Enable trace logging in WASM client

* Change priority between local and dns route handlers

* update priority tests

README.md

@@ -38,6 +38,11 @@
  
 </strong>
 <br>
+<strong>
+  🚀 <a href="https://careers.netbird.io">We are hiring! Join us at careers.netbird.io</a>
+</strong>
+<br>
+<br>
 <a href="https://registry.terraform.io/providers/netbirdio/netbird/latest">
     New: NetBird terraform provider
   </a> 

client/cmd/debug.go

@@ -314,9 +314,8 @@ func getStatusOutput(cmd *cobra.Command, anon bool) string {
 			profName = activeProf.Name
 		}
 
-		statusOutputString = nbstatus.ParseToFullDetailSummary(
-			nbstatus.ConvertToStatusOutputOverview(statusResp, anon, "", nil, nil, nil, "", profName),
-		)
+		overview := nbstatus.ConvertToStatusOutputOverview(statusResp, anon, "", nil, nil, nil, "", profName)
+		statusOutputString = overview.FullDetailSummary()
 	}
 	return statusOutputString
 }

client/cmd/status.go

@@ -103,13 +103,13 @@ func statusFunc(cmd *cobra.Command, args []string) error {
 	var statusOutputString string
 	switch {
 	case detailFlag:
-		statusOutputString = nbstatus.ParseToFullDetailSummary(outputInformationHolder)
+		statusOutputString = outputInformationHolder.FullDetailSummary()
 	case jsonFlag:
-		statusOutputString, err = nbstatus.ParseToJSON(outputInformationHolder)
+		statusOutputString, err = outputInformationHolder.JSON()
 	case yamlFlag:
-		statusOutputString, err = nbstatus.ParseToYAML(outputInformationHolder)
+		statusOutputString, err = outputInformationHolder.YAML()
 	default:
-		statusOutputString = nbstatus.ParseGeneralSummary(outputInformationHolder, false, false, false, false)
+		statusOutputString = outputInformationHolder.GeneralSummary(false, false, false, false)
 	}
 
 	if err != nil {

client/embed/embed.go

@@ -10,6 +10,7 @@ import (
 	"net/netip"
 	"os"
 	"sync"
+	"time"
 
 	"github.com/sirupsen/logrus"
 	wgnetstack "golang.zx2c4.com/wireguard/tun/netstack"
@@ -20,6 +21,7 @@ import (
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
 	sshcommon "github.com/netbirdio/netbird/client/ssh"
 	"github.com/netbirdio/netbird/client/system"
+	mgmProto "github.com/netbirdio/netbird/shared/management/proto"
 )
 
 var (
@@ -29,6 +31,11 @@ var (
 	ErrConfigNotInitialized = errors.New("config not initialized")
 )
 
+const (
+	defaultPeerConnectionTimeout = 60 * time.Second
+	peerConnectionPollInterval   = 500 * time.Millisecond
+)
+
 // Client manages a netbird embedded client instance.
 type Client struct {
 	deviceName string
@@ -38,6 +45,7 @@ type Client struct {
 	setupKey   string
 	jwtToken   string
 	connect    *internal.ConnectClient
+	recorder   *peer.Status
 }
 
 // Options configures a new Client.
@@ -161,11 +169,17 @@ func New(opts Options) (*Client, error) {
 func (c *Client) Start(startCtx context.Context) error {
 	c.mu.Lock()
 	defer c.mu.Unlock()
-	if c.cancel != nil {
+	if c.connect != nil {
 		return ErrClientAlreadyStarted
 	}
 
-	ctx := internal.CtxInitState(context.Background())
+	ctx, cancel := context.WithCancel(internal.CtxInitState(context.Background()))
+	defer func() {
+		if c.connect == nil {
+			cancel()
+		}
+	}()
+
 	// nolint:staticcheck
 	ctx = context.WithValue(ctx, system.DeviceNameCtxKey, c.deviceName)
 	if err := internal.Login(ctx, c.config, c.setupKey, c.jwtToken); err != nil {
@@ -173,7 +187,9 @@ func (c *Client) Start(startCtx context.Context) error {
 	}
 
 	recorder := peer.NewRecorder(c.config.ManagementURL.String())
+	c.recorder = recorder
 	client := internal.NewConnectClient(ctx, c.config, recorder, false)
+	client.SetSyncResponsePersistence(true)
 
 	// either startup error (permanent backoff err) or nil err (successful engine up)
 	// TODO: make after-startup backoff err available
@@ -197,6 +213,7 @@ func (c *Client) Start(startCtx context.Context) error {
 	}
 
 	c.connect = client
+	c.cancel = cancel
 
 	return nil
 }
@@ -211,17 +228,23 @@ func (c *Client) Stop(ctx context.Context) error {
 		return ErrClientNotStarted
 	}
 
+	if c.cancel != nil {
+		c.cancel()
+		c.cancel = nil
+	}
+
 	done := make(chan error, 1)
+	connect := c.connect
 	go func() {
-		done <- c.connect.Stop()
+		done <- connect.Stop()
 	}()
 
 	select {
 	case <-ctx.Done():
-		c.cancel = nil
+		c.connect = nil
 		return ctx.Err()
 	case err := <-done:
-		c.cancel = nil
+		c.connect = nil
 		if err != nil {
 			return fmt.Errorf("stop: %w", err)
 		}
@@ -241,18 +264,40 @@ func (c *Client) GetConfig() (profilemanager.Config, error) {
 
 // Dial dials a network address in the netbird network.
 // Not applicable if the userspace networking mode is disabled.
+// With lazy connections, the connection is established on first traffic.
 func (c *Client) Dial(ctx context.Context, network, address string) (net.Conn, error) {
+	logrus.Infof("embed.Dial called: network=%s, address=%s", network, address)
+
+	// Check context status upfront
+	if ctx.Err() != nil {
+		logrus.Warnf("embed.Dial: context already cancelled/expired: %v", ctx.Err())
+		return nil, ctx.Err()
+	}
+
 	engine, err := c.getEngine()
 	if err != nil {
+		logrus.Errorf("embed.Dial: getEngine failed: %v", err)
 		return nil, err
 	}
 
 	nsnet, err := engine.GetNet()
 	if err != nil {
+		logrus.Errorf("embed.Dial: GetNet failed: %v", err)
 		return nil, fmt.Errorf("get net: %w", err)
 	}
 
-	return nsnet.DialContext(ctx, network, address)
+	// Note: Don't wait for peer connection here - lazy connection manager
+	// will open the connection when DialContext is called. The netstack
+	// dial triggers WireGuard traffic which activates the lazy connection.
+
+	logrus.Debugf("embed.Dial: calling nsnet.DialContext for %s", address)
+	conn, err := nsnet.DialContext(ctx, network, address)
+	if err != nil {
+		logrus.Errorf("embed.Dial: nsnet.DialContext failed: %v", err)
+		return nil, err
+	}
+	logrus.Infof("embed.Dial: successfully connected to %s", address)
+	return conn, nil
 }
 
 // DialContext dials a network address in the netbird network with context
@@ -315,6 +360,90 @@ func (c *Client) NewHTTPClient() *http.Client {
 	}
 }
 
+// Status returns the current status of the client.
+func (c *Client) Status() (peer.FullStatus, error) {
+	c.mu.Lock()
+	recorder := c.recorder
+	connect := c.connect
+	c.mu.Unlock()
+
+	if recorder == nil {
+		return peer.FullStatus{}, errors.New("client not started")
+	}
+
+	if connect != nil {
+		engine := connect.Engine()
+		if engine != nil {
+			_ = engine.RunHealthProbes(false)
+		}
+	}
+
+	return recorder.GetFullStatus(), nil
+}
+
+// GetLatestSyncResponse returns the latest sync response from the management server.
+func (c *Client) GetLatestSyncResponse() (*mgmProto.SyncResponse, error) {
+	engine, err := c.getEngine()
+	if err != nil {
+		return nil, err
+	}
+
+	syncResp, err := engine.GetLatestSyncResponse()
+	if err != nil {
+		return nil, fmt.Errorf("get sync response: %w", err)
+	}
+
+	return syncResp, nil
+}
+
+// WaitForPeerConnection waits for a peer with the given IP to be connected.
+func (c *Client) WaitForPeerConnection(ctx context.Context, peerIP string) error {
+	logrus.Infof("Waiting for peer %s to be connected", peerIP)
+
+	ticker := time.NewTicker(peerConnectionPollInterval)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ctx.Done():
+			return fmt.Errorf("timeout waiting for peer %s to connect: %w", peerIP, ctx.Err())
+		case <-ticker.C:
+			status, err := c.Status()
+			if err != nil {
+				logrus.Debugf("Error getting status while waiting for peer: %v", err)
+				continue
+			}
+
+			for _, p := range status.Peers {
+				if p.IP == peerIP && p.ConnStatus == peer.StatusConnected {
+					logrus.Infof("Peer %s is now connected (relayed: %v)", peerIP, p.Relayed)
+					return nil
+				}
+			}
+			logrus.Tracef("Peer %s not yet connected, waiting...", peerIP)
+		}
+	}
+}
+
+// SetLogLevel sets the logging level for the client and its components.
+func (c *Client) SetLogLevel(levelStr string) error {
+	level, err := logrus.ParseLevel(levelStr)
+	if err != nil {
+		return fmt.Errorf("parse log level: %w", err)
+	}
+
+	logrus.SetLevel(level)
+
+	c.mu.Lock()
+	connect := c.connect
+	c.mu.Unlock()
+
+	// Note: ConnectClient doesn't have SetLogLevel method
+	_ = connect
+
+	return nil
+}
+
 // VerifySSHHostKey verifies an SSH host key against stored peer keys.
 // Returns nil if the key matches, ErrPeerNotFound if peer is not in network,
 // ErrNoStoredKey if peer has no stored key, or an error for verification failures.

client/iface/netstack/dialer.go

@@ -23,10 +23,10 @@ func NewNSDialer(net *netstack.Net) *NSDialer {
 }
 
 func (d *NSDialer) Dial(ctx context.Context, network, addr string) (net.Conn, error) {
-	log.Debugf("dialing %s %s", network, addr)
+	log.Infof("NSDialer.Dial: network=%s, addr=%s", network, addr)
 	conn, err := d.net.Dial(network, addr)
 	if err != nil {
-		log.Debugf("failed to deal connection: %s", err)
+		log.Warnf("NSDialer.Dial failed: %s", err)
 	}
 	return conn, err
 }

client/internal/connect.go

@@ -420,6 +420,19 @@ func (c *ConnectClient) GetLatestSyncResponse() (*mgmProto.SyncResponse, error)
 	return syncResponse, nil
 }
 
+// SetLogLevel sets the log level for the firewall manager if the engine is running.
+func (c *ConnectClient) SetLogLevel(level log.Level) {
+	engine := c.Engine()
+	if engine == nil {
+		return
+	}
+
+	fwManager := engine.GetFirewallManager()
+	if fwManager != nil {
+		fwManager.SetLogLevel(level)
+	}
+}
+
 // Status returns the current client status
 func (c *ConnectClient) Status() StatusType {
 	if c == nil {

client/internal/dns.go

@@ -76,7 +76,7 @@ func collectPTRRecords(config *nbdns.Config, prefix netip.Prefix) []nbdns.Simple
 	var records []nbdns.SimpleRecord
 
 	for _, zone := range config.CustomZones {
-		if zone.SkipPTRProcess {
+		if zone.NonAuthoritative {
 			continue
 		}
 		for _, record := range zone.Records {

client/internal/dns/handler_chain.go

@@ -3,17 +3,21 @@ package dns
 import (
 	"fmt"
 	"slices"
+	"strconv"
 	"strings"
 	"sync"
+	"time"
 
 	"github.com/miekg/dns"
 	log "github.com/sirupsen/logrus"
+
+	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 )
 
 const (
 	PriorityMgmtCache = 150
-	PriorityLocal     = 100
-	PriorityDNSRoute  = 75
+	PriorityDNSRoute  = 100
+	PriorityLocal     = 75
 	PriorityUpstream  = 50
 	PriorityDefault   = 1
 	PriorityFallback  = -100
@@ -43,7 +47,23 @@ type HandlerChain struct {
 type ResponseWriterChain struct {
 	dns.ResponseWriter
 	origPattern    string
+	requestID      string
 	shouldContinue bool
+	response       *dns.Msg
+	meta           map[string]string
+}
+
+// RequestID returns the request ID for tracing
+func (w *ResponseWriterChain) RequestID() string {
+	return w.requestID
+}
+
+// SetMeta sets a metadata key-value pair for logging
+func (w *ResponseWriterChain) SetMeta(key, value string) {
+	if w.meta == nil {
+		w.meta = make(map[string]string)
+	}
+	w.meta[key] = value
 }
 
 func (w *ResponseWriterChain) WriteMsg(m *dns.Msg) error {
@@ -52,6 +72,7 @@ func (w *ResponseWriterChain) WriteMsg(m *dns.Msg) error {
 		w.shouldContinue = true
 		return nil
 	}
+	w.response = m
 	return w.ResponseWriter.WriteMsg(m)
 }
 
@@ -101,6 +122,8 @@ func (c *HandlerChain) AddHandler(pattern string, handler dns.Handler, priority
 
 	pos := c.findHandlerPosition(entry)
 	c.handlers = append(c.handlers[:pos], append([]HandlerEntry{entry}, c.handlers[pos:]...)...)
+
+	c.logHandlers()
 }
 
 // findHandlerPosition determines where to insert a new handler based on priority and specificity
@@ -140,68 +163,109 @@ func (c *HandlerChain) removeEntry(pattern string, priority int) {
 	for i := len(c.handlers) - 1; i >= 0; i-- {
 		entry := c.handlers[i]
 		if strings.EqualFold(entry.OrigPattern, pattern) && entry.Priority == priority {
+			log.Debugf("removing handler pattern: domain=%s priority=%d", entry.OrigPattern, priority)
 			c.handlers = append(c.handlers[:i], c.handlers[i+1:]...)
+			c.logHandlers()
 			break
 		}
 	}
 }
 
