupViaChain(ctx, chain, maxPriority, q.Name, q.Qtype)
-	}
-
-	// TODO: drop once every supported OS registers a fallback resolver.
-	m.markRefreshing(q)
-	defer m.clearRefreshing(q)
-
-	return m.osLookup(ctx, d, q.Name, q.Qtype)
-}
-
-// lookupViaChain resolves via the handler chain and rewrites each RR to use
-// dnsName as owner and m.cacheTTL as TTL, so CNAME-backed domains don't cache
-// target-owned records or upstream TTLs. NODATA returns (nil, nil).
-func (m *Resolver) lookupViaChain(ctx context.Context, chain ChainResolver, maxPriority int, dnsName string, qtype uint16) ([]dns.RR, error) {
-	msg := &dns.Msg{}
-	msg.SetQuestion(dnsName, qtype)
-	msg.RecursionDesired = true
-
-	resp, err := chain.ResolveInternal(ctx, msg, maxPriority)
-	if err != nil {
-		return nil, fmt.Errorf("chain resolve: %w", err)
-	}
-	if resp == nil {
-		return nil, fmt.Errorf("chain resolve returned nil response")
-	}
-	if resp.Rcode != dns.RcodeSuccess {
-		return nil, fmt.Errorf("chain resolve rcode=%s", dns.RcodeToString[resp.Rcode])
-	}
-
-	ttl := uint32(m.cacheTTL.Seconds())
-	owners := cnameOwners(dnsName, resp.Answer)
-	var filtered []dns.RR
-	for _, rr := range resp.Answer {
-		h := rr.Header()
-		if h.Class != dns.ClassINET || h.Rrtype != qtype {
-			continue
-		}
-		if !owners[strings.ToLower(dns.Fqdn(h.Name))] {
-			continue
-		}
-		if cp := cloneIPRecord(rr, dnsName, ttl); cp != nil {
-			filtered = append(filtered, cp)
-		}
-	}
-	return filtered, nil
-}
-
-// osLookup resolves a single family via net.DefaultResolver using resutil,
-// which disambiguates NODATA from NXDOMAIN and Unmaps v4-mapped-v6. NODATA
-// returns (nil, nil).
-func (m *Resolver) osLookup(ctx context.Context, d domain.Domain, dnsName string, qtype uint16) ([]dns.RR, error) {
-	network := resutil.NetworkForQtype(qtype)
-	if network == "" {
-		return nil, fmt.Errorf("unsupported qtype %s", dns.TypeToString[qtype])
-	}
-
-	log.Infof("looking up IP for mgmt domain=%s type=%s", d.SafeString(), dns.TypeToString[qtype])
-	defer log.Infof("done looking up IP for mgmt domain=%s type=%s", d.SafeString(), dns.TypeToString[qtype])
-
-	result := resutil.LookupIP(ctx, net.DefaultResolver, network, d.PunycodeString(), qtype)
-	if result.Rcode == dns.RcodeSuccess {
-		return resutil.IPsToRRs(dnsName, result.IPs, uint32(m.cacheTTL.Seconds())), nil
-	}
-
-	if result.Err != nil {
-		return nil, fmt.Errorf("resolve %s type=%s: %w", d.SafeString(), dns.TypeToString[qtype], result.Err)
-	}
-	return nil, fmt.Errorf("resolve %s type=%s: rcode=%s", d.SafeString(), dns.TypeToString[qtype], dns.RcodeToString[result.Rcode])
-}
-
-// responseTTL returns the remaining cache lifetime in seconds (rounded up),
-// so downstream resolvers don't cache an answer for longer than we will.
-func (m *Resolver) responseTTL(cachedAt time.Time) uint32 {
-	remaining := m.cacheTTL - time.Since(cachedAt)
-	if remaining <= 0 {
-		return 0
-	}
-	return uint32((remaining + time.Second - 1) / time.Second)
+	return resp.ips, nil
 }
 
 // PopulateFromConfig extracts and caches domains from the client configuration.
@@ -456,12 +224,19 @@ func (m *Resolver) RemoveDomain(d domain.Domain) error {
 	m.mutex.Lock()
 	defer m.mutex.Unlock()
 
-	qA := dns.Question{Name: dnsName, Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	qAAAA := dns.Question{Name: dnsName, Qtype: dns.TypeAAAA, Qclass: dns.ClassINET}
-	delete(m.records, qA)
-	delete(m.records, qAAAA)
-	delete(m.refreshing, qA)
-	delete(m.refreshing, qAAAA)
+	aQuestion := dns.Question{
+		Name:   dnsName,
+		Qtype:  dns.TypeA,
+		Qclass: dns.ClassINET,
+	}
+	delete(m.records, aQuestion)
+
+	aaaaQuestion := dns.Question{
+		Name:   dnsName,
+		Qtype:  dns.TypeAAAA,
+		Qclass: dns.ClassINET,
+	}
+	delete(m.records, aaaaQuestion)
 
 	log.Debugf("removed domain=%s from cache", d.SafeString())
 	return nil
@@ -619,73 +394,3 @@ func (m *Resolver) extractDomainsFromServerDomains(serverDomains dnsconfig.Serve
 
 	return domains
 }
-
-// cloneIPRecord returns a deep copy of rr retargeted to owner with ttl. Non
-// A/AAAA records return nil.
-func cloneIPRecord(rr dns.RR, owner string, ttl uint32) dns.RR {
-	switch r := rr.(type) {
-	case *dns.A:
-		cp := *r
-		cp.Hdr.Name = owner
-		cp.Hdr.Ttl = ttl
-		cp.A = slices.Clone(r.A)
-		return &cp
-	case *dns.AAAA:
-		cp := *r
-		cp.Hdr.Name = owner
-		cp.Hdr.Ttl = ttl
-		cp.AAAA = slices.Clone(r.AAAA)
-		return &cp
-	}
-	return nil
-}
-
-// cloneRecordsWithTTL clones A/AAAA records preserving their owner and
-// stamping ttl so the response shares no memory with the cached slice.
-func cloneRecordsWithTTL(records []dns.RR, ttl uint32) []dns.RR {
-	out := make([]dns.RR, 0, len(records))
-	for _, rr := range records {
-		if cp := cloneIPRecord(rr, rr.Header().Name, ttl); cp != nil {
-			out = append(out, cp)
-		}
-	}
-	return out
-}
-
-// cnameOwners returns dnsName plus every target reachable by following CNAMEs
-// in answer, iterating until fixed point so out-of-order chains resolve.
-func cnameOwners(dnsName string, answer []dns.RR) map[string]bool {
-	owners := map[string]bool{dnsName: true}
-	for {
-		added := false
-		for _, rr := range answer {
-			cname, ok := rr.(*dns.CNAME)
-			if !ok {
-				continue
-			}
-			name := strings.ToLower(dns.Fqdn(cname.Hdr.Name))
-			if !owners[name] {
-				continue
-			}
-			target := strings.ToLower(dns.Fqdn(cname.Target))
-			if !owners[target] {
-				owners[target] = true
-				added = true
-			}
-		}
-		if !added {
-			return owners
-		}
-	}
-}
-
-// resolveCacheTTL reads the cache TTL override env var; invalid or empty
-// values fall back to defaultTTL. Called once per Resolver from NewResolver.
-func resolveCacheTTL() time.Duration {
-	if v := os.Getenv(envMgmtCacheTTL); v != "" {
-		if d, err := time.ParseDuration(v); err == nil && d > 0 {
-			return d
-		}
-	}
-	return defaultTTL
-}

client/internal/dns/mgmt/mgmt_refresh_test.go

@@ -1,408 +0,0 @@
-package mgmt
-
-import (
-	"context"
-	"errors"
-	"net"
-	"sync"
-	"sync/atomic"
-	"testing"
-	"time"
-
-	"github.com/miekg/dns"
-	"github.com/stretchr/testify/assert"
-	"github.com/stretchr/testify/require"
-
-	"github.com/netbirdio/netbird/client/internal/dns/test"
-	"github.com/netbirdio/netbird/shared/management/domain"
-)
-
-type fakeChain struct {
-	mu       sync.Mutex
-	calls    map[string]int
-	answers  map[string][]dns.RR
-	err      error
-	hasRoot  bool
-	onLookup func()
-}
-
-func newFakeChain() *fakeChain {
-	return &fakeChain{
-		calls:   map[string]int{},
-		answers: map[string][]dns.RR{},
-		hasRoot: true,
-	}
-}
-
-func (f *fakeChain) HasRootHandlerAtOrBelow(maxPriority int) bool {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	return f.hasRoot
-}
-
-func (f *fakeChain) ResolveInternal(ctx context.Context, msg *dns.Msg, maxPriority int) (*dns.Msg, error) {
-	f.mu.Lock()
-	q := msg.Question[0]
-	key := q.Name + "|" + dns.TypeToString[q.Qtype]
-	f.calls[key]++
-	answers := f.answers[key]
-	err := f.err
-	onLookup := f.onLookup
-	f.mu.Unlock()
-
-	if onLookup != nil {
-		onLookup()
-	}
-	if err != nil {
-		return nil, err
-	}
-	resp := &dns.Msg{}
-	resp.SetReply(msg)
-	resp.Answer = answers
-	return resp, nil
-}
-
-func (f *fakeChain) setAnswer(name string, qtype uint16, ip string) {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	key := name + "|" + dns.TypeToString[qtype]
-	hdr := dns.RR_Header{Name: name, Rrtype: qtype, Class: dns.ClassINET, Ttl: 60}
-	switch qtype {
-	case dns.TypeA:
-		f.answers[key] = []dns.RR{&dns.A{Hdr: hdr, A: net.ParseIP(ip).To4()}}
-	case dns.TypeAAAA:
-		f.answers[key] = []dns.RR{&dns.AAAA{Hdr: hdr, AAAA: net.ParseIP(ip).To16()}}
-	}
-}
-
-func (f *fakeChain) callCount(name string, qtype uint16) int {
-	f.mu.Lock()
-	defer f.mu.Unlock()
-	return f.calls[name+"|"+dns.TypeToString[qtype]]
-}
-
-// waitFor polls the predicate until it returns true or the deadline passes.
-func waitFor(t *testing.T, d time.Duration, fn func() bool) {
-	t.Helper()
-	deadline := time.Now().Add(d)
-	for time.Now().Before(deadline) {
-		if fn() {
-			return
-		}
-		time.Sleep(5 * time.Millisecond)
-	}
-	t.Fatalf("condition not met within %s", d)
-}
-
-func queryA(t *testing.T, r *Resolver, name string) *dns.Msg {
-	t.Helper()
-	msg := new(dns.Msg)
-	msg.SetQuestion(name, dns.TypeA)
-	w := &test.MockResponseWriter{}
-	r.ServeDNS(w, msg)
-	return w.GetLastResponse()
-}
-
-func firstA(t *testing.T, resp *dns.Msg) string {
-	t.Helper()
-	require.NotNil(t, resp)
-	require.Greater(t, len(resp.Answer), 0, "expected at least one answer")
-	a, ok := resp.Answer[0].(*dns.A)
-	require.True(t, ok, "expected A record")
-	return a.A.String()
-}
-
-func TestResolver_CacheTTLGatesRefresh(t *testing.T) {
-	// Same cached entry age, different cacheTTL values: the shorter TTL must
-	// trigger a background refresh, the longer one must not. Proves that the
-	// per-Resolver cacheTTL field actually drives the stale decision.
-	cachedAt := time.Now().Add(-100 * time.Millisecond)
-
-	newRec := func() *cachedRecord {
-		return &cachedRecord{
-			records: []dns.RR{&dns.A{
-				Hdr: dns.RR_Header{Name: "mgmt.example.com.", Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-				A:   net.ParseIP("10.0.0.1").To4(),
-			}},
-			cachedAt: cachedAt,
-		}
-	}
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-
-	t.Run("short TTL treats entry as stale and refreshes", func(t *testing.T) {
-		r := NewResolver()
-		r.cacheTTL = 10 * time.Millisecond
-		chain := newFakeChain()
-		chain.setAnswer(q.Name, dns.TypeA, "10.0.0.2")
-		r.SetChainResolver(chain, 50)
-		r.records[q] = newRec()
-
-		resp := queryA(t, r, q.Name)
-		assert.Equal(t, "10.0.0.1", firstA(t, resp), "stale entry must be served while refresh runs")
-
-		waitFor(t, time.Second, func() bool {
-			return chain.callCount(q.Name, dns.TypeA) >= 1
-		})
-	})
-
-	t.Run("long TTL keeps entry fresh and skips refresh", func(t *testing.T) {
-		r := NewResolver()
-		r.cacheTTL = time.Hour
-		chain := newFakeChain()
-		chain.setAnswer(q.Name, dns.TypeA, "10.0.0.2")
-		r.SetChainResolver(chain, 50)
-		r.records[q] = newRec()
-
-		resp := queryA(t, r, q.Name)
-		assert.Equal(t, "10.0.0.1", firstA(t, resp))
-
-		time.Sleep(50 * time.Millisecond)
-		assert.Equal(t, 0, chain.callCount(q.Name, dns.TypeA), "fresh entry must not trigger refresh")
-	})
-}
-
-func TestResolver_ServeFresh_NoRefresh(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("mgmt.example.com.", dns.TypeA, "10.0.0.2")
-	r.SetChainResolver(chain, 50)
-
-	r.records[dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: "mgmt.example.com.", Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now(), // fresh
-	}
-
-	resp := queryA(t, r, "mgmt.example.com.")
-	assert.Equal(t, "10.0.0.1", firstA(t, resp))
-
-	time.Sleep(20 * time.Millisecond)
-	assert.Equal(t, 0, chain.callCount("mgmt.example.com.", dns.TypeA), "fresh entry must not trigger refresh")
-}
-
-func TestResolver_StaleTriggersAsyncRefresh(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("mgmt.example.com.", dns.TypeA, "10.0.0.2")
-	r.SetChainResolver(chain, 50)
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now().Add(-2 * defaultTTL), // stale
-	}
-
-	// First query: serves stale immediately.
-	resp := queryA(t, r, "mgmt.example.com.")
-	assert.Equal(t, "10.0.0.1", firstA(t, resp), "stale entry must be served while refresh runs")
-
-	waitFor(t, time.Second, func() bool {
-		return chain.callCount("mgmt.example.com.", dns.TypeA) >= 1
-	})
-
-	// Next query should now return the refreshed IP.
-	waitFor(t, time.Second, func() bool {
-		resp := queryA(t, r, "mgmt.example.com.")
-		return resp != nil && len(resp.Answer) > 0 && firstA(t, resp) == "10.0.0.2"
-	})
-}
-
-func TestResolver_ConcurrentStaleHitsCollapseRefresh(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("mgmt.example.com.", dns.TypeA, "10.0.0.2")
-
-	var inflight atomic.Int32
-	var maxInflight atomic.Int32
-	chain.onLookup = func() {
-		cur := inflight.Add(1)
-		defer inflight.Add(-1)
-		for {
-			prev := maxInflight.Load()
-			if cur <= prev || maxInflight.CompareAndSwap(prev, cur) {
-				break
-			}
-		}
-		time.Sleep(50 * time.Millisecond) // hold inflight long enough to collide
-	}
-
-	r.SetChainResolver(chain, 50)
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now().Add(-2 * defaultTTL),
-	}
-
-	var wg sync.WaitGroup
-	for i := 0; i < 50; i++ {
-		wg.Add(1)
-		go func() {
-			defer wg.Done()
-			queryA(t, r, "mgmt.example.com.")
-		}()
-	}
-	wg.Wait()
-
-	waitFor(t, 2*time.Second, func() bool {
-		return inflight.Load() == 0
-	})
-
-	calls := chain.callCount("mgmt.example.com.", dns.TypeA)
-	assert.LessOrEqual(t, calls, 2, "singleflight must collapse concurrent refreshes (got %d)", calls)
-	assert.Equal(t, int32(1), maxInflight.Load(), "only one refresh should run concurrently")
-}
-
-func TestResolver_RefreshFailureArmsBackoff(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.err = errors.New("boom")
-	r.SetChainResolver(chain, 50)
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now().Add(-2 * defaultTTL),
-	}
-
-	// First stale hit triggers a refresh attempt that fails.
-	resp := queryA(t, r, "mgmt.example.com.")
-	assert.Equal(t, "10.0.0.1", firstA(t, resp), "stale entry served while refresh fails")
-
-	waitFor(t, time.Second, func() bool {
-		return chain.callCount("mgmt.example.com.", dns.TypeA) == 1
-	})
-	waitFor(t, time.Second, func() bool {
-		r.mutex.RLock()
-		defer r.mutex.RUnlock()
-		c, ok := r.records[q]
-		return ok && !c.lastFailedRefresh.IsZero()
-	})
-
-	// Subsequent stale hits within backoff window should not schedule more refreshes.
-	for i := 0; i < 10; i++ {
-		queryA(t, r, "mgmt.example.com.")
-	}
-	time.Sleep(50 * time.Millisecond)
-	assert.Equal(t, 1, chain.callCount("mgmt.example.com.", dns.TypeA), "backoff must suppress further refreshes")
-}
-
-func TestResolver_NoRootHandler_SkipsChain(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.hasRoot = false
-	chain.setAnswer("mgmt.example.com.", dns.TypeA, "10.0.0.2")
-	r.SetChainResolver(chain, 50)
-
-	// With hasRoot=false the chain must not be consulted. Use a short
-	// deadline so the OS fallback returns quickly without waiting on a
-	// real network call in CI.
-	ctx, cancel := context.WithTimeout(context.Background(), 50*time.Millisecond)
-	defer cancel()
-	_, _, _, _ = r.lookupBoth(ctx, domain.Domain("mgmt.example.com"), "mgmt.example.com.")
-
-	assert.Equal(t, 0, chain.callCount("mgmt.example.com.", dns.TypeA),
-		"chain must not be used when no root handler is registered at the bound priority")
-}
-
-func TestResolver_ServeDuringRefreshSetsLoopFlag(t *testing.T) {
-	// ServeDNS being invoked for a question while a refresh for that question
-	// is inflight indicates a resolver loop (OS resolver sent the recursive
-	// query back to us). The inflightRefresh.loopLoggedOnce flag must be set.
-	r := NewResolver()
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now(),
-	}
-
-	// Simulate an inflight refresh.
-	r.markRefreshing(q)
-	defer r.clearRefreshing(q)
-
-	resp := queryA(t, r, "mgmt.example.com.")
-	assert.Equal(t, "10.0.0.1", firstA(t, resp), "stale entry must still be served to avoid breaking external queries")
-
-	r.mutex.RLock()
-	inflight := r.refreshing[q]
-	r.mutex.RUnlock()
-	require.NotNil(t, inflight)
-	assert.True(t, inflight.Load(), "loop flag must be set once a ServeDNS during refresh was observed")
-}
-
-func TestResolver_LoopFlagOnlyTrippedOncePerRefresh(t *testing.T) {
-	r := NewResolver()
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now(),
-	}
-
-	r.markRefreshing(q)
-	defer r.clearRefreshing(q)
-
-	// Multiple ServeDNS calls during the same refresh must not re-set the flag
-	// (CompareAndSwap from false -> true returns true only on the first call).
-	for range 5 {
-		queryA(t, r, "mgmt.example.com.")
-	}
-
-	r.mutex.RLock()
-	inflight := r.refreshing[q]
-	r.mutex.RUnlock()
-	assert.True(t, inflight.Load())
-}
-
-func TestResolver_NoLoopFlagWhenNotRefreshing(t *testing.T) {
-	r := NewResolver()
-
-	q := dns.Question{Name: "mgmt.example.com.", Qtype: dns.TypeA, Qclass: dns.ClassINET}
-	r.records[q] = &cachedRecord{
-		records: []dns.RR{&dns.A{
-			Hdr: dns.RR_Header{Name: q.Name, Rrtype: dns.TypeA, Class: dns.ClassINET, Ttl: 60},
-			A:   net.ParseIP("10.0.0.1").To4(),
-		}},
-		cachedAt: time.Now(),
-	}
-
-	queryA(t, r, "mgmt.example.com.")
-
-	r.mutex.RLock()
-	_, ok := r.refreshing[q]
-	r.mutex.RUnlock()
-	assert.False(t, ok, "no refresh inflight means no loop tracking")
-}
-
-func TestResolver_AddDomain_UsesChainWhenRootRegistered(t *testing.T) {
-	r := NewResolver()
-	chain := newFakeChain()
-	chain.setAnswer("mgmt.example.com.", dns.TypeA, "10.0.0.2")
-	chain.setAnswer("mgmt.example.com.", dns.TypeAAAA, "fd00::2")
-	r.SetChainResolver(chain, 50)
-
-	require.NoError(t, r.AddDomain(context.Background(), domain.Domain("mgmt.example.com")))
-
-	resp := queryA(t, r, "mgmt.example.com.")
-	assert.Equal(t, "10.0.0.2", firstA(t, resp))
-	assert.Equal(t, 1, chain.callCount("mgmt.example.com.", dns.TypeA))
-	assert.Equal(t, 1, chain.callCount("mgmt.example.com.", dns.TypeAAAA))
-}

client/internal/dns/mgmt/mgmt_test.go

@@ -6,7 +6,6 @@ import (
 	"net/url"
 	"strings"
 	"testing"
-	"time"
 
 	"github.com/miekg/dns"
 	"github.com/stretchr/testify/assert"
@@ -24,60 +23,6 @@ func TestResolver_NewResolver(t *testing.T) {
 	assert.False(t, resolver.MatchSubdomains())
 }
 