+// logHandlers logs the current handler chain state. Caller must hold the lock.
+func (c *HandlerChain) logHandlers() {
+	if !log.IsLevelEnabled(log.TraceLevel) {
+		return
+	}
+
+	var b strings.Builder
+	b.WriteString("handler chain (" + strconv.Itoa(len(c.handlers)) + "):\n")
+	for _, h := range c.handlers {
+		b.WriteString("  - pattern: domain=" + h.Pattern + " original: domain=" + h.OrigPattern +
+			" wildcard=" + strconv.FormatBool(h.IsWildcard) +
+			" match_subdomain=" + strconv.FormatBool(h.MatchSubdomains) +
+			" priority=" + strconv.Itoa(h.Priority) + "\n")
+	}
+	log.Trace(strings.TrimSuffix(b.String(), "\n"))
+}
+
 func (c *HandlerChain) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 	if len(r.Question) == 0 {
 		return
 	}
 
-	qname := strings.ToLower(r.Question[0].Name)
+	startTime := time.Now()
+	requestID := resutil.GenerateRequestID()
+	logger := log.WithFields(log.Fields{
+		"request_id": requestID,
+		"dns_id":     fmt.Sprintf("%04x", r.Id),
+	})
+
+	question := r.Question[0]
+	qname := strings.ToLower(question.Name)
 
 	c.mu.RLock()
 	handlers := slices.Clone(c.handlers)
 	c.mu.RUnlock()
 
-	if log.IsLevelEnabled(log.TraceLevel) {
-		var b strings.Builder
-		b.WriteString(fmt.Sprintf("DNS request domain=%s, handlers (%d):\n", qname, len(handlers)))
-		for _, h := range handlers {
-			b.WriteString(fmt.Sprintf("  - pattern: domain=%s original: domain=%s wildcard=%v match_subdomain=%v priority=%d\n",
-				h.Pattern, h.OrigPattern, h.IsWildcard, h.MatchSubdomains, h.Priority))
-		}
-		log.Trace(strings.TrimSuffix(b.String(), "\n"))
-	}
-
 	// Try handlers in priority order
 	for _, entry := range handlers {
-		matched := c.isHandlerMatch(qname, entry)
-
-		if matched {
-			log.Tracef("handler matched: domain=%s -> pattern=%s wildcard=%v match_subdomain=%v priority=%d",
-				qname, entry.OrigPattern, entry.IsWildcard, entry.MatchSubdomains, entry.Priority)
-
-			chainWriter := &ResponseWriterChain{
-				ResponseWriter: w,
-				origPattern:    entry.OrigPattern,
-			}
-			entry.Handler.ServeDNS(chainWriter, r)
-
-			// If handler wants to continue, try next handler
-			if chainWriter.shouldContinue {
-				// Only log continue for non-management cache handlers to reduce noise
-				if entry.Priority != PriorityMgmtCache {
-					log.Tracef("handler requested continue to next handler for domain=%s", qname)
-				}
-				continue
-			}
-			return
+		if !c.isHandlerMatch(qname, entry) {
+			continue
 		}
+
+		handlerName := entry.OrigPattern
+		if s, ok := entry.Handler.(interface{ String() string }); ok {
+			handlerName = s.String()
+		}
+
+		logger.Tracef("question: domain=%s type=%s class=%s -> handler=%s pattern=%s wildcard=%v match_subdomain=%v priority=%d",
+			qname, dns.TypeToString[question.Qtype], dns.ClassToString[question.Qclass],
+			handlerName, entry.OrigPattern, entry.IsWildcard, entry.MatchSubdomains, entry.Priority)
+
+		chainWriter := &ResponseWriterChain{
+			ResponseWriter: w,
+			origPattern:    entry.OrigPattern,
+			requestID:      requestID,
+		}
+		entry.Handler.ServeDNS(chainWriter, r)
+
+		// If handler wants to continue, try next handler
+		if chainWriter.shouldContinue {
+			if entry.Priority != PriorityMgmtCache {
+				logger.Tracef("handler requested continue for domain=%s", qname)
+			}
+			continue
+		}
+
+		c.logResponse(logger, chainWriter, qname, startTime)
+		return
 	}
 
 	// No handler matched or all handlers passed
-	log.Tracef("no handler found for domain=%s", qname)
+	logger.Tracef("no handler found for domain=%s type=%s class=%s",
+		qname, dns.TypeToString[question.Qtype], dns.ClassToString[question.Qclass])
 	resp := &dns.Msg{}
 	resp.SetRcode(r, dns.RcodeRefused)
 	if err := w.WriteMsg(resp); err != nil {
-		log.Errorf("failed to write DNS response: %v", err)
+		logger.Errorf("failed to write DNS response: %v", err)
 	}
 }
 
+func (c *HandlerChain) logResponse(logger *log.Entry, cw *ResponseWriterChain, qname string, startTime time.Time) {
+	if cw.response == nil {
+		return
+	}
+
+	var meta string
+	for k, v := range cw.meta {
+		meta += " " + k + "=" + v
+	}
+
+	logger.Tracef("response: domain=%s rcode=%s answers=%s%s took=%s",
+		qname, dns.RcodeToString[cw.response.Rcode], resutil.FormatAnswers(cw.response.Answer),
+		meta, time.Since(startTime))
+}
+
 func (c *HandlerChain) isHandlerMatch(qname string, entry HandlerEntry) bool {
 	switch {
 	case entry.Pattern == ".":

client/internal/dns/local/local.go

@@ -1,30 +1,52 @@
 package local
 
 import (
+	"context"
+	"errors"
 	"fmt"
+	"net"
+	"net/netip"
 	"slices"
 	"strings"
 	"sync"
+	"time"
 
 	"github.com/miekg/dns"
 	log "github.com/sirupsen/logrus"
 	"golang.org/x/exp/maps"
 
+	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/client/internal/dns/types"
 	nbdns "github.com/netbirdio/netbird/dns"
 	"github.com/netbirdio/netbird/shared/management/domain"
 )
 
+const externalResolutionTimeout = 4 * time.Second
+
+type resolver interface {
+	LookupNetIP(ctx context.Context, network, host string) ([]netip.Addr, error)
+}
+
 type Resolver struct {
 	mu      sync.RWMutex
 	records map[dns.Question][]dns.RR
 	domains map[domain.Domain]struct{}
+	// zones maps zone domain -> NonAuthoritative (true = non-authoritative, user-created zone)
+	zones    map[domain.Domain]bool
+	resolver resolver
+
+	ctx    context.Context
+	cancel context.CancelFunc
 }
 
 func NewResolver() *Resolver {
+	ctx, cancel := context.WithCancel(context.Background())
 	return &Resolver{
 		records: make(map[dns.Question][]dns.RR),
 		domains: make(map[domain.Domain]struct{}),
+		zones:   make(map[domain.Domain]bool),
+		ctx:     ctx,
+		cancel:  cancel,
 	}
 }
 
@@ -37,7 +59,18 @@ func (d *Resolver) String() string {
 	return fmt.Sprintf("LocalResolver [%d records]", len(d.records))
 }
 
-func (d *Resolver) Stop() {}
+func (d *Resolver) Stop() {
+	if d.cancel != nil {
+		d.cancel()
+	}
+
+	d.mu.Lock()
+	defer d.mu.Unlock()
+
+	maps.Clear(d.records)
+	maps.Clear(d.domains)
+	maps.Clear(d.zones)
+}
 
 // ID returns the unique handler ID
 func (d *Resolver) ID() types.HandlerID {
@@ -48,35 +81,85 @@ func (d *Resolver) ProbeAvailability() {}
 
 // ServeDNS handles a DNS request
 func (d *Resolver) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
+	logger := log.WithField("request_id", resutil.GetRequestID(w))
+
 	if len(r.Question) == 0 {
-		log.Debugf("received local resolver request with no question")
+		logger.Debug("received local resolver request with no question")
 		return
 	}
 	question := r.Question[0]
 	question.Name = strings.ToLower(dns.Fqdn(question.Name))
 
-	log.Tracef("received local question: domain=%s type=%v class=%v", r.Question[0].Name, question.Qtype, question.Qclass)
-
 	replyMessage := &dns.Msg{}
 	replyMessage.SetReply(r)
 	replyMessage.RecursionAvailable = true
 
-	// lookup all records matching the question
-	records := d.lookupRecords(question)
-	if len(records) > 0 {
-		replyMessage.Rcode = dns.RcodeSuccess
-		replyMessage.Answer = append(replyMessage.Answer, records...)
-	} else {
-		// Check if we have any records for this domain name with different types
-		if d.hasRecordsForDomain(domain.Domain(question.Name)) {
-			replyMessage.Rcode = dns.RcodeSuccess // NOERROR with 0 records
-		} else {
-			replyMessage.Rcode = dns.RcodeNameError // NXDOMAIN
-		}
+	result := d.lookupRecords(logger, question)
+	replyMessage.Authoritative = !result.hasExternalData
+	replyMessage.Answer = result.records
+	replyMessage.Rcode = d.determineRcode(question, result)
+
+	if replyMessage.Rcode == dns.RcodeNameError && d.shouldFallthrough(question.Name) {
+		d.continueToNext(logger, w, r)
+		return
 	}
 
 	if err := w.WriteMsg(replyMessage); err != nil {
-		log.Warnf("failed to write the local resolver response: %v", err)
+		logger.Warnf("failed to write the local resolver response: %v", err)
+	}
+}
+
+// determineRcode returns the appropriate DNS response code.
+// Per RFC 6604, CNAME chains should return the rcode of the final target resolution,
+// even if CNAME records are included in the answer.
+func (d *Resolver) determineRcode(question dns.Question, result lookupResult) int {
+	// Use the rcode from lookup - this properly handles CNAME chains where
+	// the target may be NXDOMAIN or SERVFAIL even though we have CNAME records
+	if result.rcode != 0 {
+		return result.rcode
+	}
+
+	// No records found, but domain exists with different record types (NODATA)
+	if d.hasRecordsForDomain(domain.Domain(question.Name)) {
+		return dns.RcodeSuccess
+	}
+
+	return dns.RcodeNameError
+}
+
+// findZone finds the matching zone for a query name using reverse suffix lookup.
+// Returns (nonAuthoritative, found). This is O(k) where k = number of labels in qname.
+func (d *Resolver) findZone(qname string) (nonAuthoritative bool, found bool) {
+	qname = strings.ToLower(dns.Fqdn(qname))
+	for {
+		if nonAuth, ok := d.zones[domain.Domain(qname)]; ok {
+			return nonAuth, true
+		}
+		// Move to parent domain
+		idx := strings.Index(qname, ".")
+		if idx == -1 || idx == len(qname)-1 {
+			return false, false
+		}
+		qname = qname[idx+1:]
+	}
+}
+
+// shouldFallthrough checks if the query should fallthrough to the next handler.
+// Returns true if the queried name belongs to a non-authoritative zone.
+func (d *Resolver) shouldFallthrough(qname string) bool {
+	d.mu.RLock()
+	defer d.mu.RUnlock()
+
+	nonAuth, found := d.findZone(qname)
+	return found && nonAuth
+}
+
+func (d *Resolver) continueToNext(logger *log.Entry, w dns.ResponseWriter, r *dns.Msg) {
+	resp := &dns.Msg{}
+	resp.SetRcode(r, dns.RcodeNameError)
+	resp.MsgHdr.Zero = true
+	if err := w.WriteMsg(resp); err != nil {
+		logger.Warnf("failed to write continue signal: %v", err)
 	}
 }
 
@@ -89,8 +172,27 @@ func (d *Resolver) hasRecordsForDomain(domainName domain.Domain) bool {
 	return exists
 }
 
+// isInManagedZone checks if the given name falls within any of our managed zones.
+// This is used to avoid unnecessary external resolution for CNAME targets that
+// are within zones we manage - if we don't have a record for it, it doesn't exist.
+// Caller must NOT hold the lock.
+func (d *Resolver) isInManagedZone(name string) bool {
+	d.mu.RLock()
+	defer d.mu.RUnlock()
+
+	_, found := d.findZone(name)
+	return found
+}
+
+// lookupResult contains the result of a DNS lookup operation.
+type lookupResult struct {
+	records         []dns.RR
+	rcode           int
+	hasExternalData bool
+}
+
 // lookupRecords fetches *all* DNS records matching the first question in r.
-func (d *Resolver) lookupRecords(question dns.Question) []dns.RR {
+func (d *Resolver) lookupRecords(logger *log.Entry, question dns.Question) lookupResult {
 	d.mu.RLock()
 	records, found := d.records[question]
 
@@ -98,10 +200,14 @@ func (d *Resolver) lookupRecords(question dns.Question) []dns.RR {
 		d.mu.RUnlock()
 		// alternatively check if we have a cname
 		if question.Qtype != dns.TypeCNAME {
-			question.Qtype = dns.TypeCNAME
-			return d.lookupRecords(question)
+			cnameQuestion := dns.Question{
+				Name:   question.Name,
+				Qtype:  dns.TypeCNAME,
+				Qclass: question.Qclass,
+			}
+			return d.lookupCNAMEChain(logger, cnameQuestion, question.Qtype)
 		}
-		return nil
+		return lookupResult{rcode: dns.RcodeNameError}
 	}
 
 	recordsCopy := slices.Clone(records)
@@ -119,20 +225,178 @@ func (d *Resolver) lookupRecords(question dns.Question) []dns.RR {
 		d.mu.Unlock()
 	}
 
-	return recordsCopy
+	return lookupResult{records: recordsCopy, rcode: dns.RcodeSuccess}
 }
 