-func TestResolveCacheTTL(t *testing.T) {
-	tests := []struct {
-		name  string
-		value string
-		want  time.Duration
-	}{
-		{"unset falls back to default", "", defaultTTL},
-		{"valid duration", "45s", 45 * time.Second},
-		{"valid minutes", "2m", 2 * time.Minute},
-		{"malformed falls back to default", "not-a-duration", defaultTTL},
-		{"zero falls back to default", "0s", defaultTTL},
-		{"negative falls back to default", "-5s", defaultTTL},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			t.Setenv(envMgmtCacheTTL, tc.value)
-			got := resolveCacheTTL()
-			assert.Equal(t, tc.want, got, "parsed TTL should match")
-		})
-	}
-}
-
-func TestNewResolver_CacheTTLFromEnv(t *testing.T) {
-	t.Setenv(envMgmtCacheTTL, "7s")
-	r := NewResolver()
-	assert.Equal(t, 7*time.Second, r.cacheTTL, "NewResolver should evaluate cacheTTL once from env")
-}
-
-func TestResolver_ResponseTTL(t *testing.T) {
-	now := time.Now()
-	tests := []struct {
-		name     string
-		cacheTTL time.Duration
-		cachedAt time.Time
-		wantMin  uint32
-		wantMax  uint32
-	}{
-		{"fresh entry returns full TTL", 60 * time.Second, now, 59, 60},
-		{"half-aged entry returns half TTL", 60 * time.Second, now.Add(-30 * time.Second), 29, 31},
-		{"expired entry returns zero", 60 * time.Second, now.Add(-61 * time.Second), 0, 0},
-		{"exactly expired returns zero", 10 * time.Second, now.Add(-10 * time.Second), 0, 0},
-	}
-
-	for _, tc := range tests {
-		t.Run(tc.name, func(t *testing.T) {
-			r := &Resolver{cacheTTL: tc.cacheTTL}
-			got := r.responseTTL(tc.cachedAt)
-			assert.GreaterOrEqual(t, got, tc.wantMin, "remaining TTL should be >= wantMin")
-			assert.LessOrEqual(t, got, tc.wantMax, "remaining TTL should be <= wantMax")
-		})
-	}
-}
-
 func TestResolver_ExtractDomainFromURL(t *testing.T) {
 	tests := []struct {
 		name        string

client/internal/dns/server.go

@@ -212,7 +212,6 @@ func newDefaultServer(
 	ctx, stop := context.WithCancel(ctx)
 
 	mgmtCacheResolver := mgmt.NewResolver()
-	mgmtCacheResolver.SetChainResolver(handlerChain, PriorityUpstream)
 
 	defaultServer := &DefaultServer{
 		ctx:               ctx,

client/internal/engine.go

@@ -570,7 +570,7 @@ 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,

combined/config.yaml.example

@@ -119,8 +119,6 @@ server:
 
   # Reverse proxy settings (optional)
   # reverseProxy:
-  #   trustedHTTPProxies: []           # CIDRs of trusted reverse proxies (e.g. ["10.0.0.0/8"])
-  #   trustedHTTPProxiesCount: 0       # Number of trusted proxies in front of the server (alternative to trustedHTTPProxies)
-  #   trustedPeers: []                 # CIDRs of trusted peer networks (e.g. ["100.64.0.0/10"])
-  #   accessLogRetentionDays: 7        # Days to retain HTTP access logs. 0 (or unset) defaults to 7. Negative values disable cleanup (logs kept indefinitely).
-  #   accessLogCleanupIntervalHours: 24 # How often (in hours) to run the access-log cleanup job. 0 (or unset) is treated as "not set" and defaults to 24 hours; cleanup remains enabled. To disable cleanup, set accessLogRetentionDays to a negative value.
+  #   trustedHTTPProxies: []
+  #   trustedHTTPProxiesCount: 0
+  #   trustedPeers: []

flow/client/client_test.go

@@ -457,18 +457,6 @@ func TestReceive_ProtocolErrorStreamReconnect(t *testing.T) {
 
 	client, err := flow.NewClient("http://"+server.addr, "test-payload", "test-signature", 1*time.Second)
 	require.NoError(t, err)
-
-	// Cleanups run LIFO: the goroutine-drain registered here runs after Close below,
-	// which is when Receive has actually returned. Without this, the Receive goroutine
-	// can outlive the test and call t.Logf after teardown, panicking.
-	receiveDone := make(chan struct{})
-	t.Cleanup(func() {
-		select {
-		case <-receiveDone:
-		case <-time.After(2 * time.Second):
-			t.Error("Receive goroutine did not exit after Close")
-		}
-	})
 	t.Cleanup(func() {
 		err := client.Close()
 		assert.NoError(t, err, "failed to close flow")
@@ -480,7 +468,6 @@ func TestReceive_ProtocolErrorStreamReconnect(t *testing.T) {
 	receivedAfterReconnect := make(chan struct{})
 
 	go func() {
-		defer close(receiveDone)
 		err := client.Receive(ctx, 1*time.Second, func(msg *proto.FlowEventAck) error {
 			if msg.IsInitiator || len(msg.EventId) == 0 {
 				return nil

go.mod

@@ -323,5 +323,3 @@ replace github.com/pion/ice/v4 => github.com/netbirdio/ice/v4 v4.0.0-20250908184
 replace github.com/libp2p/go-netroute => github.com/netbirdio/go-netroute v0.0.0-20240611143515-f59b0e1d3944
 
 replace github.com/dexidp/dex => github.com/netbirdio/dex v0.244.0
-
-replace github.com/mailru/easyjson => github.com/netbirdio/easyjson v0.9.0

go.sum

@@ -400,6 +400,8 @@ github.com/lufia/plan9stats v0.0.0-20240513124658-fba389f38bae h1:dIZY4ULFcto4tA
 github.com/lufia/plan9stats v0.0.0-20240513124658-fba389f38bae/go.mod h1:ilwx/Dta8jXAgpFYFvSWEMwxmbWXyiUHkd5FwyKhb5k=
 github.com/magiconair/properties v1.8.10 h1:s31yESBquKXCV9a/ScB3ESkOjUYYv+X0rg8SYxI99mE=
 github.com/magiconair/properties v1.8.10/go.mod h1:Dhd985XPs7jluiymwWYZ0G4Z61jb3vdS329zhj2hYo0=
+github.com/mailru/easyjson v0.9.0 h1:PrnmzHw7262yW8sTBwxi1PdJA3Iw/EKBa8psRf7d9a4=
+github.com/mailru/easyjson v0.9.0/go.mod h1:1+xMtQp2MRNVL/V1bOzuP3aP8VNwRW55fQUto+XFtTU=
 github.com/mattermost/xml-roundtrip-validator v0.1.0 h1:RXbVD2UAl7A7nOTR4u7E3ILa4IbtvKBHw64LDsmu9hU=
 github.com/mattermost/xml-roundtrip-validator v0.1.0/go.mod h1:qccnGMcpgwcNaBnxqpJpWWUiPNr5H3O8eDgGV9gT5To=
 github.com/mattn/go-isatty v0.0.9/go.mod h1:YNRxwqDuOph6SZLI9vUUz6OYw3QyUt7WiY2yME+cCiQ=
@@ -447,8 +449,6 @@ github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 h1:C3w9PqII01/Oq
 github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822/go.mod h1:+n7T8mK8HuQTcFwEeznm/DIxMOiR9yIdICNftLE1DvQ=
 github.com/netbirdio/dex v0.244.0 h1:1GOvi8wnXYassnKGildzNqRHq0RbcfEUw7LKYpKIN7U=
 github.com/netbirdio/dex v0.244.0/go.mod h1:STGInJhPcAflrHmDO7vyit2kSq03PdL+8zQPoGALtcU=
-github.com/netbirdio/easyjson v0.9.0 h1:6Nw2lghSVuy8RSkAYDhDv1thBVEmfVbKZnV7T7Z6Aus=
-github.com/netbirdio/easyjson v0.9.0/go.mod h1:1+xMtQp2MRNVL/V1bOzuP3aP8VNwRW55fQUto+XFtTU=
 github.com/netbirdio/go-netroute v0.0.0-20240611143515-f59b0e1d3944 h1:TDtJKmM6Sf8uYFx/dMeqNOL90KUoRscdfpFZ3Im89uk=
 github.com/netbirdio/go-netroute v0.0.0-20240611143515-f59b0e1d3944/go.mod h1:sHA6TRxjQ6RLbnI+3R4DZo2Eseg/iKiPRfNmcuNySVQ=
 github.com/netbirdio/ice/v4 v4.0.0-20250908184934-6202be846b51 h1:Ov4qdafATOgGMB1wbSuh+0aAHcwz9hdvB6VZjh1mVMI=

infrastructure_files/getting-started.sh

@@ -472,7 +472,7 @@ start_services_and_show_instructions() {
       if [[ "$ENABLE_CROWDSEC" == "true" ]]; then
         echo "Registering CrowdSec bouncer..."
         local cs_retries=0
-        while ! $DOCKER_COMPOSE_COMMAND exec -T crowdsec cscli lapi status >/dev/null 2>&1; do
+        while ! $DOCKER_COMPOSE_COMMAND exec -T crowdsec cscli capi status >/dev/null 2>&1; do
           cs_retries=$((cs_retries + 1))
           if [[ $cs_retries -ge 30 ]]; then
             echo "WARNING: CrowdSec did not become ready. Skipping CrowdSec setup." > /dev/stderr

management/internals/server/boot.go

@@ -30,7 +30,6 @@ import (
 	nbcache "github.com/netbirdio/netbird/management/server/cache"
 	nbContext "github.com/netbirdio/netbird/management/server/context"
 	nbhttp "github.com/netbirdio/netbird/management/server/http"
-	"github.com/netbirdio/netbird/management/server/http/middleware"
 	"github.com/netbirdio/netbird/management/server/store"
 	"github.com/netbirdio/netbird/management/server/telemetry"
 	mgmtProto "github.com/netbirdio/netbird/shared/management/proto"
@@ -110,7 +109,7 @@ func (s *BaseServer) EventStore() activity.Store {
 
 func (s *BaseServer) APIHandler() http.Handler {
 	return Create(s, func() http.Handler {
-		httpAPIHandler, err := nbhttp.NewAPIHandler(context.Background(), s.AccountManager(), s.NetworksManager(), s.ResourcesManager(), s.RoutesManager(), s.GroupsManager(), s.GeoLocationManager(), s.AuthManager(), s.Metrics(), s.IntegratedValidator(), s.ProxyController(), s.PermissionsManager(), s.PeersManager(), s.SettingsManager(), s.ZonesManager(), s.RecordsManager(), s.NetworkMapController(), s.IdpManager(), s.ServiceManager(), s.ReverseProxyDomainManager(), s.AccessLogsManager(), s.ReverseProxyGRPCServer(), s.Config.ReverseProxy.TrustedHTTPProxies, s.RateLimiter())
+		httpAPIHandler, err := nbhttp.NewAPIHandler(context.Background(), s.AccountManager(), s.NetworksManager(), s.ResourcesManager(), s.RoutesManager(), s.GroupsManager(), s.GeoLocationManager(), s.AuthManager(), s.Metrics(), s.IntegratedValidator(), s.ProxyController(), s.PermissionsManager(), s.PeersManager(), s.SettingsManager(), s.ZonesManager(), s.RecordsManager(), s.NetworkMapController(), s.IdpManager(), s.ServiceManager(), s.ReverseProxyDomainManager(), s.AccessLogsManager(), s.ReverseProxyGRPCServer(), s.Config.ReverseProxy.TrustedHTTPProxies)
 		if err != nil {
 			log.Fatalf("failed to create API handler: %v", err)
 		}
@@ -118,15 +117,6 @@ func (s *BaseServer) APIHandler() http.Handler {
 	})
 }
 
-func (s *BaseServer) RateLimiter() *middleware.APIRateLimiter {
-	return Create(s, func() *middleware.APIRateLimiter {
-		cfg, enabled := middleware.RateLimiterConfigFromEnv()
-		limiter := middleware.NewAPIRateLimiter(cfg)
-		limiter.SetEnabled(enabled)
-		return limiter
-	})
-}
-
 func (s *BaseServer) GRPCServer() *grpc.Server {
 	return Create(s, func() *grpc.Server {
 		trustedPeers := s.Config.ReverseProxy.TrustedPeers

management/server/account_test.go

@@ -2311,29 +2311,6 @@ 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
@@ -3253,13 +3230,6 @@ 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 {
@@ -3278,7 +3248,7 @@ 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(peerUpdateTimeout):
+	case <-time.After(500 * time.Millisecond):
 		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(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -478,7 +478,7 @@ func TestDNSAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -518,7 +518,7 @@ func TestDNSAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			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(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -638,7 +638,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -656,7 +656,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -689,7 +689,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -730,7 +730,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -757,7 +757,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -804,7 +804,7 @@ func TestGroupAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/http/handler.go

@@ -5,6 +5,9 @@ import (
 	"fmt"
 	"net/http"
 	"net/netip"
+	"os"
+	"strconv"
+	"time"
 
 	"github.com/gorilla/mux"
 	"github.com/rs/cors"
@@ -63,11 +66,14 @@ import (
 )
 
 const (
-	apiPrefix = "/api"
+	apiPrefix              = "/api"
+	rateLimitingEnabledKey = "NB_API_RATE_LIMITING_ENABLED"
+	rateLimitingBurstKey   = "NB_API_RATE_LIMITING_BURST"
+	rateLimitingRPMKey     = "NB_API_RATE_LIMITING_RPM"
 )
 
 // NewAPIHandler creates the Management service HTTP API handler registering all the available endpoints.
-func NewAPIHandler(ctx context.Context, accountManager account.Manager, networksManager nbnetworks.Manager, resourceManager resources.Manager, routerManager routers.Manager, groupsManager nbgroups.Manager, LocationManager geolocation.Geolocation, authManager auth.Manager, appMetrics telemetry.AppMetrics, integratedValidator integrated_validator.IntegratedValidator, proxyController port_forwarding.Controller, permissionsManager permissions.Manager, peersManager nbpeers.Manager, settingsManager settings.Manager, zManager zones.Manager, rManager records.Manager, networkMapController network_map.Controller, idpManager idpmanager.Manager, serviceManager service.Manager, reverseProxyDomainManager *manager.Manager, reverseProxyAccessLogsManager accesslogs.Manager, proxyGRPCServer *nbgrpc.ProxyServiceServer, trustedHTTPProxies []netip.Prefix, rateLimiter *middleware.APIRateLimiter) (http.Handler, error) {
+func NewAPIHandler(ctx context.Context, accountManager account.Manager, networksManager nbnetworks.Manager, resourceManager resources.Manager, routerManager routers.Manager, groupsManager nbgroups.Manager, LocationManager geolocation.Geolocation, authManager auth.Manager, appMetrics telemetry.AppMetrics, integratedValidator integrated_validator.IntegratedValidator, proxyController port_forwarding.Controller, permissionsManager permissions.Manager, peersManager nbpeers.Manager, settingsManager settings.Manager, zManager zones.Manager, rManager records.Manager, networkMapController network_map.Controller, idpManager idpmanager.Manager, serviceManager service.Manager, reverseProxyDomainManager *manager.Manager, reverseProxyAccessLogsManager accesslogs.Manager, proxyGRPCServer *nbgrpc.ProxyServiceServer, trustedHTTPProxies []netip.Prefix) (http.Handler, error) {
 
 	// Register bypass paths for unauthenticated endpoints
 	if err := bypass.AddBypassPath("/api/instance"); err != nil {
@@ -88,10 +94,34 @@ func NewAPIHandler(ctx context.Context, accountManager account.Manager, networks
 		return nil, fmt.Errorf("failed to add bypass path: %w", err)
 	}
 
-	if rateLimiter == nil {
-		log.Warn("NewAPIHandler: nil rate limiter, rate limiting disabled")
-		rateLimiter = middleware.NewAPIRateLimiter(nil)
-		rateLimiter.SetEnabled(false)
+	var rateLimitingConfig *middleware.RateLimiterConfig
+	if os.Getenv(rateLimitingEnabledKey) == "true" {
+		rpm := 6
+		if v := os.Getenv(rateLimitingRPMKey); v != "" {
+			value, err := strconv.Atoi(v)
+			if err != nil {
+				log.Warnf("parsing %s env var: %v, using default %d", rateLimitingRPMKey, err, rpm)
+			} else {
+				rpm = value
+			}
+		}
+
+		burst := 500
+		if v := os.Getenv(rateLimitingBurstKey); v != "" {
+			value, err := strconv.Atoi(v)
+			if err != nil {
+				log.Warnf("parsing %s env var: %v, using default %d", rateLimitingBurstKey, err, burst)
+			} else {
+				burst = value
+			}
+		}
+
+		rateLimitingConfig = &middleware.RateLimiterConfig{
+			RequestsPerMinute: float64(rpm),
+			Burst:             burst,
+			CleanupInterval:   6 * time.Hour,
+			LimiterTTL:        24 * time.Hour,
+		}
 	}
 
 	authMiddleware := middleware.NewAuthMiddleware(
@@ -99,7 +129,7 @@ func NewAPIHandler(ctx context.Context, accountManager account.Manager, networks
 		accountManager.GetAccountIDFromUserAuth,
 		accountManager.SyncUserJWTGroups,
 		accountManager.GetUserFromUserAuth,
-		rateLimiter,
+		rateLimitingConfig,
 		appMetrics.GetMeter(),
 	)
 

management/server/http/middleware/auth_middleware.go

@@ -12,7 +12,6 @@ import (
 	"go.opentelemetry.io/otel/metric"
 
 	"github.com/netbirdio/management-integrations/integrations"
-
 	serverauth "github.com/netbirdio/netbird/management/server/auth"
 	nbcontext "github.com/netbirdio/netbird/management/server/context"
 	"github.com/netbirdio/netbird/management/server/http/middleware/bypass"
@@ -43,9 +42,14 @@ func NewAuthMiddleware(
 	ensureAccount EnsureAccountFunc,
 	syncUserJWTGroups SyncUserJWTGroupsFunc,
 	getUserFromUserAuth GetUserFromUserAuthFunc,
-	rateLimiter *APIRateLimiter,
+	rateLimiterConfig *RateLimiterConfig,
 	meter metric.Meter,
 ) *AuthMiddleware {
+	var rateLimiter *APIRateLimiter
+	if rateLimiterConfig != nil {
+		rateLimiter = NewAPIRateLimiter(rateLimiterConfig)
+	}
+
 	var patUsageTracker *PATUsageTracker
 	if meter != nil {
 		var err error
@@ -83,14 +87,17 @@ func (m *AuthMiddleware) Handler(h http.Handler) http.Handler {
 
 		switch authType {
 		case "bearer":
-			if err := m.checkJWTFromRequest(r, authHeader); err != nil {
+			request, err := m.checkJWTFromRequest(r, authHeader)
+			if err != nil {
 				log.WithContext(r.Context()).Errorf("Error when validating JWT: %s", err.Error())
 				util.WriteError(r.Context(), status.Errorf(status.Unauthorized, "token invalid"), w)
 				return
 			}
-			h.ServeHTTP(w, r)
+
+			h.ServeHTTP(w, request)
 		case "token":
-			if err := m.checkPATFromRequest(r, authHeader); err != nil {
+			request, err := m.checkPATFromRequest(r, authHeader)
+			if err != nil {
 				log.WithContext(r.Context()).Debugf("Error when validating PAT: %s", err.Error())
 				// Check if it's a status error, otherwise default to Unauthorized
 				if _, ok := status.FromError(err); !ok {
@@ -99,7 +106,7 @@ func (m *AuthMiddleware) Handler(h http.Handler) http.Handler {
 				util.WriteError(r.Context(), err, w)
 				return
 			}
-			h.ServeHTTP(w, r)
+			h.ServeHTTP(w, request)
 		default:
 			util.WriteError(r.Context(), status.Errorf(status.Unauthorized, "no valid authentication provided"), w)
 			return
@@ -108,19 +115,19 @@ func (m *AuthMiddleware) Handler(h http.Handler) http.Handler {
 }
 
 // CheckJWTFromRequest checks if the JWT is valid
-func (m *AuthMiddleware) checkJWTFromRequest(r *http.Request, authHeaderParts []string) error {
+func (m *AuthMiddleware) checkJWTFromRequest(r *http.Request, authHeaderParts []string) (*http.Request, error) {
 	token, err := getTokenFromJWTRequest(authHeaderParts)
 
 	// If an error occurs, call the error handler and return an error
 	if err != nil {
-		return fmt.Errorf("error extracting token: %w", err)
+		return r, fmt.Errorf("error extracting token: %w", err)
 	}
 
 	ctx := r.Context()
 
 	userAuth, validatedToken, err := m.authManager.ValidateAndParseToken(ctx, token)
 	if err != nil {
-		return err
+		return r, err
 	}
 
 	if impersonate, ok := r.URL.Query()["account"]; ok && len(impersonate) == 1 {
@@ -136,7 +143,7 @@ func (m *AuthMiddleware) checkJWTFromRequest(r *http.Request, authHeaderParts []
 	// we need to call this method because if user is new, we will automatically add it to existing or create a new account
 	accountId, _, err := m.ensureAccount(ctx, userAuth)
 	if err != nil {
-		return err
+		return r, err
 	}
 
 	if userAuth.AccountId != accountId {
@@ -146,7 +153,7 @@ func (m *AuthMiddleware) checkJWTFromRequest(r *http.Request, authHeaderParts []
 
 	userAuth, err = m.authManager.EnsureUserAccessByJWTGroups(ctx, userAuth, validatedToken)
 	if err != nil {
-		return err
+		return r, err
 	}
 
 	err = m.syncUserJWTGroups(ctx, userAuth)
@@ -157,41 +164,41 @@ func (m *AuthMiddleware) checkJWTFromRequest(r *http.Request, authHeaderParts []
 	_, err = m.getUserFromUserAuth(ctx, userAuth)
 	if err != nil {
 		log.WithContext(ctx).Errorf("HTTP server failed to update user from user auth: %s", err)
-		return err
+		return r, err
 	}
 
-	// propagates ctx change to upstream middleware
-	*r = *nbcontext.SetUserAuthInRequest(r, userAuth)
-	return nil
+	return nbcontext.SetUserAuthInRequest(r, userAuth), nil
 }
 