-func (d *Resolver) Update(update []nbdns.SimpleRecord) {
+// lookupCNAMEChain follows a CNAME chain and returns the CNAME records along with
+// the final resolved record of the requested type. This is required for musl libc
+// compatibility, which expects the full answer chain rather than just the CNAME.
+func (d *Resolver) lookupCNAMEChain(logger *log.Entry, cnameQuestion dns.Question, targetType uint16) lookupResult {
+	const maxDepth = 8
+	var chain []dns.RR
+
+	for range maxDepth {
+		cnameRecords := d.getRecords(cnameQuestion)
+		if len(cnameRecords) == 0 {
+			break
+		}
+
+		chain = append(chain, cnameRecords...)
+
+		cname, ok := cnameRecords[0].(*dns.CNAME)
+		if !ok {
+			break
+		}
+
+		targetName := strings.ToLower(cname.Target)
+		targetResult := d.resolveCNAMETarget(logger, targetName, targetType, cnameQuestion.Qclass)
+
+		// keep following chain
+		if targetResult.rcode == -1 {
+			cnameQuestion = dns.Question{Name: targetName, Qtype: dns.TypeCNAME, Qclass: cnameQuestion.Qclass}
+			continue
+		}
+
+		return d.buildChainResult(chain, targetResult)
+	}
+
+	if len(chain) > 0 {
+		return lookupResult{records: chain, rcode: dns.RcodeSuccess}
+	}
+	return lookupResult{rcode: dns.RcodeSuccess}
+}
+
+// buildChainResult combines CNAME chain records with the target resolution result.
+// Per RFC 6604, the final rcode is propagated through the chain.
+func (d *Resolver) buildChainResult(chain []dns.RR, target lookupResult) lookupResult {
+	records := chain
+	if len(target.records) > 0 {
+		records = append(records, target.records...)
+	}
+
+	// preserve hasExternalData for SERVFAIL so caller knows the error came from upstream
+	if target.hasExternalData && target.rcode == dns.RcodeServerFailure {
+		return lookupResult{
+			records:         records,
+			rcode:           dns.RcodeServerFailure,
+			hasExternalData: true,
+		}
+	}
+
+	return lookupResult{
+		records:         records,
+		rcode:           target.rcode,
+		hasExternalData: target.hasExternalData,
+	}
+}
+
+// resolveCNAMETarget attempts to resolve a CNAME target name.
+// Returns rcode=-1 to signal "keep following the chain".
+func (d *Resolver) resolveCNAMETarget(logger *log.Entry, targetName string, targetType uint16, qclass uint16) lookupResult {
+	if records := d.getRecords(dns.Question{Name: targetName, Qtype: targetType, Qclass: qclass}); len(records) > 0 {
+		return lookupResult{records: records, rcode: dns.RcodeSuccess}
+	}
+
+	// another CNAME, keep following
+	if d.hasRecord(dns.Question{Name: targetName, Qtype: dns.TypeCNAME, Qclass: qclass}) {
+		return lookupResult{rcode: -1}
+	}
+
+	// domain exists locally but not this record type (NODATA)
+	if d.hasRecordsForDomain(domain.Domain(targetName)) {
+		return lookupResult{rcode: dns.RcodeSuccess}
+	}
+
+	// in our zone but doesn't exist (NXDOMAIN)
+	if d.isInManagedZone(targetName) {
+		return lookupResult{rcode: dns.RcodeNameError}
+	}
+
+	return d.resolveExternal(logger, targetName, targetType)
+}
+
+func (d *Resolver) getRecords(q dns.Question) []dns.RR {
+	d.mu.RLock()
+	defer d.mu.RUnlock()
+	return d.records[q]
+}
+
+func (d *Resolver) hasRecord(q dns.Question) bool {
+	d.mu.RLock()
+	defer d.mu.RUnlock()
+	_, ok := d.records[q]
+	return ok
+}
+
+// resolveExternal resolves a domain name using the system resolver.
+// This is used to resolve CNAME targets that point outside our local zone,
+// which is required for musl libc compatibility (musl expects complete answers).
+func (d *Resolver) resolveExternal(logger *log.Entry, name string, qtype uint16) lookupResult {
+	network := resutil.NetworkForQtype(qtype)
+	if network == "" {
+		return lookupResult{rcode: dns.RcodeNotImplemented}
+	}
+
+	resolver := d.resolver
+	if resolver == nil {
+		resolver = net.DefaultResolver
+	}
+
+	ctx, cancel := context.WithTimeout(d.ctx, externalResolutionTimeout)
+	defer cancel()
+
+	result := resutil.LookupIP(ctx, resolver, network, name, qtype)
+	if result.Err != nil {
+		d.logDNSError(logger, name, qtype, result.Err)
+		return lookupResult{rcode: result.Rcode, hasExternalData: true}
+	}
+
+	return lookupResult{
+		records:         resutil.IPsToRRs(name, result.IPs, 60),
+		rcode:           dns.RcodeSuccess,
+		hasExternalData: true,
+	}
+}
+
+// logDNSError logs DNS resolution errors for debugging.
+func (d *Resolver) logDNSError(logger *log.Entry, hostname string, qtype uint16, err error) {
+	qtypeName := dns.TypeToString[qtype]
+
+	var dnsErr *net.DNSError
+	if !errors.As(err, &dnsErr) {
+		logger.Debugf("DNS resolution failed for %s type %s: %v", hostname, qtypeName, err)
+		return
+	}
+
+	if dnsErr.IsNotFound {
+		logger.Tracef("DNS target not found: %s type %s", hostname, qtypeName)
+		return
+	}
+
+	if dnsErr.Server != "" {
+		logger.Debugf("DNS resolution failed for %s type %s server=%s: %v", hostname, qtypeName, dnsErr.Server, err)
+	} else {
+		logger.Debugf("DNS resolution failed for %s type %s: %v", hostname, qtypeName, err)
+	}
+}
+
+// Update replaces all zones and their records
+func (d *Resolver) Update(customZones []nbdns.CustomZone) {
 	d.mu.Lock()
 	defer d.mu.Unlock()
 
 	maps.Clear(d.records)
 	maps.Clear(d.domains)
+	maps.Clear(d.zones)
 
-	for _, rec := range update {
-		if err := d.registerRecord(rec); err != nil {
-			log.Warnf("failed to register the record (%s): %v", rec, err)
-			continue
+	for _, zone := range customZones {
+		zoneDomain := domain.Domain(strings.ToLower(dns.Fqdn(zone.Domain)))
+		d.zones[zoneDomain] = zone.NonAuthoritative
+
+		for _, rec := range zone.Records {
+			if err := d.registerRecord(rec); err != nil {
+				log.Warnf("failed to register the record (%s): %v", rec, err)
+			}
 		}
 	}
 }

client/internal/dns/local/local_test.go

@@ -1,8 +1,14 @@
 package local
 
 import (
+	"context"
+	"fmt"
+	"net"
+	"net/netip"
 	"strings"
+	"sync"
 	"testing"
+	"time"
 
 	"github.com/miekg/dns"
 	"github.com/stretchr/testify/assert"
@@ -12,6 +18,18 @@ import (
 	nbdns "github.com/netbirdio/netbird/dns"
 )
 
+// mockResolver implements resolver for testing
+type mockResolver struct {
+	lookupFunc func(ctx context.Context, network, host string) ([]netip.Addr, error)
+}
+
+func (m *mockResolver) LookupNetIP(ctx context.Context, network, host string) ([]netip.Addr, error) {
+	if m.lookupFunc != nil {
+		return m.lookupFunc(ctx, network, host)
+	}
+	return nil, nil
+}
+
 func TestLocalResolver_ServeDNS(t *testing.T) {
 	recordA := nbdns.SimpleRecord{
 		Name:  "peera.netbird.cloud.",
@@ -106,11 +124,11 @@ func TestLocalResolver_Update_StaleRecord(t *testing.T) {
 
 	resolver := NewResolver()
 
-	update1 := []nbdns.SimpleRecord{record1}
-	update2 := []nbdns.SimpleRecord{record2}
+	zone1 := []nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{record1}}}
+	zone2 := []nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{record2}}}
 
 	// Apply first update
-	resolver.Update(update1)
+	resolver.Update(zone1)
 
 	// Verify first update
 	resolver.mu.RLock()
@@ -122,7 +140,7 @@ func TestLocalResolver_Update_StaleRecord(t *testing.T) {
 	assert.Contains(t, rrSlice1[0].String(), record1.RData, "Record after first update should be %s", record1.RData)
 
 	// Apply second update
-	resolver.Update(update2)
+	resolver.Update(zone2)
 
 	// Verify second update
 	resolver.mu.RLock()
@@ -151,10 +169,10 @@ func TestLocalResolver_MultipleRecords_SameQuestion(t *testing.T) {
 		Name: recordName, Type: int(recordType), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.2",
 	}
 
-	update := []nbdns.SimpleRecord{record1, record2}
+	zones := []nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{record1, record2}}}
 
 	// Apply update with both records
-	resolver.Update(update)
+	resolver.Update(zones)
 
 	// Create question that matches both records
 	question := dns.Question{
@@ -195,10 +213,10 @@ func TestLocalResolver_RecordRotation(t *testing.T) {
 		Name: recordName, Type: int(recordType), Class: nbdns.DefaultClass, TTL: 300, RData: "192.168.1.3",
 	}
 
-	update := []nbdns.SimpleRecord{record1, record2, record3}
+	zones := []nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{record1, record2, record3}}}
 
 	// Apply update with all three records
-	resolver.Update(update)
+	resolver.Update(zones)
 
 	msg := new(dns.Msg).SetQuestion(recordName, recordType)
 
@@ -264,7 +282,7 @@ func TestLocalResolver_CaseInsensitiveMatching(t *testing.T) {
 	}
 
 	// Update resolver with the records
-	resolver.Update([]nbdns.SimpleRecord{lowerCaseRecord, mixedCaseRecord})
+	resolver.Update([]nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{lowerCaseRecord, mixedCaseRecord}}})
 
 	testCases := []struct {
 		name          string
@@ -379,7 +397,7 @@ func TestLocalResolver_CNAMEFallback(t *testing.T) {
 	}
 
 	// Update resolver with both records
-	resolver.Update([]nbdns.SimpleRecord{cnameRecord, targetRecord})
+	resolver.Update([]nbdns.CustomZone{{Domain: "example.com.", Records: []nbdns.SimpleRecord{cnameRecord, targetRecord}}})
 
 	testCases := []struct {
 		name          string
@@ -476,6 +494,20 @@ func TestLocalResolver_CNAMEFallback(t *testing.T) {
 // with 0 records instead of NXDOMAIN
 func TestLocalResolver_NoErrorWithDifferentRecordType(t *testing.T) {
 	resolver := NewResolver()
+	// Mock external resolver for CNAME target resolution
+	resolver.resolver = &mockResolver{
+		lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+			if host == "target.example.com." {
+				if network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+				}
+				if network == "ip6" {
+					return []netip.Addr{netip.MustParseAddr("2606:2800:220:1:248:1893:25c8:1946")}, nil
+				}
+			}
+			return nil, &net.DNSError{IsNotFound: true, Name: host}
+		},
+	}
 
 	recordA := nbdns.SimpleRecord{
 		Name:  "example.netbird.cloud.",
@@ -493,7 +525,7 @@ func TestLocalResolver_NoErrorWithDifferentRecordType(t *testing.T) {
 		RData: "target.example.com.",
 	}
 
-	resolver.Update([]nbdns.SimpleRecord{recordA, recordCNAME})
+	resolver.Update([]nbdns.CustomZone{{Domain: "netbird.cloud.", Records: []nbdns.SimpleRecord{recordA, recordCNAME}}})
 
 	testCases := []struct {
 		name           string
@@ -582,3 +614,808 @@ func TestLocalResolver_NoErrorWithDifferentRecordType(t *testing.T) {
 		})
 	}
 }
+
+// TestLocalResolver_CNAMEChainResolution tests comprehensive CNAME chain following
+func TestLocalResolver_CNAMEChainResolution(t *testing.T) {
+	t.Run("simple internal CNAME chain", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "example.com.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.example.com.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "target.example.com."},
+				{Name: "target.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "192.168.1.1"},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.example.com.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
+		require.Len(t, resp.Answer, 2)
+
+		cname, ok := resp.Answer[0].(*dns.CNAME)
+		require.True(t, ok)
+		assert.Equal(t, "target.example.com.", cname.Target)
+
+		a, ok := resp.Answer[1].(*dns.A)
+		require.True(t, ok)
+		assert.Equal(t, "192.168.1.1", a.A.String())
+	})
+
+	t.Run("multi-hop CNAME chain", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "hop1.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "hop2.test."},
+				{Name: "hop2.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "hop3.test."},
+				{Name: "hop3.test.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("hop1.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
+		require.Len(t, resp.Answer, 3)
+	})
+
+	t.Run("CNAME to non-existent internal target returns only CNAME", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "nonexistent.test."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Len(t, resp.Answer, 1)
+		_, ok := resp.Answer[0].(*dns.CNAME)
+		assert.True(t, ok)
+	})
+}
+
+// TestLocalResolver_CNAMEMaxDepth tests the maximum depth limit for CNAME chains
+func TestLocalResolver_CNAMEMaxDepth(t *testing.T) {
+	t.Run("chain at max depth resolves", func(t *testing.T) {
+		resolver := NewResolver()
+		var records []nbdns.SimpleRecord
+		// Create chain of 7 CNAMEs (under max of 8)
+		for i := 1; i <= 7; i++ {
+			records = append(records, nbdns.SimpleRecord{
+				Name:  fmt.Sprintf("hop%d.test.", i),
+				Type:  int(dns.TypeCNAME),
+				Class: nbdns.DefaultClass,
+				TTL:   300,
+				RData: fmt.Sprintf("hop%d.test.", i+1),
+			})
+		}
+		records = append(records, nbdns.SimpleRecord{
+			Name: "hop8.test.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.10.10.10",
+		})
+
+		resolver.Update([]nbdns.CustomZone{{Domain: "test.", Records: records}})
+
+		msg := new(dns.Msg).SetQuestion("hop1.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Equal(t, dns.RcodeSuccess, resp.Rcode)
+		require.Len(t, resp.Answer, 8)
+	})
+
+	t.Run("chain exceeding max depth stops", func(t *testing.T) {
+		resolver := NewResolver()
+		var records []nbdns.SimpleRecord
+		// Create chain of 10 CNAMEs (exceeds max of 8)
+		for i := 1; i <= 10; i++ {
+			records = append(records, nbdns.SimpleRecord{
+				Name:  fmt.Sprintf("deep%d.test.", i),
+				Type:  int(dns.TypeCNAME),
+				Class: nbdns.DefaultClass,
+				TTL:   300,
+				RData: fmt.Sprintf("deep%d.test.", i+1),
+			})
+		}
+		records = append(records, nbdns.SimpleRecord{
+			Name: "deep11.test.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.10.10.10",
+		})
+
+		resolver.Update([]nbdns.CustomZone{{Domain: "test.", Records: records}})
+
+		msg := new(dns.Msg).SetQuestion("deep1.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		// Should NOT have the final A record (chain too deep)
+		assert.LessOrEqual(t, len(resp.Answer), 8)
+	})
+
+	t.Run("circular CNAME is protected by max depth", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "loop1.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "loop2.test."},
+				{Name: "loop2.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "loop1.test."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("loop1.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.LessOrEqual(t, len(resp.Answer), 8)
+	})
+}
+
+// TestLocalResolver_ExternalCNAMEResolution tests CNAME resolution to external domains
+func TestLocalResolver_ExternalCNAMEResolution(t *testing.T) {
+	t.Run("CNAME to external domain resolves via external resolver", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.example.com." && network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+				}
+				return nil, nil
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Len(t, resp.Answer, 2, "Should have CNAME + A record")
+
+		cname, ok := resp.Answer[0].(*dns.CNAME)
+		require.True(t, ok)
+		assert.Equal(t, "external.example.com.", cname.Target)
+
+		a, ok := resp.Answer[1].(*dns.A)
+		require.True(t, ok)
+		assert.Equal(t, "93.184.216.34", a.A.String())
+	})
+
+	t.Run("CNAME to external domain resolves IPv6", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.example.com." && network == "ip6" {
+					return []netip.Addr{netip.MustParseAddr("2606:2800:220:1:248:1893:25c8:1946")}, nil
+				}
+				return nil, nil
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeAAAA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Len(t, resp.Answer, 2, "Should have CNAME + AAAA record")
+
+		cname, ok := resp.Answer[0].(*dns.CNAME)
+		require.True(t, ok)
+		assert.Equal(t, "external.example.com.", cname.Target)
+
+		aaaa, ok := resp.Answer[1].(*dns.AAAA)
+		require.True(t, ok)
+		assert.Equal(t, "2606:2800:220:1:248:1893:25c8:1946", aaaa.AAAA.String())
+	})
+
+	t.Run("concurrent external resolution", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.example.com." && network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+				}
+				return nil, nil
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "concurrent.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		var wg sync.WaitGroup
+		results := make([]*dns.Msg, 10)
+
+		for i := 0; i < 10; i++ {
+			wg.Add(1)
+			go func(idx int) {
+				defer wg.Done()
+				msg := new(dns.Msg).SetQuestion("concurrent.test.", dns.TypeA)
+				var resp *dns.Msg
+				resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+				results[idx] = resp
+			}(i)
+		}
+		wg.Wait()
+
+		for i, resp := range results {
+			require.NotNil(t, resp, "Response %d should not be nil", i)
+			require.Len(t, resp.Answer, 2, "Response %d should have CNAME + A", i)
+		}
+	})
+}
+
+// TestLocalResolver_ZoneManagement tests zone-aware CNAME resolution
+func TestLocalResolver_ZoneManagement(t *testing.T) {
+	t.Run("Update sets zones correctly", func(t *testing.T) {
+		resolver := NewResolver()
+
+		resolver.Update([]nbdns.CustomZone{
+			{Domain: "example.com.", Records: []nbdns.SimpleRecord{
+				{Name: "host.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"},
+			}},
+			{Domain: "test.local."},
+		})
+
+		assert.True(t, resolver.isInManagedZone("host.example.com."))
+		assert.True(t, resolver.isInManagedZone("other.example.com."))
+		assert.True(t, resolver.isInManagedZone("sub.test.local."))
+		assert.False(t, resolver.isInManagedZone("external.com."))
+	})
+
+	t.Run("isInManagedZone case insensitive", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{Domain: "Example.COM."}})
+
+		assert.True(t, resolver.isInManagedZone("host.example.com."))
+		assert.True(t, resolver.isInManagedZone("HOST.EXAMPLE.COM."))
+	})
+
+	t.Run("Update clears zones", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{Domain: "example.com."}})
+		assert.True(t, resolver.isInManagedZone("host.example.com."))
+
+		resolver.Update(nil)
+		assert.False(t, resolver.isInManagedZone("host.example.com."))
+	})
+}
+
+// TestLocalResolver_CNAMEZoneAwareResolution tests CNAME resolution with zone awareness
+func TestLocalResolver_CNAMEZoneAwareResolution(t *testing.T) {
+	t.Run("CNAME target in managed zone returns NXDOMAIN per RFC 6604", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "myzone.test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.myzone.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "nonexistent.myzone.test."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.myzone.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.Equal(t, dns.RcodeNameError, resp.Rcode, "Should return NXDOMAIN")
+		require.Len(t, resp.Answer, 1, "Should include CNAME in answer")
+	})
+
+	t.Run("CNAME to external domain skips zone check", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.other.com." && network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("203.0.113.1")}, nil
+				}
+				return nil, nil
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "myzone.test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.myzone.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.other.com."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.myzone.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.Equal(t, dns.RcodeSuccess, resp.Rcode)
+		require.Len(t, resp.Answer, 2, "Should have CNAME + A from external resolution")
+	})
+
+	t.Run("CNAME target exists with different type returns NODATA not NXDOMAIN", func(t *testing.T) {
+		resolver := NewResolver()
+		// CNAME points to target that has A but no AAAA - query for AAAA should be NODATA
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "myzone.test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.myzone.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "target.myzone.test."},
+				{Name: "target.myzone.test.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "1.1.1.1"},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.myzone.test.", dns.TypeAAAA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.Equal(t, dns.RcodeSuccess, resp.Rcode, "Should return NODATA (success), not NXDOMAIN")
+		require.Len(t, resp.Answer, 1, "Should have only CNAME, no AAAA")
+		_, ok := resp.Answer[0].(*dns.CNAME)
+		assert.True(t, ok, "Answer should be CNAME record")
+	})
+
+	t.Run("external CNAME target exists but no AAAA records (NODATA)", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.example.com." {
+					if network == "ip6" {
+						// No AAAA records
+						return nil, &net.DNSError{IsNotFound: true, Name: host}
+					}
+					if network == "ip4" {
+						// But A records exist - domain exists
+						return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+					}
+				}
+				return nil, &net.DNSError{IsNotFound: true, Name: host}
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeAAAA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.Equal(t, dns.RcodeSuccess, resp.Rcode, "Should return NODATA (success), not NXDOMAIN")
+		require.Len(t, resp.Answer, 1, "Should have only CNAME")
+		_, ok := resp.Answer[0].(*dns.CNAME)
+		assert.True(t, ok, "Answer should be CNAME record")
+	})
+
+	// Table-driven test for all external resolution outcomes
+	externalCases := []struct {
+		name           string
+		lookupFunc     func(context.Context, string, string) ([]netip.Addr, error)
+		expectedRcode  int
+		expectedAnswer int
+	}{
+		{
+			name: "external NXDOMAIN (both A and AAAA not found)",
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				return nil, &net.DNSError{IsNotFound: true, Name: host}
+			},
+			expectedRcode:  dns.RcodeNameError,
+			expectedAnswer: 1, // CNAME only
+		},
+		{
+			name: "external SERVFAIL (temporary error)",
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				return nil, &net.DNSError{IsTemporary: true, Name: host}
+			},
+			expectedRcode:  dns.RcodeServerFailure,
+			expectedAnswer: 1, // CNAME only
+		},
+		{
+			name: "external SERVFAIL (timeout)",
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				return nil, &net.DNSError{IsTimeout: true, Name: host}
+			},
+			expectedRcode:  dns.RcodeServerFailure,
+			expectedAnswer: 1, // CNAME only
+		},
+		{
+			name: "external SERVFAIL (generic error)",
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				return nil, fmt.Errorf("connection refused")
+			},
+			expectedRcode:  dns.RcodeServerFailure,
+			expectedAnswer: 1, // CNAME only
+		},
+		{
+			name: "external success with IPs",
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+				}
+				return nil, &net.DNSError{IsNotFound: true, Name: host}
+			},
+			expectedRcode:  dns.RcodeSuccess,
+			expectedAnswer: 2, // CNAME + A
+		},
+	}
+
+	for _, tc := range externalCases {
+		t.Run(tc.name, func(t *testing.T) {
+			resolver := NewResolver()
+			resolver.resolver = &mockResolver{lookupFunc: tc.lookupFunc}
+
+			resolver.Update([]nbdns.CustomZone{{
+				Domain: "test.",
+				Records: []nbdns.SimpleRecord{
+					{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+				},
+			}})
+
+			msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeA)
+			var resp *dns.Msg
+			resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+			require.NotNil(t, resp)
+			assert.Equal(t, tc.expectedRcode, resp.Rcode, "rcode mismatch")
+			assert.Len(t, resp.Answer, tc.expectedAnswer, "answer count mismatch")
+			if tc.expectedAnswer > 0 {
+				_, ok := resp.Answer[0].(*dns.CNAME)
+				assert.True(t, ok, "first answer should be CNAME")
+			}
+		})
+	}
+}
+
+// TestLocalResolver_Fallthrough verifies that non-authoritative zones
+// trigger fallthrough (Zero bit set) when no records match
+func TestLocalResolver_Fallthrough(t *testing.T) {
+	resolver := NewResolver()
+
+	record := nbdns.SimpleRecord{
+		Name:  "existing.custom.zone.",
+		Type:  int(dns.TypeA),
+		Class: nbdns.DefaultClass,
+		TTL:   300,
+		RData: "10.0.0.1",
+	}
+
+	testCases := []struct {
+		name              string
+		zones             []nbdns.CustomZone
+		queryName         string
+		expectFallthrough bool
+		expectRecord      bool
+	}{
+		{
+			name: "Authoritative zone returns NXDOMAIN without fallthrough",
+			zones: []nbdns.CustomZone{{
+				Domain:  "custom.zone.",
+				Records: []nbdns.SimpleRecord{record},
+			}},
+			queryName:         "nonexistent.custom.zone.",
+			expectFallthrough: false,
+			expectRecord:      false,
+		},
+		{
+			name: "Non-authoritative zone triggers fallthrough",
+			zones: []nbdns.CustomZone{{
+				Domain:           "custom.zone.",
+				Records:          []nbdns.SimpleRecord{record},
+				NonAuthoritative: true,
+			}},
+			queryName:         "nonexistent.custom.zone.",
+			expectFallthrough: true,
+			expectRecord:      false,
+		},
+		{
+			name: "Record found in non-authoritative zone returns normally",
+			zones: []nbdns.CustomZone{{
+				Domain:           "custom.zone.",
+				Records:          []nbdns.SimpleRecord{record},
+				NonAuthoritative: true,
+			}},
+			queryName:         "existing.custom.zone.",
+			expectFallthrough: false,
+			expectRecord:      true,
+		},
+		{
+			name: "Record found in authoritative zone returns normally",
+			zones: []nbdns.CustomZone{{
+				Domain:  "custom.zone.",
+				Records: []nbdns.SimpleRecord{record},
+			}},
+			queryName:         "existing.custom.zone.",
+			expectFallthrough: false,
+			expectRecord:      true,
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			resolver.Update(tc.zones)
+
+			var responseMSG *dns.Msg
+			responseWriter := &test.MockResponseWriter{
+				WriteMsgFunc: func(m *dns.Msg) error {
+					responseMSG = m
+					return nil
+				},
+			}
+
+			msg := new(dns.Msg).SetQuestion(tc.queryName, dns.TypeA)
+			resolver.ServeDNS(responseWriter, msg)
+
+			require.NotNil(t, responseMSG, "Should have received a response")
+
+			if tc.expectFallthrough {
+				assert.True(t, responseMSG.MsgHdr.Zero, "Zero bit should be set for fallthrough")
+				assert.Equal(t, dns.RcodeNameError, responseMSG.Rcode, "Should return NXDOMAIN")
+			} else {
+				assert.False(t, responseMSG.MsgHdr.Zero, "Zero bit should not be set")
+			}
+
+			if tc.expectRecord {
+				assert.Greater(t, len(responseMSG.Answer), 0, "Should have answer records")
+				assert.Equal(t, dns.RcodeSuccess, responseMSG.Rcode)
+			}
+		})
+	}
+}
+
+// TestLocalResolver_AuthoritativeFlag tests the AA flag behavior
+func TestLocalResolver_AuthoritativeFlag(t *testing.T) {
+	t.Run("direct record lookup is authoritative", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "example.com.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "host.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("host.example.com.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.True(t, resp.Authoritative)
+	})
+
+	t.Run("external resolution is not authoritative", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(_ context.Context, network, host string) ([]netip.Addr, error) {
+				if host == "external.example.com." && network == "ip4" {
+					return []netip.Addr{netip.MustParseAddr("93.184.216.34")}, nil
+				}
+				return nil, nil
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		require.Len(t, resp.Answer, 2)
+		assert.False(t, resp.Authoritative)
+	})
+}
+
+// TestLocalResolver_Stop tests cleanup on Stop
+func TestLocalResolver_Stop(t *testing.T) {
+	t.Run("Stop clears all state", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "example.com.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "host.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"},
+			},
+		}})
+
+		resolver.Stop()
+
+		msg := new(dns.Msg).SetQuestion("host.example.com.", dns.TypeA)
+		var resp *dns.Msg
+		resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { resp = m; return nil }}, msg)
+
+		require.NotNil(t, resp)
+		assert.Len(t, resp.Answer, 0)
+		assert.False(t, resolver.isInManagedZone("host.example.com."))
+	})
+
+	t.Run("Stop is safe to call multiple times", func(t *testing.T) {
+		resolver := NewResolver()
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "example.com.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "host.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"},
+			},
+		}})
+
+		resolver.Stop()
+		resolver.Stop()
+		resolver.Stop()
+	})
+
+	t.Run("Stop cancels in-flight external resolution", func(t *testing.T) {
+		resolver := NewResolver()
+
+		lookupStarted := make(chan struct{})
+		lookupCtxCanceled := make(chan struct{})
+
+		resolver.resolver = &mockResolver{
+			lookupFunc: func(ctx context.Context, network, host string) ([]netip.Addr, error) {
+				close(lookupStarted)
+				<-ctx.Done()
+				close(lookupCtxCanceled)
+				return nil, ctx.Err()
+			},
+		}
+
+		resolver.Update([]nbdns.CustomZone{{
+			Domain: "test.",
+			Records: []nbdns.SimpleRecord{
+				{Name: "alias.test.", Type: int(dns.TypeCNAME), Class: nbdns.DefaultClass, TTL: 300, RData: "external.example.com."},
+			},
+		}})
+
+		done := make(chan struct{})
+		go func() {
+			msg := new(dns.Msg).SetQuestion("alias.test.", dns.TypeA)
+			resolver.ServeDNS(&test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { return nil }}, msg)
+			close(done)
+		}()
+
+		<-lookupStarted
+		resolver.Stop()
+
+		select {
+		case <-lookupCtxCanceled:
+		case <-time.After(time.Second):
+			t.Fatal("external lookup context was not canceled")
+		}
+
+		select {
+		case <-done:
+		case <-time.After(time.Second):
+			t.Fatal("ServeDNS did not return after Stop")
+		}
+	})
+}
+
+// TestLocalResolver_FallthroughCaseInsensitive verifies case-insensitive domain matching for fallthrough
+func TestLocalResolver_FallthroughCaseInsensitive(t *testing.T) {
+	resolver := NewResolver()
+
+	resolver.Update([]nbdns.CustomZone{{
+		Domain:           "EXAMPLE.COM.",
+		Records:          []nbdns.SimpleRecord{{Name: "host.example.com.", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "1.2.3.4"}},
+		NonAuthoritative: true,
+	}})
+
+	var responseMSG *dns.Msg
+	responseWriter := &test.MockResponseWriter{
+		WriteMsgFunc: func(m *dns.Msg) error {
+			responseMSG = m
+			return nil
+		},
+	}
+
+	msg := new(dns.Msg).SetQuestion("nonexistent.example.com.", dns.TypeA)
+	resolver.ServeDNS(responseWriter, msg)
+
+	require.NotNil(t, responseMSG)
+	assert.True(t, responseMSG.MsgHdr.Zero, "Should fallthrough for non-authoritative zone with case-insensitive match")
+}
+
+// BenchmarkFindZone_BestCase benchmarks zone lookup with immediate match (first label)
+func BenchmarkFindZone_BestCase(b *testing.B) {
+	resolver := NewResolver()
+
+	// Single zone that matches immediately
+	resolver.Update([]nbdns.CustomZone{{
+		Domain:           "example.com.",
+		NonAuthoritative: true,
+	}})
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		resolver.shouldFallthrough("example.com.")
+	}
+}
+
+// BenchmarkFindZone_WorstCase benchmarks zone lookup with many zones, no match, many labels
+func BenchmarkFindZone_WorstCase(b *testing.B) {
+	resolver := NewResolver()
+
+	// 100 zones that won't match
+	var zones []nbdns.CustomZone
+	for i := 0; i < 100; i++ {
+		zones = append(zones, nbdns.CustomZone{
+			Domain:           fmt.Sprintf("zone%d.internal.", i),
+			NonAuthoritative: true,
+		})
+	}
+	resolver.Update(zones)
+
+	// Query with many labels that won't match any zone
+	qname := "a.b.c.d.e.f.g.h.external.com."
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		resolver.shouldFallthrough(qname)
+	}
+}
+
+// BenchmarkFindZone_TypicalCase benchmarks typical usage: few zones, subdomain match
+func BenchmarkFindZone_TypicalCase(b *testing.B) {
+	resolver := NewResolver()
+
+	// Typical setup: peer zone (authoritative) + one user zone (non-authoritative)
+	resolver.Update([]nbdns.CustomZone{
+		{Domain: "netbird.cloud.", NonAuthoritative: false},
+		{Domain: "custom.local.", NonAuthoritative: true},
+	})
+
+	// Query for subdomain of user zone
+	qname := "myhost.custom.local."
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		resolver.shouldFallthrough(qname)
+	}
+}
+
+// BenchmarkIsInManagedZone_ManyZones benchmarks isInManagedZone with 100 zones
+func BenchmarkIsInManagedZone_ManyZones(b *testing.B) {
+	resolver := NewResolver()
+
+	var zones []nbdns.CustomZone
+	for i := 0; i < 100; i++ {
+		zones = append(zones, nbdns.CustomZone{
+			Domain: fmt.Sprintf("zone%d.internal.", i),
+		})
+	}
+	resolver.Update(zones)
+
+	// Query that matches zone50
+	qname := "host.zone50.internal."
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		resolver.isInManagedZone(qname)
+	}
+}