 // CheckPATFromRequest checks if the PAT is valid
-func (m *AuthMiddleware) checkPATFromRequest(r *http.Request, authHeaderParts []string) error {
+func (m *AuthMiddleware) checkPATFromRequest(r *http.Request, authHeaderParts []string) (*http.Request, error) {
 	token, err := getTokenFromPATRequest(authHeaderParts)
 	if err != nil {
-		return fmt.Errorf("error extracting token: %w", err)
+		return r, fmt.Errorf("error extracting token: %w", err)
 	}
 
 	if m.patUsageTracker != nil {
 		m.patUsageTracker.IncrementUsage(token)
 	}
 
-	if !isTerraformRequest(r) && !m.rateLimiter.Allow(token) {
-		return status.Errorf(status.TooManyRequests, "too many requests")
+	if m.rateLimiter != nil && !isTerraformRequest(r) {
+		if !m.rateLimiter.Allow(token) {
+			return r, status.Errorf(status.TooManyRequests, "too many requests")
+		}
 	}
 
 	ctx := r.Context()
 	user, pat, accDomain, accCategory, err := m.authManager.GetPATInfo(ctx, token)
 	if err != nil {
-		return fmt.Errorf("invalid Token: %w", err)
+		return r, fmt.Errorf("invalid Token: %w", err)
 	}
 	if time.Now().After(pat.GetExpirationDate()) {
-		return fmt.Errorf("token expired")
+		return r, fmt.Errorf("token expired")
 	}
 
 	err = m.authManager.MarkPATUsed(ctx, pat.ID)
 	if err != nil {
-		return err
+		return r, err
 	}
 
 	userAuth := auth.UserAuth{
@@ -209,9 +216,7 @@ func (m *AuthMiddleware) checkPATFromRequest(r *http.Request, authHeaderParts []
 		}
 	}
 
-	// propagates ctx change to upstream middleware
-	*r = *nbcontext.SetUserAuthInRequest(r, userAuth)
-	return nil
+	return nbcontext.SetUserAuthInRequest(r, userAuth), nil
 }
 
 func isTerraformRequest(r *http.Request) bool {

management/server/http/middleware/auth_middleware_test.go

@@ -196,8 +196,6 @@ func TestAuthMiddleware_Handler(t *testing.T) {
 		GetPATInfoFunc:                  mockGetAccountInfoFromPAT,
 	}
 
-	disabledLimiter := NewAPIRateLimiter(nil)
-	disabledLimiter.SetEnabled(false)
 	authMiddleware := NewAuthMiddleware(
 		mockAuth,
 		func(ctx context.Context, userAuth nbauth.UserAuth) (string, string, error) {
@@ -209,7 +207,7 @@ func TestAuthMiddleware_Handler(t *testing.T) {
 		func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 			return &types.User{}, nil
 		},
-		disabledLimiter,
+		nil,
 		nil,
 	)
 
@@ -268,7 +266,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -320,7 +318,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -363,7 +361,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -407,7 +405,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -471,7 +469,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -530,7 +528,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -585,7 +583,7 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 				return &types.User{}, nil
 			},
-			NewAPIRateLimiter(rateLimitConfig),
+			rateLimitConfig,
 			nil,
 		)
 
@@ -672,8 +670,6 @@ func TestAuthMiddleware_Handler_Child(t *testing.T) {
 		GetPATInfoFunc:                  mockGetAccountInfoFromPAT,
 	}
 
-	disabledLimiter := NewAPIRateLimiter(nil)
-	disabledLimiter.SetEnabled(false)
 	authMiddleware := NewAuthMiddleware(
 		mockAuth,
 		func(ctx context.Context, userAuth nbauth.UserAuth) (string, string, error) {
@@ -685,7 +681,7 @@ func TestAuthMiddleware_Handler_Child(t *testing.T) {
 		func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
 			return &types.User{}, nil
 		},
-		disabledLimiter,
+		nil,
 		nil,
 	)
 

management/server/http/middleware/rate_limiter.go

@@ -4,27 +4,14 @@ import (
 	"context"
 	"net"
 	"net/http"
-	"os"
-	"strconv"
 	"sync"
-	"sync/atomic"
 	"time"
 
-	log "github.com/sirupsen/logrus"
 	"golang.org/x/time/rate"
 
 	"github.com/netbirdio/netbird/shared/management/http/util"
 )
 
-const (
-	RateLimitingEnabledEnv = "NB_API_RATE_LIMITING_ENABLED"
-	RateLimitingBurstEnv   = "NB_API_RATE_LIMITING_BURST"
-	RateLimitingRPMEnv     = "NB_API_RATE_LIMITING_RPM"
-
-	defaultAPIRPM   = 6
-	defaultAPIBurst = 500
-)
-
 // RateLimiterConfig holds configuration for the API rate limiter
 type RateLimiterConfig struct {
 	// RequestsPerMinute defines the rate at which tokens are replenished
@@ -47,43 +34,6 @@ func DefaultRateLimiterConfig() *RateLimiterConfig {
 	}
 }
 
-func RateLimiterConfigFromEnv() (cfg *RateLimiterConfig, enabled bool) {
-	rpm := defaultAPIRPM
-	if v := os.Getenv(RateLimitingRPMEnv); v != "" {
-		value, err := strconv.Atoi(v)
-		if err != nil {
-			log.Warnf("parsing %s env var: %v, using default %d", RateLimitingRPMEnv, err, rpm)
-		} else {
-			rpm = value
-		}
-	}
-	if rpm <= 0 {
-		log.Warnf("%s=%d is non-positive, using default %d", RateLimitingRPMEnv, rpm, defaultAPIRPM)
-		rpm = defaultAPIRPM
-	}
-
-	burst := defaultAPIBurst
-	if v := os.Getenv(RateLimitingBurstEnv); v != "" {
-		value, err := strconv.Atoi(v)
-		if err != nil {
-			log.Warnf("parsing %s env var: %v, using default %d", RateLimitingBurstEnv, err, burst)
-		} else {
-			burst = value
-		}
-	}
-	if burst <= 0 {
-		log.Warnf("%s=%d is non-positive, using default %d", RateLimitingBurstEnv, burst, defaultAPIBurst)
-		burst = defaultAPIBurst
-	}
-
-	return &RateLimiterConfig{
-		RequestsPerMinute: float64(rpm),
-		Burst:             burst,
-		CleanupInterval:   6 * time.Hour,
-		LimiterTTL:        24 * time.Hour,
-	}, os.Getenv(RateLimitingEnabledEnv) == "true"
-}
-
 // limiterEntry holds a rate limiter and its last access time
 type limiterEntry struct {
 	limiter    *rate.Limiter
@@ -96,7 +46,6 @@ type APIRateLimiter struct {
 	limiters map[string]*limiterEntry
 	mu       sync.RWMutex
 	stopChan chan struct{}
-	enabled  atomic.Bool
 }
 
 // NewAPIRateLimiter creates a new API rate limiter with the given configuration
@@ -110,53 +59,14 @@ func NewAPIRateLimiter(config *RateLimiterConfig) *APIRateLimiter {
 		limiters: make(map[string]*limiterEntry),
 		stopChan: make(chan struct{}),
 	}
-	rl.enabled.Store(true)
 
 	go rl.cleanupLoop()
 
 	return rl
 }
 
-func (rl *APIRateLimiter) SetEnabled(enabled bool) {
-	rl.enabled.Store(enabled)
-}
-
-func (rl *APIRateLimiter) Enabled() bool {
-	return rl.enabled.Load()
-}
-
-func (rl *APIRateLimiter) UpdateConfig(config *RateLimiterConfig) {
-	if config == nil {
-		return
-	}
-	if config.RequestsPerMinute <= 0 || config.Burst <= 0 {
-		log.Warnf("UpdateConfig: ignoring invalid rpm=%v burst=%d", config.RequestsPerMinute, config.Burst)
-		return
-	}
-
-	newRPS := rate.Limit(config.RequestsPerMinute / 60.0)
-	newBurst := config.Burst
-
-	rl.mu.Lock()
-	rl.config.RequestsPerMinute = config.RequestsPerMinute
-	rl.config.Burst = newBurst
-	snapshot := make([]*rate.Limiter, 0, len(rl.limiters))
-	for _, entry := range rl.limiters {
-		snapshot = append(snapshot, entry.limiter)
-	}
-	rl.mu.Unlock()
-
-	for _, l := range snapshot {
-		l.SetLimit(newRPS)
-		l.SetBurst(newBurst)
-	}
-}
-
 // Allow checks if a request for the given key (token) is allowed
 func (rl *APIRateLimiter) Allow(key string) bool {
-	if !rl.enabled.Load() {
-		return true
-	}
 	limiter := rl.getLimiter(key)
 	return limiter.Allow()
 }
@@ -164,9 +74,6 @@ func (rl *APIRateLimiter) Allow(key string) bool {
 // Wait blocks until the rate limiter allows another request for the given key
 // Returns an error if the context is canceled
 func (rl *APIRateLimiter) Wait(ctx context.Context, key string) error {
-	if !rl.enabled.Load() {
-		return nil
-	}
 	limiter := rl.getLimiter(key)
 	return limiter.Wait(ctx)
 }
@@ -246,10 +153,6 @@ func (rl *APIRateLimiter) Reset(key string) {
 // Returns 429 Too Many Requests if the rate limit is exceeded.
 func (rl *APIRateLimiter) Middleware(next http.Handler) http.Handler {
 	return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
-		if !rl.enabled.Load() {
-			next.ServeHTTP(w, r)
-			return
-		}
 		clientIP := getClientIP(r)
 		if !rl.Allow(clientIP) {
 			util.WriteErrorResponse("rate limit exceeded, please try again later", http.StatusTooManyRequests, w)

management/server/http/middleware/rate_limiter_test.go

@@ -1,10 +1,8 @@
 package middleware
 
 import (
-	"fmt"
 	"net/http"
 	"net/http/httptest"
-	"sync"
 	"testing"
 	"time"
 