client/internal/dns/resutil/resolve.go

@@ -0,0 +1,197 @@
+// Package resutil provides shared DNS resolution utilities
+package resutil
+
+import (
+	"context"
+	"crypto/rand"
+	"encoding/hex"
+	"errors"
+	"net"
+	"net/netip"
+	"strings"
+
+	"github.com/miekg/dns"
+	log "github.com/sirupsen/logrus"
+)
+
+// GenerateRequestID creates a random 8-character hex string for request tracing.
+func GenerateRequestID() string {
+	bytes := make([]byte, 4)
+	if _, err := rand.Read(bytes); err != nil {
+		log.Errorf("generate request ID: %v", err)
+		return ""
+	}
+	return hex.EncodeToString(bytes)
+}
+
+// IPsToRRs converts a slice of IP addresses to DNS resource records.
+// IPv4 addresses become A records, IPv6 addresses become AAAA records.
+func IPsToRRs(name string, ips []netip.Addr, ttl uint32) []dns.RR {
+	var result []dns.RR
+
+	for _, ip := range ips {
+		if ip.Is6() {
+			result = append(result, &dns.AAAA{
+				Hdr: dns.RR_Header{
+					Name:   name,
+					Rrtype: dns.TypeAAAA,
+					Class:  dns.ClassINET,
+					Ttl:    ttl,
+				},
+				AAAA: ip.AsSlice(),
+			})
+		} else {
+			result = append(result, &dns.A{
+				Hdr: dns.RR_Header{
+					Name:   name,
+					Rrtype: dns.TypeA,
+					Class:  dns.ClassINET,
+					Ttl:    ttl,
+				},
+				A: ip.AsSlice(),
+			})
+		}
+	}
+
+	return result
+}
+
+// NetworkForQtype returns the network string ("ip4" or "ip6") for a DNS query type.
+// Returns empty string for unsupported types.
+func NetworkForQtype(qtype uint16) string {
+	switch qtype {
+	case dns.TypeA:
+		return "ip4"
+	case dns.TypeAAAA:
+		return "ip6"
+	default:
+		return ""
+	}
+}
+
+type resolver interface {
+	LookupNetIP(ctx context.Context, network, host string) ([]netip.Addr, error)
+}
+
+// chainedWriter is implemented by ResponseWriters that carry request metadata
+type chainedWriter interface {
+	RequestID() string
+	SetMeta(key, value string)
+}
+
+// GetRequestID extracts a request ID from the ResponseWriter if available,
+// otherwise generates a new one.
+func GetRequestID(w dns.ResponseWriter) string {
+	if cw, ok := w.(chainedWriter); ok {
+		if id := cw.RequestID(); id != "" {
+			return id
+		}
+	}
+	return GenerateRequestID()
+}
+
+// SetMeta sets metadata on the ResponseWriter if it supports it.
+func SetMeta(w dns.ResponseWriter, key, value string) {
+	if cw, ok := w.(chainedWriter); ok {
+		cw.SetMeta(key, value)
+	}
+}
+
+// LookupResult contains the result of an external DNS lookup
+type LookupResult struct {
+	IPs   []netip.Addr
+	Rcode int
+	Err   error // Original error for caller's logging needs
+}
+
+// LookupIP performs a DNS lookup and determines the appropriate rcode.
+func LookupIP(ctx context.Context, r resolver, network, host string, qtype uint16) LookupResult {
+	ips, err := r.LookupNetIP(ctx, network, host)
+	if err != nil {
+		return LookupResult{
+			Rcode: getRcodeForError(ctx, r, host, qtype, err),
+			Err:   err,
+		}
+	}
+
+	// Unmap IPv4-mapped IPv6 addresses that some resolvers may return
+	for i, ip := range ips {
+		ips[i] = ip.Unmap()
+	}
+
+	return LookupResult{
+		IPs:   ips,
+		Rcode: dns.RcodeSuccess,
+	}
+}
+
+func getRcodeForError(ctx context.Context, r resolver, host string, qtype uint16, err error) int {
+	var dnsErr *net.DNSError
+	if !errors.As(err, &dnsErr) {
+		return dns.RcodeServerFailure
+	}
+
+	if dnsErr.IsNotFound {
+		return getRcodeForNotFound(ctx, r, host, qtype)
+	}
+
+	return dns.RcodeServerFailure
+}
+
+// getRcodeForNotFound distinguishes between NXDOMAIN (domain doesn't exist) and NODATA
+// (domain exists but no records of requested type) by checking the opposite record type.
+//
+// musl libc (the reason we need this distinction) only queries A/AAAA pairs in getaddrinfo,
+// so checking the opposite A/AAAA type is sufficient. Other record types (MX, TXT, etc.)
+// are not queried by musl and don't need this handling.
+func getRcodeForNotFound(ctx context.Context, r resolver, domain string, originalQtype uint16) int {
+	// Try querying for a different record type to see if the domain exists
+	// If the original query was for AAAA, try A. If it was for A, try AAAA.
+	// This helps distinguish between NXDOMAIN and NODATA.
+	var alternativeNetwork string
+	switch originalQtype {
+	case dns.TypeAAAA:
+		alternativeNetwork = "ip4"
+	case dns.TypeA:
+		alternativeNetwork = "ip6"
+	default:
+		return dns.RcodeNameError
+	}
+
+	if _, err := r.LookupNetIP(ctx, alternativeNetwork, domain); err != nil {
+		var dnsErr *net.DNSError
+		if errors.As(err, &dnsErr) && dnsErr.IsNotFound {
+			// Alternative query also returned not found - domain truly doesn't exist
+			return dns.RcodeNameError
+		}
+		// Some other error (timeout, server failure, etc.) - can't determine, assume domain exists
+		return dns.RcodeSuccess
+	}
+
+	// Alternative query succeeded - domain exists but has no records of this type
+	return dns.RcodeSuccess
+}
+
+// FormatAnswers formats DNS resource records for logging.
+func FormatAnswers(answers []dns.RR) string {
+	if len(answers) == 0 {
+		return "[]"
+	}
+
+	parts := make([]string, 0, len(answers))
+	for _, rr := range answers {
+		switch r := rr.(type) {
+		case *dns.A:
+			parts = append(parts, r.A.String())
+		case *dns.AAAA:
+			parts = append(parts, r.AAAA.String())
+		case *dns.CNAME:
+			parts = append(parts, "CNAME:"+r.Target)
+		case *dns.PTR:
+			parts = append(parts, "PTR:"+r.Ptr)
+		default:
+			parts = append(parts, dns.TypeToString[rr.Header().Rrtype])
+		}
+	}
+	return "[" + strings.Join(parts, ", ") + "]"
+}

client/internal/dns/server.go

@@ -485,7 +485,7 @@ func (s *DefaultServer) applyConfiguration(update nbdns.Config) error {
 		}
 	}
 
-	localMuxUpdates, localRecords, err := s.buildLocalHandlerUpdate(update.CustomZones)
+	localMuxUpdates, localZones, err := s.buildLocalHandlerUpdate(update.CustomZones)
 	if err != nil {
 		return fmt.Errorf("local handler updater: %w", err)
 	}
@@ -498,8 +498,7 @@ func (s *DefaultServer) applyConfiguration(update nbdns.Config) error {
 
 	s.updateMux(muxUpdates)
 
-	// register local records
-	s.localResolver.Update(localRecords)
+	s.localResolver.Update(localZones)
 
 	s.currentConfig = dnsConfigToHostDNSConfig(update, s.service.RuntimeIP(), s.service.RuntimePort())
 
@@ -632,9 +631,7 @@ func (s *DefaultServer) registerFallback(config HostDNSConfig) {
 
 	handler, err := newUpstreamResolver(
 		s.ctx,
-		s.wgInterface.Name(),
-		s.wgInterface.Address().IP,
-		s.wgInterface.Address().Network,
+		s.wgInterface,
 		s.statusRecorder,
 		s.hostsDNSHolder,
 		nbdns.RootZone,
@@ -659,9 +656,9 @@ func (s *DefaultServer) registerFallback(config HostDNSConfig) {
 	s.registerHandler([]string{nbdns.RootZone}, handler, PriorityFallback)
 }
 
-func (s *DefaultServer) buildLocalHandlerUpdate(customZones []nbdns.CustomZone) ([]handlerWrapper, []nbdns.SimpleRecord, error) {
+func (s *DefaultServer) buildLocalHandlerUpdate(customZones []nbdns.CustomZone) ([]handlerWrapper, []nbdns.CustomZone, error) {
 	var muxUpdates []handlerWrapper
-	var localRecords []nbdns.SimpleRecord
+	var zones []nbdns.CustomZone
 
 	for _, customZone := range customZones {
 		if len(customZone.Records) == 0 {
@@ -675,17 +672,20 @@ func (s *DefaultServer) buildLocalHandlerUpdate(customZones []nbdns.CustomZone)
 			priority: PriorityLocal,
 		})
 
+		// zone records contain the fqdn, so we can just flatten them
+		var localRecords []nbdns.SimpleRecord
 		for _, record := range customZone.Records {
 			if record.Class != nbdns.DefaultClass {
 				log.Warnf("received an invalid class type: %s", record.Class)
 				continue
 			}
-			// zone records contain the fqdn, so we can just flatten them
 			localRecords = append(localRecords, record)
 		}
+		customZone.Records = localRecords
+		zones = append(zones, customZone)
 	}
 
-	return muxUpdates, localRecords, nil
+	return muxUpdates, zones, nil
 }
 
 func (s *DefaultServer) buildUpstreamHandlerUpdate(nameServerGroups []*nbdns.NameServerGroup) ([]handlerWrapper, error) {
@@ -741,9 +741,7 @@ func (s *DefaultServer) createHandlersForDomainGroup(domainGroup nsGroupsByDomai
 		log.Debugf("creating handler for domain=%s with priority=%d", domainGroup.domain, priority)
 		handler, err := newUpstreamResolver(
 			s.ctx,
-			s.wgInterface.Name(),
-			s.wgInterface.Address().IP,
-			s.wgInterface.Address().Network,
+			s.wgInterface,
 			s.statusRecorder,
 			s.hostsDNSHolder,
 			domainGroup.domain,
@@ -924,9 +922,7 @@ func (s *DefaultServer) addHostRootZone() {
 
 	handler, err := newUpstreamResolver(
 		s.ctx,
-		s.wgInterface.Name(),
-		s.wgInterface.Address().IP,
-		s.wgInterface.Address().Network,
+		s.wgInterface,
 		s.statusRecorder,
 		s.hostsDNSHolder,
 		nbdns.RootZone,

client/internal/dns/server_test.go

@@ -15,6 +15,7 @@ import (
 	log "github.com/sirupsen/logrus"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/mock"
+	"golang.zx2c4.com/wireguard/tun/netstack"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
 