@@ -158,172 +156,3 @@ func TestAPIRateLimiter_Reset(t *testing.T) {
 	// Should be allowed again
 	assert.True(t, rl.Allow("test-key"))
 }
-
-func TestAPIRateLimiter_SetEnabled(t *testing.T) {
-	rl := NewAPIRateLimiter(&RateLimiterConfig{
-		RequestsPerMinute: 60,
-		Burst:             1,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-	defer rl.Stop()
-
-	assert.True(t, rl.Allow("key"))
-	assert.False(t, rl.Allow("key"), "burst exhausted while enabled")
-
-	rl.SetEnabled(false)
-	assert.False(t, rl.Enabled())
-	for i := 0; i < 5; i++ {
-		assert.True(t, rl.Allow("key"), "disabled limiter must always allow")
-	}
-
-	rl.SetEnabled(true)
-	assert.True(t, rl.Enabled())
-	assert.False(t, rl.Allow("key"), "re-enabled limiter retains prior bucket state")
-}
-
-func TestAPIRateLimiter_UpdateConfig(t *testing.T) {
-	rl := NewAPIRateLimiter(&RateLimiterConfig{
-		RequestsPerMinute: 60,
-		Burst:             2,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-	defer rl.Stop()
-
-	assert.True(t, rl.Allow("k1"))
-	assert.True(t, rl.Allow("k1"))
-	assert.False(t, rl.Allow("k1"), "burst=2 exhausted")
-
-	rl.UpdateConfig(&RateLimiterConfig{
-		RequestsPerMinute: 60,
-		Burst:             10,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-
-	// New burst applies to existing keys in place; bucket refills up to new burst over time,
-	// but importantly newly-added keys use the updated config immediately.
-	assert.True(t, rl.Allow("k2"))
-	for i := 0; i < 9; i++ {
-		assert.True(t, rl.Allow("k2"))
-	}
-	assert.False(t, rl.Allow("k2"), "new burst=10 exhausted")
-}
-
-func TestAPIRateLimiter_UpdateConfig_NilIgnored(t *testing.T) {
-	rl := NewAPIRateLimiter(&RateLimiterConfig{
-		RequestsPerMinute: 60,
-		Burst:             1,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-	defer rl.Stop()
-
-	rl.UpdateConfig(nil) // must not panic or zero the config
-
-	assert.True(t, rl.Allow("k"))
-	assert.False(t, rl.Allow("k"))
-}
-
-func TestAPIRateLimiter_UpdateConfig_NonPositiveIgnored(t *testing.T) {
-	rl := NewAPIRateLimiter(&RateLimiterConfig{
-		RequestsPerMinute: 60,
-		Burst:             1,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-	defer rl.Stop()
-
-	assert.True(t, rl.Allow("k"))
-	assert.False(t, rl.Allow("k"))
-
-	rl.UpdateConfig(&RateLimiterConfig{RequestsPerMinute: 0, Burst: 0, CleanupInterval: time.Minute, LimiterTTL: time.Minute})
-	rl.UpdateConfig(&RateLimiterConfig{RequestsPerMinute: -1, Burst: 5, CleanupInterval: time.Minute, LimiterTTL: time.Minute})
-	rl.UpdateConfig(&RateLimiterConfig{RequestsPerMinute: 60, Burst: -1, CleanupInterval: time.Minute, LimiterTTL: time.Minute})
-
-	rl.Reset("k")
-	assert.True(t, rl.Allow("k"))
-	assert.False(t, rl.Allow("k"), "burst should still be 1 — invalid UpdateConfig calls were ignored")
-}
-
-func TestAPIRateLimiter_ConcurrentAllowAndUpdate(t *testing.T) {
-	rl := NewAPIRateLimiter(&RateLimiterConfig{
-		RequestsPerMinute: 600,
-		Burst:             10,
-		CleanupInterval:   time.Minute,
-		LimiterTTL:        time.Minute,
-	})
-	defer rl.Stop()
-
-	var wg sync.WaitGroup
-	stop := make(chan struct{})
-
-	for i := 0; i < 8; i++ {
-		wg.Add(1)
-		go func(id int) {
-			defer wg.Done()
-			key := fmt.Sprintf("k%d", id)
-			for {
-				select {
-				case <-stop:
-					return
-				default:
-					rl.Allow(key)
-				}
-			}
-		}(i)
-	}
-
-	wg.Add(1)
-	go func() {
-		defer wg.Done()
-		for i := 0; i < 200; i++ {
-			select {
-			case <-stop:
-				return
-			default:
-				rl.UpdateConfig(&RateLimiterConfig{
-					RequestsPerMinute: float64(30 + (i % 90)),
-					Burst:             1 + (i % 20),
-					CleanupInterval:   time.Minute,
-					LimiterTTL:        time.Minute,
-				})
-				rl.SetEnabled(i%2 == 0)
-			}
-		}
-	}()
-
-	time.Sleep(100 * time.Millisecond)
-	close(stop)
-	wg.Wait()
-}
-
-func TestRateLimiterConfigFromEnv(t *testing.T) {
-	t.Setenv(RateLimitingEnabledEnv, "true")
-	t.Setenv(RateLimitingRPMEnv, "42")
-	t.Setenv(RateLimitingBurstEnv, "7")
-
-	cfg, enabled := RateLimiterConfigFromEnv()
-	assert.True(t, enabled)
-	assert.Equal(t, float64(42), cfg.RequestsPerMinute)
-	assert.Equal(t, 7, cfg.Burst)
-
-	t.Setenv(RateLimitingEnabledEnv, "false")
-	_, enabled = RateLimiterConfigFromEnv()
-	assert.False(t, enabled)
-
-	t.Setenv(RateLimitingEnabledEnv, "")
-	t.Setenv(RateLimitingRPMEnv, "")
-	t.Setenv(RateLimitingBurstEnv, "")
-	cfg, enabled = RateLimiterConfigFromEnv()
-	assert.False(t, enabled)
-	assert.Equal(t, float64(defaultAPIRPM), cfg.RequestsPerMinute)
-	assert.Equal(t, defaultAPIBurst, cfg.Burst)
-
-	t.Setenv(RateLimitingRPMEnv, "0")
-	t.Setenv(RateLimitingBurstEnv, "-5")
-	cfg, _ = RateLimiterConfigFromEnv()
-	assert.Equal(t, float64(defaultAPIRPM), cfg.RequestsPerMinute, "non-positive rpm must fall back to default")
-	assert.Equal(t, defaultAPIBurst, cfg.Burst, "non-positive burst must fall back to default")
-}

management/server/http/testing/testing_tools/channel/channel.go

@@ -135,7 +135,7 @@ func BuildApiBlackBoxWithDBState(t testing_tools.TB, sqlFile string, expectedPee
 	customZonesManager := zonesManager.NewManager(store, am, permissionsManager, "")
 	zoneRecordsManager := recordsManager.NewManager(store, am, permissionsManager)
 
-	apiHandler, err := http2.NewAPIHandler(context.Background(), am, networksManager, resourcesManager, routersManager, groupsManager, geoMock, authManagerMock, metrics, validatorMock, proxyController, permissionsManager, peersManager, settingsManager, customZonesManager, zoneRecordsManager, networkMapController, nil, serviceManager, nil, nil, nil, nil, nil)
+	apiHandler, err := http2.NewAPIHandler(context.Background(), am, networksManager, resourcesManager, routersManager, groupsManager, geoMock, authManagerMock, metrics, validatorMock, proxyController, permissionsManager, peersManager, settingsManager, customZonesManager, zoneRecordsManager, networkMapController, nil, serviceManager, nil, nil, nil, nil)
 	if err != nil {
 		t.Fatalf("Failed to create API handler: %v", err)
 	}
@@ -264,7 +264,7 @@ func BuildApiBlackBoxWithDBStateAndPeerChannel(t testing_tools.TB, sqlFile strin
 	customZonesManager := zonesManager.NewManager(store, am, permissionsManager, "")
 	zoneRecordsManager := recordsManager.NewManager(store, am, permissionsManager)
 
-	apiHandler, err := http2.NewAPIHandler(context.Background(), am, networksManager, resourcesManager, routersManager, groupsManager, geoMock, authManagerMock, metrics, validatorMock, proxyController, permissionsManager, peersManager, settingsManager, customZonesManager, zoneRecordsManager, networkMapController, nil, serviceManager, nil, nil, nil, nil, nil)
+	apiHandler, err := http2.NewAPIHandler(context.Background(), am, networksManager, resourcesManager, routersManager, groupsManager, geoMock, authManagerMock, metrics, validatorMock, proxyController, permissionsManager, peersManager, settingsManager, customZonesManager, zoneRecordsManager, networkMapController, nil, serviceManager, nil, nil, nil, nil)
 	if err != nil {
 		t.Fatalf("Failed to create API handler: %v", err)
 	}

management/server/management_proto_test.go

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

management/server/nameserver_test.go

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

management/server/peer.go

@@ -1405,10 +1405,6 @@ 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(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1929,7 +1929,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1994,7 +1994,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2012,7 +2012,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2058,7 +2058,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2076,7 +2076,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2113,7 +2113,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2131,7 +2131,7 @@ func TestPeerAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})

management/server/policy.go

@@ -5,7 +5,6 @@ import (
 	_ "embed"
 
 	"github.com/rs/xid"
-	"github.com/sirupsen/logrus"
 
 	"github.com/netbirdio/netbird/management/server/permissions/modules"
 	"github.com/netbirdio/netbird/management/server/permissions/operations"
@@ -47,40 +46,25 @@ func (am *DefaultAccountManager) SavePolicy(ctx context.Context, accountID, user
 	var isUpdate = policy.ID != ""
 	var updateAccountPeers bool
 	var action = activity.PolicyAdded
-	var unchanged bool
 
 	err = am.Store.ExecuteInTransaction(ctx, func(transaction store.Store) error {
-		existingPolicy, err := validatePolicy(ctx, transaction, accountID, policy)
+		if err = validatePolicy(ctx, transaction, accountID, policy); err != nil {
+			return err
+		}
+
+		updateAccountPeers, err = arePolicyChangesAffectPeers(ctx, transaction, accountID, policy, isUpdate)
 		if err != nil {
 			return err
 		}
 
+		saveFunc := transaction.CreatePolicy
 		if isUpdate {
-			if policy.Equal(existingPolicy) {
-				logrus.WithContext(ctx).Tracef("policy update skipped because equal to stored one - policy id %s", policy.ID)
-				unchanged = true
-				return nil
-			}
-
 			action = activity.PolicyUpdated
+			saveFunc = transaction.SavePolicy
+		}
 
-			updateAccountPeers, err = arePolicyChangesAffectPeersWithExisting(ctx, transaction, policy, existingPolicy)
-			if err != nil {
-				return err
-			}
-
-			if err = transaction.SavePolicy(ctx, policy); err != nil {
-				return err
-			}
-		} else {
-			updateAccountPeers, err = arePolicyChangesAffectPeers(ctx, transaction, policy)
-			if err != nil {
-				return err
-			}
-
-			if err = transaction.CreatePolicy(ctx, policy); err != nil {
-				return err
-			}
+		if err = saveFunc(ctx, policy); err != nil {
+			return err
 		}
 
 		return transaction.IncrementNetworkSerial(ctx, accountID)
@@ -89,10 +73,6 @@ func (am *DefaultAccountManager) SavePolicy(ctx context.Context, accountID, user
 		return nil, err
 	}
 
-	if unchanged {
-		return policy, nil
-	}
-
 	am.StoreEvent(ctx, userID, policy.ID, accountID, action, policy.EventMeta())
 
 	if updateAccountPeers {
@@ -121,7 +101,7 @@ func (am *DefaultAccountManager) DeletePolicy(ctx context.Context, accountID, po
 			return err
 		}
 
-		updateAccountPeers, err = arePolicyChangesAffectPeers(ctx, transaction, policy)
+		updateAccountPeers, err = arePolicyChangesAffectPeers(ctx, transaction, accountID, policy, false)
 		if err != nil {
 			return err
 		}
@@ -158,55 +138,49 @@ func (am *DefaultAccountManager) ListPolicies(ctx context.Context, accountID, us
 	return am.Store.GetAccountPolicies(ctx, store.LockingStrengthNone, accountID)
 }
 