 	"github.com/netbirdio/netbird/client/firewall/uspfilter"
@@ -81,6 +82,10 @@ func (w *mocWGIface) GetStats(_ string) (configurer.WGStats, error) {
 	return configurer.WGStats{}, nil
 }
 
+func (w *mocWGIface) GetNet() *netstack.Net {
+	return nil
+}
+
 var zoneRecords = []nbdns.SimpleRecord{
 	{
 		Name:  "peera.netbird.cloud",
@@ -128,7 +133,7 @@ func TestUpdateDNSServer(t *testing.T) {
 	testCases := []struct {
 		name                string
 		initUpstreamMap     registeredHandlerMap
-		initLocalRecords    []nbdns.SimpleRecord
+		initLocalZones      []nbdns.CustomZone
 		initSerial          uint64
 		inputSerial         uint64
 		inputUpdate         nbdns.Config
@@ -180,8 +185,8 @@ func TestUpdateDNSServer(t *testing.T) {
 			expectedLocalQs: []dns.Question{{Name: "peera.netbird.cloud.", Qtype: dns.TypeA, Qclass: dns.ClassINET}},
 		},
 		{
-			name:             "New Config Should Succeed",
-			initLocalRecords: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}},
+			name:           "New Config Should Succeed",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: 1, Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
 			initUpstreamMap: registeredHandlerMap{
 				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
 					domain:   "netbird.cloud",
@@ -221,19 +226,19 @@ func TestUpdateDNSServer(t *testing.T) {
 			expectedLocalQs: []dns.Question{{Name: zoneRecords[0].Name, Qtype: 1, Qclass: 1}},
 		},
 		{
-			name:             "Smaller Config Serial Should Be Skipped",
-			initLocalRecords: []nbdns.SimpleRecord{},
-			initUpstreamMap:  make(registeredHandlerMap),
-			initSerial:       2,
-			inputSerial:      1,
-			shouldFail:       true,
+			name:            "Smaller Config Serial Should Be Skipped",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      2,
+			inputSerial:     1,
+			shouldFail:      true,
 		},
 		{
-			name:             "Empty NS Group Domain Or Not Primary Element Should Fail",
-			initLocalRecords: []nbdns.SimpleRecord{},
-			initUpstreamMap:  make(registeredHandlerMap),
-			initSerial:       0,
-			inputSerial:      1,
+			name:            "Empty NS Group Domain Or Not Primary Element Should Fail",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
 			inputUpdate: nbdns.Config{
 				ServiceEnable: true,
 				CustomZones: []nbdns.CustomZone{
@@ -251,11 +256,11 @@ func TestUpdateDNSServer(t *testing.T) {
 			shouldFail: true,
 		},
 		{
-			name:             "Invalid NS Group Nameservers list Should Fail",
-			initLocalRecords: []nbdns.SimpleRecord{},
-			initUpstreamMap:  make(registeredHandlerMap),
-			initSerial:       0,
-			inputSerial:      1,
+			name:            "Invalid NS Group Nameservers list Should Fail",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
 			inputUpdate: nbdns.Config{
 				ServiceEnable: true,
 				CustomZones: []nbdns.CustomZone{
@@ -273,11 +278,11 @@ func TestUpdateDNSServer(t *testing.T) {
 			shouldFail: true,
 		},
 		{
-			name:             "Invalid Custom Zone Records list Should Skip",
-			initLocalRecords: []nbdns.SimpleRecord{},
-			initUpstreamMap:  make(registeredHandlerMap),
-			initSerial:       0,
-			inputSerial:      1,
+			name:            "Invalid Custom Zone Records list Should Skip",
+			initLocalZones:  []nbdns.CustomZone{},
+			initUpstreamMap: make(registeredHandlerMap),
+			initSerial:      0,
+			inputSerial:     1,
 			inputUpdate: nbdns.Config{
 				ServiceEnable: true,
 				CustomZones: []nbdns.CustomZone{
@@ -299,8 +304,8 @@ func TestUpdateDNSServer(t *testing.T) {
 			}},
 		},
 		{
-			name:             "Empty Config Should Succeed and Clean Maps",
-			initLocalRecords: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}},
+			name:           "Empty Config Should Succeed and Clean Maps",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
 			initUpstreamMap: registeredHandlerMap{
 				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
 					domain:   zoneRecords[0].Name,
@@ -315,8 +320,8 @@ func TestUpdateDNSServer(t *testing.T) {
 			expectedLocalQs:     []dns.Question{},
 		},
 		{
-			name:             "Disabled Service Should clean map",
-			initLocalRecords: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}},
+			name:           "Disabled Service Should clean map",
+			initLocalZones: []nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}},
 			initUpstreamMap: registeredHandlerMap{
 				generateDummyHandler(zoneRecords[0].Name, nameServers).ID(): handlerWrapper{
 					domain:   zoneRecords[0].Name,
@@ -385,7 +390,7 @@ func TestUpdateDNSServer(t *testing.T) {
 			}()
 
 			dnsServer.dnsMuxMap = testCase.initUpstreamMap
-			dnsServer.localResolver.Update(testCase.initLocalRecords)
+			dnsServer.localResolver.Update(testCase.initLocalZones)
 			dnsServer.updateSerial = testCase.initSerial
 
 			err = dnsServer.UpdateDNSServer(testCase.inputSerial, testCase.inputUpdate)
@@ -510,8 +515,7 @@ func TestDNSFakeResolverHandleUpdates(t *testing.T) {
 			priority: PriorityUpstream,
 		},
 	}
-	//dnsServer.localResolver.RegisteredMap = local.RegistrationMap{local.BuildRecordKey("netbird.cloud", dns.ClassINET, dns.TypeA): struct{}{}}
-	dnsServer.localResolver.Update([]nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}})
+	dnsServer.localResolver.Update([]nbdns.CustomZone{{Domain: "netbird.cloud", Records: []nbdns.SimpleRecord{{Name: "netbird.cloud", Type: int(dns.TypeA), Class: nbdns.DefaultClass, TTL: 300, RData: "10.0.0.1"}}}})
 	dnsServer.updateSerial = 0
 
 	nameServers := []nbdns.NameServer{
@@ -2048,7 +2052,7 @@ func TestLocalResolverPriorityInServer(t *testing.T) {
 
 func TestLocalResolverPriorityConstants(t *testing.T) {
 	// Test that priority constants are ordered correctly
-	assert.Greater(t, PriorityLocal, PriorityDNSRoute, "Local priority should be higher than DNS route")
+	assert.Greater(t, PriorityDNSRoute, PriorityLocal, "DNS Route should be higher than Local priority")
 	assert.Greater(t, PriorityLocal, PriorityUpstream, "Local priority should be higher than upstream")
 	assert.Greater(t, PriorityUpstream, PriorityDefault, "Upstream priority should be higher than default")
 

client/internal/dns/upstream.go

@@ -2,7 +2,6 @@ package dns
 
 import (
 	"context"
-	"crypto/rand"
 	"crypto/sha256"
 	"encoding/hex"
 	"errors"
@@ -19,8 +18,10 @@ import (
 	"github.com/hashicorp/go-multierror"
 	"github.com/miekg/dns"
 	log "github.com/sirupsen/logrus"
+	"golang.zx2c4.com/wireguard/tun/netstack"
 
 	"github.com/netbirdio/netbird/client/iface"
+	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/client/internal/dns/types"
 	"github.com/netbirdio/netbird/client/internal/peer"
 	"github.com/netbirdio/netbird/client/proto"
@@ -113,10 +114,7 @@ func (u *upstreamResolverBase) Stop() {
 
 // ServeDNS handles a DNS request
 func (u *upstreamResolverBase) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	requestID := GenerateRequestID()
-	logger := log.WithField("request_id", requestID)
-
-	logger.Tracef("received upstream question: domain=%s type=%v class=%v", r.Question[0].Name, r.Question[0].Qtype, r.Question[0].Qclass)
+	logger := log.WithField("request_id", resutil.GetRequestID(w))
 
 	u.prepareRequest(r)
 
@@ -202,11 +200,18 @@ func (u *upstreamResolverBase) handleUpstreamError(err error, upstream netip.Add
 
 func (u *upstreamResolverBase) writeSuccessResponse(w dns.ResponseWriter, rm *dns.Msg, upstream netip.AddrPort, domain string, t time.Duration, logger *log.Entry) bool {
 	u.successCount.Add(1)
-	logger.Tracef("took %s to query the upstream %s for question domain=%s", t, upstream, domain)
+
+	resutil.SetMeta(w, "upstream", upstream.String())
+
+	// Clear Zero bit from external responses to prevent upstream servers from
+	// manipulating our internal fallthrough signaling mechanism
+	rm.MsgHdr.Zero = false
 
 	if err := w.WriteMsg(rm); err != nil {
 		logger.Errorf("failed to write DNS response for question domain=%s: %s", domain, err)
+		return true
 	}
+
 	return true
 }
 
@@ -414,16 +419,56 @@ func ExchangeWithFallback(ctx context.Context, client *dns.Client, r *dns.Msg, u
 	return rm, t, nil
 }
 
-func GenerateRequestID() string {
-	bytes := make([]byte, 4)
-	_, err := rand.Read(bytes)
+// ExchangeWithNetstack performs a DNS exchange using netstack for dialing.
+// This is needed when netstack is enabled to reach peer IPs through the tunnel.
+func ExchangeWithNetstack(ctx context.Context, nsNet *netstack.Net, r *dns.Msg, upstream string) (*dns.Msg, error) {
+	reply, err := netstackExchange(ctx, nsNet, r, upstream, "udp")
 	if err != nil {
-		log.Errorf("failed to generate request ID: %v", err)
-		return ""
+		return nil, err
 	}
-	return hex.EncodeToString(bytes)
+
+	// If response is truncated, retry with TCP
+	if reply != nil && reply.MsgHdr.Truncated {
+		log.Tracef("udp response for domain=%s type=%v class=%v is truncated, trying TCP",
+			r.Question[0].Name, r.Question[0].Qtype, r.Question[0].Qclass)
+		return netstackExchange(ctx, nsNet, r, upstream, "tcp")
+	}
+
+	return reply, nil
 }
 
+func netstackExchange(ctx context.Context, nsNet *netstack.Net, r *dns.Msg, upstream, network string) (*dns.Msg, error) {
+	conn, err := nsNet.DialContext(ctx, network, upstream)
+	if err != nil {
+		return nil, fmt.Errorf("with %s: %w", network, err)
+	}
+	defer func() {
+		if err := conn.Close(); err != nil {
+			log.Debugf("failed to close DNS connection: %v", err)
+		}
+	}()
+
+	if deadline, ok := ctx.Deadline(); ok {
+		if err := conn.SetDeadline(deadline); err != nil {
+			return nil, fmt.Errorf("set deadline: %w", err)
+		}
+	}
+
+	dnsConn := &dns.Conn{Conn: conn}
+
+	if err := dnsConn.WriteMsg(r); err != nil {
+		return nil, fmt.Errorf("write %s message: %w", network, err)
+	}
+
+	reply, err := dnsConn.ReadMsg()
+	if err != nil {
+		return nil, fmt.Errorf("read %s message: %w", network, err)
+	}
+
+	return reply, nil
+}
+
+
 // FormatPeerStatus formats peer connection status information for debugging DNS timeouts
 func FormatPeerStatus(peerState *peer.State) string {
 	isConnected := peerState.ConnStatus == peer.StatusConnected

client/internal/dns/upstream_android.go

@@ -23,9 +23,7 @@ type upstreamResolver struct {
 // first time, and we need to wait for a while to start to use again the proper DNS resolver.
 func newUpstreamResolver(
 	ctx context.Context,
-	_ string,
-	_ netip.Addr,
-	_ netip.Prefix,
+	_ WGIface,
 	statusRecorder *peer.Status,
 	hostsDNSHolder *hostsDNSHolder,
 	domain string,

client/internal/dns/upstream_general.go

@@ -5,22 +5,23 @@ package dns
 import (
 	"context"
 	"net/netip"
+	"runtime"
 	"time"
 
 	"github.com/miekg/dns"
+	"golang.zx2c4.com/wireguard/tun/netstack"
 
 	"github.com/netbirdio/netbird/client/internal/peer"
 )
 
 type upstreamResolver struct {
 	*upstreamResolverBase
+	nsNet *netstack.Net
 }
 
 func newUpstreamResolver(
 	ctx context.Context,
-	_ string,
-	_ netip.Addr,
-	_ netip.Prefix,
+	wgIface WGIface,
 	statusRecorder *peer.Status,
 	_ *hostsDNSHolder,
 	domain string,
@@ -28,12 +29,23 @@ func newUpstreamResolver(
 	upstreamResolverBase := newUpstreamResolverBase(ctx, statusRecorder, domain)
 	nonIOS := &upstreamResolver{
 		upstreamResolverBase: upstreamResolverBase,
+		nsNet:                wgIface.GetNet(),
 	}
 	upstreamResolverBase.upstreamClient = nonIOS
 	return nonIOS, nil
 }
 
 func (u *upstreamResolver) exchange(ctx context.Context, upstream string, r *dns.Msg) (rm *dns.Msg, t time.Duration, err error) {
+	// TODO: Check if upstream DNS server is routed through a peer before using netstack.
+	// Similar to iOS logic, we should determine if the DNS server is reachable directly
+	// or needs to go through the tunnel, and only use netstack when necessary.
+	// For now, only use netstack on JS platform where direct access is not possible.
+	if u.nsNet != nil && runtime.GOOS == "js" {
+		start := time.Now()
+		reply, err := ExchangeWithNetstack(ctx, u.nsNet, r, upstream)
+		return reply, time.Since(start), err
+	}
+
 	client := &dns.Client{
 		Timeout: ClientTimeout,
 	}

client/internal/dns/upstream_ios.go

@@ -26,9 +26,7 @@ type upstreamResolverIOS struct {
 
 func newUpstreamResolver(
 	ctx context.Context,
-	interfaceName string,
-	ip netip.Addr,
-	net netip.Prefix,
+	wgIface WGIface,
 	statusRecorder *peer.Status,
 	_ *hostsDNSHolder,
 	domain string,
@@ -37,9 +35,9 @@ func newUpstreamResolver(
 
 	ios := &upstreamResolverIOS{
 		upstreamResolverBase: upstreamResolverBase,
-		lIP:                  ip,
-		lNet:                 net,
-		interfaceName:        interfaceName,
+		lIP:                  wgIface.Address().IP,
+		lNet:                 wgIface.Address().Network,
+		interfaceName:        wgIface.Name(),
 	}
 	ios.upstreamClient = ios
 

client/internal/dns/upstream_test.go

@@ -2,13 +2,17 @@ package dns
 
 import (
 	"context"
+	"net"
 	"net/netip"
 	"strings"
 	"testing"
 	"time"
 