-// arePolicyChangesAffectPeers checks if a policy (being created or deleted) will affect any associated peers.
-func arePolicyChangesAffectPeers(ctx context.Context, transaction store.Store, policy *types.Policy) (bool, error) {
-	for _, rule := range policy.Rules {
-		if rule.SourceResource.Type != "" || rule.DestinationResource.Type != "" {
-			return true, nil
-		}
-	}
-
-	return anyGroupHasPeersOrResources(ctx, transaction, policy.AccountID, policy.RuleGroups())
-}
-
-func arePolicyChangesAffectPeersWithExisting(ctx context.Context, transaction store.Store, policy *types.Policy, existingPolicy *types.Policy) (bool, error) {
-	if !policy.Enabled && !existingPolicy.Enabled {
-		return false, nil
-	}
-
-	for _, rule := range existingPolicy.Rules {
-		if rule.SourceResource.Type != "" || rule.DestinationResource.Type != "" {
-			return true, nil
-		}
-	}
-
-	hasPeers, err := anyGroupHasPeersOrResources(ctx, transaction, policy.AccountID, existingPolicy.RuleGroups())
-	if err != nil {
-		return false, err
-	}
-
-	if hasPeers {
-		return true, nil
-	}
-
-	for _, rule := range policy.Rules {
-		if rule.SourceResource.Type != "" || rule.DestinationResource.Type != "" {
-			return true, nil
-		}
-	}
-
-	return anyGroupHasPeersOrResources(ctx, transaction, policy.AccountID, policy.RuleGroups())
-}
-
-// validatePolicy validates the policy and its rules. For updates it returns
-// the existing policy loaded from the store so callers can avoid a second read.
-func validatePolicy(ctx context.Context, transaction store.Store, accountID string, policy *types.Policy) (*types.Policy, error) {
-	var existingPolicy *types.Policy
-	if policy.ID != "" {
-		var err error
-		existingPolicy, err = transaction.GetPolicyByID(ctx, store.LockingStrengthNone, accountID, policy.ID)
+// arePolicyChangesAffectPeers checks if changes to a policy will affect any associated peers.
+func arePolicyChangesAffectPeers(ctx context.Context, transaction store.Store, accountID string, policy *types.Policy, isUpdate bool) (bool, error) {
+	if isUpdate {
+		existingPolicy, err := transaction.GetPolicyByID(ctx, store.LockingStrengthNone, accountID, policy.ID)
 		if err != nil {
-			return nil, err
+			return false, err
+		}
+
+		if !policy.Enabled && !existingPolicy.Enabled {
+			return false, nil
+		}
+
+		for _, rule := range existingPolicy.Rules {
+			if rule.SourceResource.Type != "" || rule.DestinationResource.Type != "" {
+				return true, nil
+			}
+		}
+
+		hasPeers, err := anyGroupHasPeersOrResources(ctx, transaction, policy.AccountID, existingPolicy.RuleGroups())
+		if err != nil {
+			return false, err
+		}
+
+		if hasPeers {
+			return true, nil
+		}
+	}
+
+	for _, rule := range policy.Rules {
+		if rule.SourceResource.Type != "" || rule.DestinationResource.Type != "" {
+			return true, nil
+		}
+	}
+
+	return anyGroupHasPeersOrResources(ctx, transaction, policy.AccountID, policy.RuleGroups())
+}
+
+// validatePolicy validates the policy and its rules.
+func validatePolicy(ctx context.Context, transaction store.Store, accountID string, policy *types.Policy) error {
+	if policy.ID != "" {
+		existingPolicy, err := transaction.GetPolicyByID(ctx, store.LockingStrengthNone, accountID, policy.ID)
+		if err != nil {
+			return err
 		}
 
 		// TODO: Refactor to support multiple rules per policy
@@ -217,7 +191,7 @@ func validatePolicy(ctx context.Context, transaction store.Store, accountID stri
 
 		for _, rule := range policy.Rules {
 			if rule.ID != "" && !existingRuleIDs[rule.ID] {
-				return nil, status.Errorf(status.InvalidArgument, "invalid rule ID: %s", rule.ID)
+				return status.Errorf(status.InvalidArgument, "invalid rule ID: %s", rule.ID)
 			}
 		}
 	} else {
@@ -227,12 +201,12 @@ func validatePolicy(ctx context.Context, transaction store.Store, accountID stri
 
 	groups, err := transaction.GetGroupsByIDs(ctx, store.LockingStrengthNone, accountID, policy.RuleGroups())
 	if err != nil {
-		return nil, err
+		return err
 	}
 
 	postureChecks, err := transaction.GetPostureChecksByIDs(ctx, store.LockingStrengthNone, accountID, policy.SourcePostureChecks)
 	if err != nil {
-		return nil, err
+		return err
 	}
 
 	for i, rule := range policy.Rules {
@@ -251,7 +225,7 @@ func validatePolicy(ctx context.Context, transaction store.Store, accountID stri
 		policy.SourcePostureChecks = getValidPostureCheckIDs(postureChecks, policy.SourcePostureChecks)
 	}
 
-	return existingPolicy, nil
+	return nil
 }
 
 // getValidPostureCheckIDs filters and returns only the valid posture check IDs from the provided list.

management/server/policy_test.go

@@ -1231,7 +1231,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1263,7 +1263,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1294,7 +1294,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1314,7 +1314,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1355,7 +1355,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -1373,7 +1373,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 
@@ -1393,7 +1393,7 @@ func TestPolicyAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			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(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -273,7 +273,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -292,7 +292,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -395,7 +395,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -438,7 +438,7 @@ func TestPostureCheckAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			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(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 
@@ -2107,7 +2107,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2127,7 +2127,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2145,7 +2145,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2185,7 +2185,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			t.Error("timeout waiting for peerShouldReceiveUpdate")
 		}
 	})
@@ -2225,7 +2225,7 @@ func TestRouteAccountPeersUpdate(t *testing.T) {
 
 		select {
 		case <-done:
-		case <-time.After(peerUpdateTimeout):
+		case <-time.After(time.Second):
 			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 peer_status_login_expired != ? AND user_id IS NOT NULL AND user_id != ''", true, true).
+		Where("login_expiration_enabled = ? AND user_id IS NOT NULL AND user_id != ''", 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: 5,
+			expectedCount: 4,
 		},
 		{
 			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: 4,
+			expectedCount: 3,
 		},
 		{
 			name:          "should filter peers by ip",
@@ -2777,16 +2777,14 @@ func TestSqlStore_GetAccountPeersWithExpiration(t *testing.T) {
 	require.NoError(t, err)
 
 	tests := []struct {
-		name            string
-		accountID       string
-		expectedCount   int
-		expectedPeerIDs []string
+		name          string
+		accountID     string
+		expectedCount int
 	}{
 		{
-			name:            "should retrieve only non-expired peers with expiration enabled",
-			accountID:       "bf1c8084-ba50-4ce7-9439-34653001fc3b",
-			expectedCount:   1,
-			expectedPeerIDs: []string{"notexpired01"},
+			name:          "should retrieve peers with expiration for an existing account ID",
+			accountID:     "bf1c8084-ba50-4ce7-9439-34653001fc3b",
+			expectedCount: 1,
 		},
 		{
 			name:          "should return no peers with expiration for a non-existing account ID",
@@ -2805,30 +2803,10 @@ 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)
@@ -2909,7 +2887,7 @@ 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: 3,
+			expectedCount: 2,
 		},
 		{
 			name:          "should return no peers for existing account ID with empty user ID",

management/server/telemetry/http_api_metrics.go

@@ -193,12 +193,20 @@ func (m *HTTPMiddleware) Handler(h http.Handler) http.Handler {
 			}
 		})
 
-		// Hold on to req so auth's in-place ctx update is visible after ServeHTTP.
-		req := r.WithContext(ctx)
-		h.ServeHTTP(w, req)
+		h.ServeHTTP(w, r.WithContext(ctx))
 		close(handlerDone)
 
-		ctx = req.Context()
+		userAuth, err := nbContext.GetUserAuthFromContext(r.Context())
+		if err == nil {
+			if userAuth.AccountId != "" {
+				//nolint
+				ctx = context.WithValue(ctx, nbContext.AccountIDKey, userAuth.AccountId)
+			}
+			if userAuth.UserId != "" {
+				//nolint
+				ctx = context.WithValue(ctx, nbContext.UserIDKey, userAuth.UserId)
+			}
+		}
 
 		if w.Status() > 399 {
 			log.WithContext(ctx).Errorf("HTTP response %v: %v %v status %v", reqID, r.Method, r.URL, w.Status())

management/server/testdata/store_with_expired_peers.sql

@@ -31,7 +31,6 @@ 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/types/policy.go

@@ -93,44 +93,6 @@ func (p *Policy) Copy() *Policy {
 	return c
 }
 
-func (p *Policy) Equal(other *Policy) bool {
-	if p == nil || other == nil {
-		return p == other
-	}
-
-	if p.ID != other.ID ||
-		p.AccountID != other.AccountID ||
-		p.Name != other.Name ||
-		p.Description != other.Description ||
-		p.Enabled != other.Enabled {
-		return false
-	}
-
-	if !stringSlicesEqualUnordered(p.SourcePostureChecks, other.SourcePostureChecks) {
-		return false
-	}
-
-	if len(p.Rules) != len(other.Rules) {
-		return false
-	}
-
-	otherRules := make(map[string]*PolicyRule, len(other.Rules))
-	for _, r := range other.Rules {
-		otherRules[r.ID] = r
-	}
-	for _, r := range p.Rules {
-		otherRule, ok := otherRules[r.ID]
-		if !ok {
-			return false
-		}
-		if !r.Equal(otherRule) {
-			return false
-		}
-	}
-
-	return true
-}
-
 // EventMeta returns activity event meta related to this policy
 func (p *Policy) EventMeta() map[string]any {
 	return map[string]any{"name": p.Name}

management/server/types/policy_test.go

@@ -1,193 +0,0 @@
-package types
-
-import (
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-)
-
-func TestPolicyEqual_SameRulesDifferentOrder(t *testing.T) {
-	a := &Policy{
-		ID:        "pol1",
-		AccountID: "acc1",
-		Name:      "test",
-		Enabled:   true,
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1", Ports: []string{"80"}},
-			{ID: "r2", PolicyID: "pol1", Ports: []string{"443"}},
-		},
-	}
-	b := &Policy{
-		ID:        "pol1",
-		AccountID: "acc1",
-		Name:      "test",
-		Enabled:   true,
-		Rules: []*PolicyRule{
-			{ID: "r2", PolicyID: "pol1", Ports: []string{"443"}},
-			{ID: "r1", PolicyID: "pol1", Ports: []string{"80"}},
-		},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyEqual_DifferentRules(t *testing.T) {
-	a := &Policy{
-		ID:      "pol1",
-		Enabled: true,
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1", Ports: []string{"80"}},
-		},
-	}
-	b := &Policy{
-		ID:      "pol1",
-		Enabled: true,
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1", Ports: []string{"443"}},
-		},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyEqual_DifferentRuleCount(t *testing.T) {
-	a := &Policy{
-		ID: "pol1",
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1"},
-		},
-	}
-	b := &Policy{
-		ID: "pol1",
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1"},
-			{ID: "r2", PolicyID: "pol1"},
-		},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyEqual_PostureChecksDifferentOrder(t *testing.T) {
-	a := &Policy{
-		ID:                  "pol1",
-		SourcePostureChecks: []string{"pc3", "pc1", "pc2"},
-	}
-	b := &Policy{
-		ID:                  "pol1",
-		SourcePostureChecks: []string{"pc1", "pc2", "pc3"},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyEqual_DifferentPostureChecks(t *testing.T) {
-	a := &Policy{
-		ID:                  "pol1",
-		SourcePostureChecks: []string{"pc1", "pc2"},
-	}
-	b := &Policy{
-		ID:                  "pol1",
-		SourcePostureChecks: []string{"pc1", "pc3"},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyEqual_DifferentScalarFields(t *testing.T) {
-	base := Policy{
-		ID:          "pol1",
-		AccountID:   "acc1",
-		Name:        "test",
-		Description: "desc",
-		Enabled:     true,
-	}
-
-	other := base
-	other.Name = "changed"
-	assert.False(t, base.Equal(&other))
-
-	other = base
-	other.Enabled = false
-	assert.False(t, base.Equal(&other))
-
-	other = base
-	other.Description = "changed"
-	assert.False(t, base.Equal(&other))
-}
-
-func TestPolicyEqual_NilCases(t *testing.T) {
-	var a *Policy
-	var b *Policy
-	assert.True(t, a.Equal(b))
-
-	a = &Policy{ID: "pol1"}
-	assert.False(t, a.Equal(nil))
-}
-
-func TestPolicyEqual_RulesMismatchByID(t *testing.T) {
-	a := &Policy{
-		ID: "pol1",
-		Rules: []*PolicyRule{
-			{ID: "r1", PolicyID: "pol1"},
-		},
-	}
-	b := &Policy{
-		ID: "pol1",
-		Rules: []*PolicyRule{
-			{ID: "r2", PolicyID: "pol1"},
-		},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyEqual_FullScenario(t *testing.T) {
-	a := &Policy{
-		ID:                  "pol1",
-		AccountID:           "acc1",
-		Name:                "Web Access",
-		Description:         "Allow web access",
-		Enabled:             true,
-		SourcePostureChecks: []string{"pc2", "pc1"},
-		Rules: []*PolicyRule{
-			{
-				ID:            "r1",
-				PolicyID:      "pol1",
-				Name:          "HTTP",
-				Enabled:       true,
-				Action:        PolicyTrafficActionAccept,
-				Protocol:      PolicyRuleProtocolTCP,
-				Bidirectional: true,
-				Sources:       []string{"g2", "g1"},
-				Destinations:  []string{"g4", "g3"},
-				Ports:         []string{"443", "80", "8080"},
-				PortRanges: []RulePortRange{
-					{Start: 8000, End: 9000},
-					{Start: 80, End: 80},
-				},
-			},
-		},
-	}
-	b := &Policy{
-		ID:                  "pol1",
-		AccountID:           "acc1",
-		Name:                "Web Access",
-		Description:         "Allow web access",
-		Enabled:             true,
-		SourcePostureChecks: []string{"pc1", "pc2"},
-		Rules: []*PolicyRule{
-			{
-				ID:            "r1",
-				PolicyID:      "pol1",
-				Name:          "HTTP",
-				Enabled:       true,
-				Action:        PolicyTrafficActionAccept,
-				Protocol:      PolicyRuleProtocolTCP,
-				Bidirectional: true,
-				Sources:       []string{"g1", "g2"},
-				Destinations:  []string{"g3", "g4"},
-				Ports:         []string{"80", "8080", "443"},
-				PortRanges: []RulePortRange{
-					{Start: 80, End: 80},
-					{Start: 8000, End: 9000},
-				},
-			},
-		},
-	}
-	assert.True(t, a.Equal(b))
-}

management/server/types/policyrule.go

@@ -1,8 +1,6 @@
 package types
 
 import (
-	"slices"
-
 	"github.com/netbirdio/netbird/shared/management/proto"
 )
 
@@ -120,106 +118,3 @@ func (pm *PolicyRule) Copy() *PolicyRule {
 	}
 	return rule
 }
-
-func (pm *PolicyRule) Equal(other *PolicyRule) bool {
-	if pm == nil || other == nil {
-		return pm == other
-	}
-
-	if pm.ID != other.ID ||
-		pm.PolicyID != other.PolicyID ||
-		pm.Name != other.Name ||
-		pm.Description != other.Description ||
-		pm.Enabled != other.Enabled ||
-		pm.Action != other.Action ||
-		pm.Bidirectional != other.Bidirectional ||
-		pm.Protocol != other.Protocol ||
-		pm.SourceResource != other.SourceResource ||
-		pm.DestinationResource != other.DestinationResource ||
-		pm.AuthorizedUser != other.AuthorizedUser {
-		return false
-	}
-
-	if !stringSlicesEqualUnordered(pm.Sources, other.Sources) {
-		return false
-	}
-	if !stringSlicesEqualUnordered(pm.Destinations, other.Destinations) {
-		return false
-	}
-	if !stringSlicesEqualUnordered(pm.Ports, other.Ports) {
-		return false
-	}
-	if !portRangeSlicesEqualUnordered(pm.PortRanges, other.PortRanges) {
-		return false
-	}
-	if !authorizedGroupsEqual(pm.AuthorizedGroups, other.AuthorizedGroups) {
-		return false
-	}
-
-	return true
-}
-
-func stringSlicesEqualUnordered(a, b []string) bool {
-	if len(a) != len(b) {
-		return false
-	}
-	if len(a) == 0 {
-		return true
-	}
-	sorted1 := make([]string, len(a))
-	sorted2 := make([]string, len(b))
-	copy(sorted1, a)
-	copy(sorted2, b)
-	slices.Sort(sorted1)
-	slices.Sort(sorted2)
-	return slices.Equal(sorted1, sorted2)
-}
-
-func portRangeSlicesEqualUnordered(a, b []RulePortRange) bool {
-	if len(a) != len(b) {
-		return false
-	}
-	if len(a) == 0 {
-		return true
-	}
-	cmp := func(x, y RulePortRange) int {
-		if x.Start != y.Start {
-			if x.Start < y.Start {
-				return -1
-			}
-			return 1
-		}
-		if x.End != y.End {
-			if x.End < y.End {
-				return -1
-			}
-			return 1
-		}
-		return 0
-	}
-	sorted1 := make([]RulePortRange, len(a))
-	sorted2 := make([]RulePortRange, len(b))
-	copy(sorted1, a)
-	copy(sorted2, b)
-	slices.SortFunc(sorted1, cmp)
-	slices.SortFunc(sorted2, cmp)
-	return slices.EqualFunc(sorted1, sorted2, func(x, y RulePortRange) bool {
-		return x.Start == y.Start && x.End == y.End
-	})
-}
-
-func authorizedGroupsEqual(a, b map[string][]string) bool {
-	if len(a) != len(b) {
-		return false
-	}
-	for k, va := range a {
-		vb, ok := b[k]
-		if !ok {
-			return false
-		}
-		if !stringSlicesEqualUnordered(va, vb) {
-			return false
-		}
-	}
-	return true
-}

management/server/types/policyrule_test.go

@@ -1,194 +0,0 @@
-package types
-
-import (
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-)
-
-func TestPolicyRuleEqual_SamePortsDifferentOrder(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    []string{"443", "80", "22"},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    []string{"22", "443", "80"},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_DifferentPorts(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    []string{"443", "80"},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    []string{"443", "22"},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_SourcesDestinationsDifferentOrder(t *testing.T) {
-	a := &PolicyRule{
-		ID:           "rule1",
-		PolicyID:     "pol1",
-		Sources:      []string{"g1", "g2", "g3"},
-		Destinations: []string{"g4", "g5"},
-	}
-	b := &PolicyRule{
-		ID:           "rule1",
-		PolicyID:     "pol1",
-		Sources:      []string{"g3", "g1", "g2"},
-		Destinations: []string{"g5", "g4"},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_DifferentSources(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Sources:  []string{"g1", "g2"},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Sources:  []string{"g1", "g3"},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_PortRangesDifferentOrder(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		PortRanges: []RulePortRange{
-			{Start: 8000, End: 9000},
-			{Start: 80, End: 80},
-		},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		PortRanges: []RulePortRange{
-			{Start: 80, End: 80},
-			{Start: 8000, End: 9000},
-		},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_DifferentPortRanges(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		PortRanges: []RulePortRange{
-			{Start: 80, End: 80},
-		},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		PortRanges: []RulePortRange{
-			{Start: 80, End: 443},
-		},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_AuthorizedGroupsDifferentValueOrder(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		AuthorizedGroups: map[string][]string{
-			"g1": {"u1", "u2", "u3"},
-		},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		AuthorizedGroups: map[string][]string{
-			"g1": {"u3", "u1", "u2"},
-		},
-	}
-	assert.True(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_DifferentAuthorizedGroups(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		AuthorizedGroups: map[string][]string{
-			"g1": {"u1"},
-		},
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		AuthorizedGroups: map[string][]string{
-			"g2": {"u1"},
-		},
-	}
-	assert.False(t, a.Equal(b))
-}
-
-func TestPolicyRuleEqual_DifferentScalarFields(t *testing.T) {
-	base := PolicyRule{
-		ID:            "rule1",
-		PolicyID:      "pol1",
-		Name:          "test",
-		Description:   "desc",
-		Enabled:       true,
-		Action:        PolicyTrafficActionAccept,
-		Bidirectional: true,
-		Protocol:      PolicyRuleProtocolTCP,
-	}
-
-	other := base
-	other.Name = "changed"
-	assert.False(t, base.Equal(&other))
-
-	other = base
-	other.Enabled = false
-	assert.False(t, base.Equal(&other))
-
-	other = base
-	other.Action = PolicyTrafficActionDrop
-	assert.False(t, base.Equal(&other))
-
-	other = base
-	other.Protocol = PolicyRuleProtocolUDP
-	assert.False(t, base.Equal(&other))
-}
-
-func TestPolicyRuleEqual_NilCases(t *testing.T) {
-	var a *PolicyRule
-	var b *PolicyRule
-	assert.True(t, a.Equal(b))
-
-	a = &PolicyRule{ID: "rule1"}
-	assert.False(t, a.Equal(nil))
-}
-
-func TestPolicyRuleEqual_EmptySlices(t *testing.T) {
-	a := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    []string{},
-		Sources:  nil,
-	}
-	b := &PolicyRule{
-		ID:       "rule1",
-		PolicyID: "pol1",
-		Ports:    nil,
-		Sources:  []string{},
-	}
-	assert.True(t, a.Equal(b))
-}
-

management/server/user_test.go

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

proxy/internal/auth/middleware.go

@@ -433,7 +433,6 @@ 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

@@ -391,15 +391,6 @@ 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/management/client/grpc.go

@@ -30,8 +30,6 @@ import (
 
 const ConnectTimeout = 10 * time.Second
 
-const healthCheckTimeout = 5 * time.Second
-
 const (
 	// EnvMaxRecvMsgSize overrides the default gRPC max receive message size (4 MB)
 	// for the management client connection. Value is in bytes.
@@ -534,7 +532,7 @@ func (c *GrpcClient) IsHealthy() bool {
 	case connectivity.Ready:
 	}
 
-	ctx, cancel := context.WithTimeout(c.ctx, healthCheckTimeout)
+	ctx, cancel := context.WithTimeout(c.ctx, 1*time.Second)
 	defer cancel()
 
 	_, err := c.realClient.GetServerKey(ctx, &proto.Empty{})

shared/signal/client/grpc.go

@@ -23,8 +23,6 @@ import (
 	"github.com/netbirdio/netbird/util/wsproxy"
 )
 
-const healthCheckTimeout = 5 * time.Second
-
 // ConnStateNotifier is a wrapper interface of the status recorder
 type ConnStateNotifier interface {
 	MarkSignalDisconnected(error)
@@ -265,7 +263,7 @@ func (c *GrpcClient) IsHealthy() bool {
 	case connectivity.Ready:
 	}
 
-	ctx, cancel := context.WithTimeout(c.ctx, healthCheckTimeout)
+	ctx, cancel := context.WithTimeout(c.ctx, 1*time.Second)
 	defer cancel()
 	_, err := c.realClient.Send(ctx, &proto.EncryptedMessage{
 		Key:       c.key.PublicKey().String(),