 	"github.com/miekg/dns"
+	"golang.zx2c4.com/wireguard/tun/netstack"
 
+	"github.com/netbirdio/netbird/client/iface/device"
+	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	"github.com/netbirdio/netbird/client/internal/dns/test"
 )
 
@@ -58,7 +62,7 @@ func TestUpstreamResolver_ServeDNS(t *testing.T) {
 	for _, testCase := range testCases {
 		t.Run(testCase.name, func(t *testing.T) {
 			ctx, cancel := context.WithCancel(context.TODO())
-			resolver, _ := newUpstreamResolver(ctx, "", netip.Addr{}, netip.Prefix{}, nil, nil, ".")
+			resolver, _ := newUpstreamResolver(ctx, &mockNetstackProvider{}, nil, nil, ".")
 			// Convert test servers to netip.AddrPort
 			var servers []netip.AddrPort
 			for _, server := range testCase.InputServers {
@@ -112,6 +116,19 @@ func TestUpstreamResolver_ServeDNS(t *testing.T) {
 	}
 }
 
+type mockNetstackProvider struct{}
+
+func (m *mockNetstackProvider) Name() string                      { return "mock" }
+func (m *mockNetstackProvider) Address() wgaddr.Address           { return wgaddr.Address{} }
+func (m *mockNetstackProvider) ToInterface() *net.Interface       { return nil }
+func (m *mockNetstackProvider) IsUserspaceBind() bool             { return false }
+func (m *mockNetstackProvider) GetFilter() device.PacketFilter    { return nil }
+func (m *mockNetstackProvider) GetDevice() *device.FilteredDevice { return nil }
+func (m *mockNetstackProvider) GetNet() *netstack.Net             { return nil }
+func (m *mockNetstackProvider) GetInterfaceGUIDString() (string, error) {
+	return "", nil
+}
+
 type mockUpstreamResolver struct {
 	r   *dns.Msg
 	rtt time.Duration

client/internal/dns/wgiface.go

@@ -5,6 +5,8 @@ package dns
 import (
 	"net"
 
+	"golang.zx2c4.com/wireguard/tun/netstack"
+
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 )
@@ -17,4 +19,5 @@ type WGIface interface {
 	IsUserspaceBind() bool
 	GetFilter() device.PacketFilter
 	GetDevice() *device.FilteredDevice
+	GetNet() *netstack.Net
 }

client/internal/dns/wgiface_windows.go

@@ -1,6 +1,8 @@
 package dns
 
 import (
+	"golang.zx2c4.com/wireguard/tun/netstack"
+
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 )
@@ -12,5 +14,6 @@ type WGIface interface {
 	IsUserspaceBind() bool
 	GetFilter() device.PacketFilter
 	GetDevice() *device.FilteredDevice
+	GetNet() *netstack.Net
 	GetInterfaceGUIDString() (string, error)
 }

client/internal/dnsfwd/forwarder.go

@@ -18,6 +18,7 @@ import (
 
 	nberrors "github.com/netbirdio/netbird/client/errors"
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
+	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/client/internal/peer"
 	"github.com/netbirdio/netbird/route"
 )
@@ -189,29 +190,22 @@ func (f *DNSForwarder) Close(ctx context.Context) error {
 	return nberrors.FormatErrorOrNil(result)
 }
 
-func (f *DNSForwarder) handleDNSQuery(w dns.ResponseWriter, query *dns.Msg) *dns.Msg {
+func (f *DNSForwarder) handleDNSQuery(logger *log.Entry, w dns.ResponseWriter, query *dns.Msg) *dns.Msg {
 	if len(query.Question) == 0 {
 		return nil
 	}
 	question := query.Question[0]
-	log.Tracef("received DNS request for DNS forwarder: domain=%v type=%v class=%v",
-		question.Name, question.Qtype, question.Qclass)
+	logger.Tracef("received DNS request for DNS forwarder: domain=%s type=%s class=%s",
+		question.Name, dns.TypeToString[question.Qtype], dns.ClassToString[question.Qclass])
 
 	domain := strings.ToLower(question.Name)
 
 	resp := query.SetReply(query)
-	var network string
-	switch question.Qtype {
-	case dns.TypeA:
-		network = "ip4"
-	case dns.TypeAAAA:
-		network = "ip6"
-	default:
-		// TODO: Handle other types
-
+	network := resutil.NetworkForQtype(question.Qtype)
+	if network == "" {
 		resp.Rcode = dns.RcodeNotImplemented
 		if err := w.WriteMsg(resp); err != nil {
-			log.Errorf("failed to write DNS response: %v", err)
+			logger.Errorf("failed to write DNS response: %v", err)
 		}
 		return nil
 	}
@@ -221,33 +215,35 @@ func (f *DNSForwarder) handleDNSQuery(w dns.ResponseWriter, query *dns.Msg) *dns
 	if mostSpecificResId == "" {
 		resp.Rcode = dns.RcodeRefused
 		if err := w.WriteMsg(resp); err != nil {
-			log.Errorf("failed to write DNS response: %v", err)
+			logger.Errorf("failed to write DNS response: %v", err)
 		}
 		return nil
 	}
 
 	ctx, cancel := context.WithTimeout(context.Background(), upstreamTimeout)
 	defer cancel()
-	ips, err := f.resolver.LookupNetIP(ctx, network, domain)
-	if err != nil {
-		f.handleDNSError(ctx, w, question, resp, domain, err)
+
+	result := resutil.LookupIP(ctx, f.resolver, network, domain, question.Qtype)
+	if result.Err != nil {
+		f.handleDNSError(ctx, logger, w, question, resp, domain, result)
 		return nil
 	}
 
-	// Unmap IPv4-mapped IPv6 addresses that some resolvers may return
-	for i, ip := range ips {
-		ips[i] = ip.Unmap()
-	}
-
-	f.updateInternalState(ips, mostSpecificResId, matchingEntries)
-	f.addIPsToResponse(resp, domain, ips)
-	f.cache.set(domain, question.Qtype, ips)
+	f.updateInternalState(result.IPs, mostSpecificResId, matchingEntries)
+	resp.Answer = append(resp.Answer, resutil.IPsToRRs(domain, result.IPs, f.ttl)...)
+	f.cache.set(domain, question.Qtype, result.IPs)
 
 	return resp
 }
 
 func (f *DNSForwarder) handleDNSQueryUDP(w dns.ResponseWriter, query *dns.Msg) {
-	resp := f.handleDNSQuery(w, query)
+	startTime := time.Now()
+	logger := log.WithFields(log.Fields{
+		"request_id": resutil.GenerateRequestID(),
+		"dns_id":     fmt.Sprintf("%04x", query.Id),
+	})
+
+	resp := f.handleDNSQuery(logger, w, query)
 	if resp == nil {
 		return
 	}
@@ -265,19 +261,33 @@ func (f *DNSForwarder) handleDNSQueryUDP(w dns.ResponseWriter, query *dns.Msg) {
 	}
 
 	if err := w.WriteMsg(resp); err != nil {
-		log.Errorf("failed to write DNS response: %v", err)
+		logger.Errorf("failed to write DNS response: %v", err)
+		return
 	}
+
+	logger.Tracef("response: domain=%s rcode=%s answers=%s took=%s",
+		query.Question[0].Name, dns.RcodeToString[resp.Rcode], resutil.FormatAnswers(resp.Answer), time.Since(startTime))
 }
 
 func (f *DNSForwarder) handleDNSQueryTCP(w dns.ResponseWriter, query *dns.Msg) {
-	resp := f.handleDNSQuery(w, query)
+	startTime := time.Now()
+	logger := log.WithFields(log.Fields{
+		"request_id": resutil.GenerateRequestID(),
+		"dns_id":     fmt.Sprintf("%04x", query.Id),
+	})
+
+	resp := f.handleDNSQuery(logger, w, query)
 	if resp == nil {
 		return
 	}
 
 	if err := w.WriteMsg(resp); err != nil {
-		log.Errorf("failed to write DNS response: %v", err)
+		logger.Errorf("failed to write DNS response: %v", err)
+		return
 	}
+
+	logger.Tracef("response: domain=%s rcode=%s answers=%s took=%s",
+		query.Question[0].Name, dns.RcodeToString[resp.Rcode], resutil.FormatAnswers(resp.Answer), time.Since(startTime))
 }
 
 func (f *DNSForwarder) updateInternalState(ips []netip.Addr, mostSpecificResId route.ResID, matchingEntries []*ForwarderEntry) {
@@ -315,140 +325,64 @@ func (f *DNSForwarder) updateFirewall(matchingEntries []*ForwarderEntry, prefixe
 	}
 }
 
-// setResponseCodeForNotFound determines and sets the appropriate response code when IsNotFound is true
-// It distinguishes between NXDOMAIN (domain doesn't exist) and NODATA (domain exists but no records of requested type)
-//
-// LIMITATION: This function only checks A and AAAA record types to determine domain existence.
-// If a domain has only other record types (MX, TXT, CNAME, etc.) but no A/AAAA records,
-// it may incorrectly return NXDOMAIN instead of NODATA. This is acceptable since the forwarder
-// only handles A/AAAA queries and returns NOTIMP for other types.
-func (f *DNSForwarder) setResponseCodeForNotFound(ctx context.Context, resp *dns.Msg, domain string, originalQtype uint16) {
-	// Try querying for a different record type to see if the domain exists
-	// If the original query was for AAAA, try A. If it was for A, try AAAA.
-	// This helps distinguish between NXDOMAIN and NODATA.
-	var alternativeNetwork string
-	switch originalQtype {
-	case dns.TypeAAAA:
-		alternativeNetwork = "ip4"
-	case dns.TypeA:
-		alternativeNetwork = "ip6"
-	default:
-		resp.Rcode = dns.RcodeNameError
-		return
-	}
-
-	if _, err := f.resolver.LookupNetIP(ctx, alternativeNetwork, domain); err != nil {
-		var dnsErr *net.DNSError
-		if errors.As(err, &dnsErr) && dnsErr.IsNotFound {
-			// Alternative query also returned not found - domain truly doesn't exist
-			resp.Rcode = dns.RcodeNameError
-			return
-		}
-		// Some other error (timeout, server failure, etc.) - can't determine, assume domain exists
-		resp.Rcode = dns.RcodeSuccess
-		return
-	}
-
-	// Alternative query succeeded - domain exists but has no records of this type
-	resp.Rcode = dns.RcodeSuccess
-}
-
 // handleDNSError processes DNS lookup errors and sends an appropriate error response.
 func (f *DNSForwarder) handleDNSError(
 	ctx context.Context,
+	logger *log.Entry,
 	w dns.ResponseWriter,
 	question dns.Question,
 	resp *dns.Msg,
 	domain string,
-	err error,
+	result resutil.LookupResult,
 ) {
-	// Default to SERVFAIL; override below when appropriate.
-	resp.Rcode = dns.RcodeServerFailure
-
 	qType := question.Qtype
 	qTypeName := dns.TypeToString[qType]
 
-	// Prefer typed DNS errors; fall back to generic logging otherwise.
-	var dnsErr *net.DNSError
-	if !errors.As(err, &dnsErr) {
-		log.Warnf(errResolveFailed, domain, err)
-		if writeErr := w.WriteMsg(resp); writeErr != nil {
-			log.Errorf("failed to write failure DNS response: %v", writeErr)
-		}
-		return
-	}
+	resp.Rcode = result.Rcode
 
-	// NotFound: set NXDOMAIN / appropriate code via helper.
-	if dnsErr.IsNotFound {
-		f.setResponseCodeForNotFound(ctx, resp, domain, qType)
-		if writeErr := w.WriteMsg(resp); writeErr != nil {
-			log.Errorf("failed to write failure DNS response: %v", writeErr)
-		}
+	// NotFound: cache negative result and respond
+	if result.Rcode == dns.RcodeNameError || result.Rcode == dns.RcodeSuccess {
 		f.cache.set(domain, question.Qtype, nil)
+		if writeErr := w.WriteMsg(resp); writeErr != nil {
+			logger.Errorf("failed to write failure DNS response: %v", writeErr)
+		}
 		return
 	}
 
 	// Upstream failed but we might have a cached answer—serve it if present.
 	if ips, ok := f.cache.get(domain, qType); ok {
 		if len(ips) > 0 {
-			log.Debugf("serving cached DNS response after upstream failure: domain=%s type=%s", domain, qTypeName)
-			f.addIPsToResponse(resp, domain, ips)
+			logger.Debugf("serving cached DNS response after upstream failure: domain=%s type=%s", domain, qTypeName)
+			resp.Answer = append(resp.Answer, resutil.IPsToRRs(domain, ips, f.ttl)...)
 			resp.Rcode = dns.RcodeSuccess
 			if writeErr := w.WriteMsg(resp); writeErr != nil {
-				log.Errorf("failed to write cached DNS response: %v", writeErr)
-			}
-		} else { // send NXDOMAIN / appropriate code if cache is empty
-			f.setResponseCodeForNotFound(ctx, resp, domain, qType)
-			if writeErr := w.WriteMsg(resp); writeErr != nil {
-				log.Errorf("failed to write failure DNS response: %v", writeErr)
+				logger.Errorf("failed to write cached DNS response: %v", writeErr)
 			}
+			return
+		}
+
+		// Cached negative result - re-verify NXDOMAIN vs NODATA
+		verifyResult := resutil.LookupIP(ctx, f.resolver, resutil.NetworkForQtype(qType), domain, qType)
+		if verifyResult.Rcode == dns.RcodeNameError || verifyResult.Rcode == dns.RcodeSuccess {
+			resp.Rcode = verifyResult.Rcode
+			if writeErr := w.WriteMsg(resp); writeErr != nil {
+				logger.Errorf("failed to write failure DNS response: %v", writeErr)
+			}
+			return
 		}
-		return
 	}
 
-	// No cache. Log with or without the server field for more context.
-	if dnsErr.Server != "" {
-		log.Warnf("failed to resolve: type=%s domain=%s server=%s: %v", qTypeName, domain, dnsErr.Server, err)
+	// No cache or verification failed. Log with or without the server field for more context.
+	var dnsErr *net.DNSError
+	if errors.As(result.Err, &dnsErr) && dnsErr.Server != "" {
+		logger.Warnf("failed to resolve: type=%s domain=%s server=%s: %v", qTypeName, domain, dnsErr.Server, result.Err)
 	} else {
-		log.Warnf(errResolveFailed, domain, err)
+		logger.Warnf(errResolveFailed, domain, result.Err)
 	}
 
 	// Write final failure response.
 	if writeErr := w.WriteMsg(resp); writeErr != nil {
-		log.Errorf("failed to write failure DNS response: %v", writeErr)
-	}
-}
-
-// addIPsToResponse adds IP addresses to the DNS response as appropriate A or AAAA records
-func (f *DNSForwarder) addIPsToResponse(resp *dns.Msg, domain string, ips []netip.Addr) {
-	for _, ip := range ips {
-		var respRecord dns.RR
-		if ip.Is6() {
-			log.Tracef("resolved domain=%s to IPv6=%s", domain, ip)
-			rr := dns.AAAA{
-				AAAA: ip.AsSlice(),
-				Hdr: dns.RR_Header{
-					Name:   domain,
-					Rrtype: dns.TypeAAAA,
-					Class:  dns.ClassINET,
-					Ttl:    f.ttl,
-				},
-			}
-			respRecord = &rr
-		} else {
-			log.Tracef("resolved domain=%s to IPv4=%s", domain, ip)
-			rr := dns.A{
-				A: ip.AsSlice(),
-				Hdr: dns.RR_Header{
-					Name:   domain,
-					Rrtype: dns.TypeA,
-					Class:  dns.ClassINET,
-					Ttl:    f.ttl,
-				},
-			}
-			respRecord = &rr
-		}
-		resp.Answer = append(resp.Answer, respRecord)
+		logger.Errorf("failed to write failure DNS response: %v", writeErr)
 	}
 }
 

client/internal/dnsfwd/forwarder_test.go

@@ -10,6 +10,7 @@ import (
 	"time"
 
 	"github.com/miekg/dns"
+	log "github.com/sirupsen/logrus"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/mock"
 	"github.com/stretchr/testify/require"
@@ -317,7 +318,7 @@ func TestDNSForwarder_UnauthorizedDomainAccess(t *testing.T) {
 			query.SetQuestion(dns.Fqdn(tt.queryDomain), dns.TypeA)
 
 			mockWriter := &test.MockResponseWriter{}
-			resp := forwarder.handleDNSQuery(mockWriter, query)
+			resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 			if tt.shouldResolve {
 				require.NotNil(t, resp, "Expected response for authorized domain")
@@ -465,7 +466,7 @@ func TestDNSForwarder_FirewallSetUpdates(t *testing.T) {
 			dnsQuery.SetQuestion(dns.Fqdn(tt.query), dns.TypeA)
 
 			mockWriter := &test.MockResponseWriter{}
-			resp := forwarder.handleDNSQuery(mockWriter, dnsQuery)
+			resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, dnsQuery)
 
 			// Verify response
 			if tt.shouldResolve {
@@ -527,7 +528,7 @@ func TestDNSForwarder_MultipleIPsInSingleUpdate(t *testing.T) {
 	query.SetQuestion("example.com.", dns.TypeA)
 
 	mockWriter := &test.MockResponseWriter{}
-	resp := forwarder.handleDNSQuery(mockWriter, query)
+	resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 	// Verify response contains all IPs
 	require.NotNil(t, resp)
@@ -604,7 +605,7 @@ func TestDNSForwarder_ResponseCodes(t *testing.T) {
 				},
 			}
 
-			_ = forwarder.handleDNSQuery(mockWriter, query)
+			_ = forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 			// Check the response written to the writer
 			require.NotNil(t, writtenResp, "Expected response to be written")
@@ -674,7 +675,7 @@ func TestDNSForwarder_ServeFromCacheOnUpstreamFailure(t *testing.T) {
 	q1 := &dns.Msg{}
 	q1.SetQuestion(dns.Fqdn("example.com"), dns.TypeA)
 	w1 := &test.MockResponseWriter{}
-	resp1 := forwarder.handleDNSQuery(w1, q1)
+	resp1 := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), w1, q1)
 	require.NotNil(t, resp1)
 	require.Equal(t, dns.RcodeSuccess, resp1.Rcode)
 	require.Len(t, resp1.Answer, 1)
@@ -684,7 +685,7 @@ func TestDNSForwarder_ServeFromCacheOnUpstreamFailure(t *testing.T) {
 	q2.SetQuestion(dns.Fqdn("example.com"), dns.TypeA)
 	var writtenResp *dns.Msg
 	w2 := &test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { writtenResp = m; return nil }}
-	_ = forwarder.handleDNSQuery(w2, q2)
+	_ = forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), w2, q2)
 
 	require.NotNil(t, writtenResp, "expected response to be written")
 	require.Equal(t, dns.RcodeSuccess, writtenResp.Rcode)
@@ -714,7 +715,7 @@ func TestDNSForwarder_CacheNormalizationCasingAndDot(t *testing.T) {
 	q1 := &dns.Msg{}
 	q1.SetQuestion(mixedQuery+".", dns.TypeA)
 	w1 := &test.MockResponseWriter{}
-	resp1 := forwarder.handleDNSQuery(w1, q1)
+	resp1 := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), w1, q1)
 	require.NotNil(t, resp1)
 	require.Equal(t, dns.RcodeSuccess, resp1.Rcode)
 	require.Len(t, resp1.Answer, 1)
@@ -728,7 +729,7 @@ func TestDNSForwarder_CacheNormalizationCasingAndDot(t *testing.T) {
 	q2.SetQuestion("EXAMPLE.COM", dns.TypeA)
 	var writtenResp *dns.Msg
 	w2 := &test.MockResponseWriter{WriteMsgFunc: func(m *dns.Msg) error { writtenResp = m; return nil }}
-	_ = forwarder.handleDNSQuery(w2, q2)
+	_ = forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), w2, q2)
 
 	require.NotNil(t, writtenResp)
 	require.Equal(t, dns.RcodeSuccess, writtenResp.Rcode)
@@ -783,7 +784,7 @@ func TestDNSForwarder_MultipleOverlappingPatterns(t *testing.T) {
 	query.SetQuestion("smtp.mail.example.com.", dns.TypeA)
 
 	mockWriter := &test.MockResponseWriter{}
-	resp := forwarder.handleDNSQuery(mockWriter, query)
+	resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 	require.NotNil(t, resp)
 	assert.Equal(t, dns.RcodeSuccess, resp.Rcode)
@@ -904,7 +905,7 @@ func TestDNSForwarder_NodataVsNxdomain(t *testing.T) {
 				},
 			}
 
-			resp := forwarder.handleDNSQuery(mockWriter, query)
+			resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 			// If a response was returned, it means it should be written (happens in wrapper functions)
 			if resp != nil && writtenResp == nil {
@@ -937,7 +938,7 @@ func TestDNSForwarder_EmptyQuery(t *testing.T) {
 			return nil
 		},
 	}
-	resp := forwarder.handleDNSQuery(mockWriter, query)
+	resp := forwarder.handleDNSQuery(log.NewEntry(log.StandardLogger()), mockWriter, query)
 
 	assert.Nil(t, resp, "Should return nil for empty query")
 	assert.False(t, writeCalled, "Should not write response for empty query")

client/internal/engine.go

@@ -1251,11 +1251,16 @@ func toDNSConfig(protoDNSConfig *mgmProto.DNSConfig, network netip.Prefix) nbdns
 		ForwarderPort:    forwarderPort,
 	}
 
-	for _, zone := range protoDNSConfig.GetCustomZones() {
+	protoZones := protoDNSConfig.GetCustomZones()
+	// Treat single zone as authoritative for backward compatibility with old servers
+	// that only send the peer FQDN zone without setting field 4.
+	singleZoneCompat := len(protoZones) == 1
+
+	for _, zone := range protoZones {
 		dnsZone := nbdns.CustomZone{
 			Domain:               zone.GetDomain(),
 			SearchDomainDisabled: zone.GetSearchDomainDisabled(),
-			SkipPTRProcess:       zone.GetSkipPTRProcess(),
+			NonAuthoritative:     zone.GetNonAuthoritative() && !singleZoneCompat,
 		}
 		for _, record := range zone.Records {
 			dnsRecord := nbdns.SimpleRecord{
@@ -1743,22 +1748,26 @@ func (e *Engine) RunHealthProbes(waitForResult bool) bool {
 	}
 
 	e.syncMsgMux.Unlock()
-	var results []relay.ProbeResult
-	if waitForResult {
-		results = e.probeStunTurn.ProbeAllWaitResult(e.ctx, stuns, turns)
-	} else {
-		results = e.probeStunTurn.ProbeAll(e.ctx, stuns, turns)
-	}
-	e.statusRecorder.UpdateRelayStates(results)
 
+	// Skip STUN/TURN probing for JS/WASM as it's not available
 	relayHealthy := true
-	for _, res := range results {
-		if res.Err != nil {
-			relayHealthy = false
-			break
+	if runtime.GOOS != "js" {
+		var results []relay.ProbeResult
+		if waitForResult {
+			results = e.probeStunTurn.ProbeAllWaitResult(e.ctx, stuns, turns)
+		} else {
+			results = e.probeStunTurn.ProbeAll(e.ctx, stuns, turns)
 		}
+		e.statusRecorder.UpdateRelayStates(results)
+
+		for _, res := range results {
+			if res.Err != nil {
+				relayHealthy = false
+				break
+			}
+		}
+		log.Debugf("relay health check: healthy=%t", relayHealthy)
 	}
-	log.Debugf("relay health check: healthy=%t", relayHealthy)
 
 	allHealthy := signalHealthy && managementHealthy && relayHealthy
 	log.Debugf("all health checks completed: healthy=%t", allHealthy)

client/internal/peer/conn.go

@@ -669,10 +669,17 @@ func (conn *Conn) isConnectedOnAllWay() (connected bool) {
 		}
 	}()
 
-	if runtime.GOOS != "js" && conn.statusICE.Get() == worker.StatusDisconnected && !conn.workerICE.InProgress() {
+	// For JS platform: only relay connection is supported
+	if runtime.GOOS == "js" {
+		return conn.statusRelay.Get() == worker.StatusConnected
+	}
+
+	// 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

client/internal/peer/status.go

@@ -14,6 +14,7 @@ import (
 	"golang.org/x/exp/maps"
 	"google.golang.org/grpc/codes"
 	gstatus "google.golang.org/grpc/status"
+	"google.golang.org/protobuf/types/known/durationpb"
 	"google.golang.org/protobuf/types/known/timestamppb"
 
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
@@ -158,6 +159,7 @@ type FullStatus struct {
 	NSGroupStates         []NSGroupState
 	NumOfForwardingRules  int
 	LazyConnectionEnabled bool
+	Events                []*proto.SystemEvent
 }
 
 type StatusChangeSubscription struct {
@@ -981,6 +983,7 @@ func (d *Status) GetFullStatus() FullStatus {
 	}
 
 	fullStatus.Peers = append(fullStatus.Peers, d.offlinePeers...)
+	fullStatus.Events = d.GetEventHistory()
 	return fullStatus
 }
 
@@ -1181,3 +1184,97 @@ type EventSubscription struct {
 func (s *EventSubscription) Events() <-chan *proto.SystemEvent {
 	return s.events
 }
+
+// ToProto converts FullStatus to proto.FullStatus.
+func (fs FullStatus) ToProto() *proto.FullStatus {
+	pbFullStatus := proto.FullStatus{
+		ManagementState: &proto.ManagementState{},
+		SignalState:     &proto.SignalState{},
+		LocalPeerState:  &proto.LocalPeerState{},
+		Peers:           []*proto.PeerState{},
+	}
+
+	pbFullStatus.ManagementState.URL = fs.ManagementState.URL
+	pbFullStatus.ManagementState.Connected = fs.ManagementState.Connected
+	if err := fs.ManagementState.Error; err != nil {
+		pbFullStatus.ManagementState.Error = err.Error()
+	}
+
+	pbFullStatus.SignalState.URL = fs.SignalState.URL
+	pbFullStatus.SignalState.Connected = fs.SignalState.Connected
+	if err := fs.SignalState.Error; err != nil {
+		pbFullStatus.SignalState.Error = err.Error()
+	}
+
+	pbFullStatus.LocalPeerState.IP = fs.LocalPeerState.IP
+	pbFullStatus.LocalPeerState.PubKey = fs.LocalPeerState.PubKey
+	pbFullStatus.LocalPeerState.KernelInterface = fs.LocalPeerState.KernelInterface
+	pbFullStatus.LocalPeerState.Fqdn = fs.LocalPeerState.FQDN
+	pbFullStatus.LocalPeerState.RosenpassPermissive = fs.RosenpassState.Permissive
+	pbFullStatus.LocalPeerState.RosenpassEnabled = fs.RosenpassState.Enabled
+	pbFullStatus.NumberOfForwardingRules = int32(fs.NumOfForwardingRules)
+	pbFullStatus.LazyConnectionEnabled = fs.LazyConnectionEnabled
+
+	pbFullStatus.LocalPeerState.Networks = maps.Keys(fs.LocalPeerState.Routes)
+
+	for _, peerState := range fs.Peers {
+		networks := maps.Keys(peerState.GetRoutes())
+
+		pbPeerState := &proto.PeerState{
+			IP:                         peerState.IP,
+			PubKey:                     peerState.PubKey,
+			ConnStatus:                 peerState.ConnStatus.String(),
+			ConnStatusUpdate:           timestamppb.New(peerState.ConnStatusUpdate),
+			Relayed:                    peerState.Relayed,
+			LocalIceCandidateType:      peerState.LocalIceCandidateType,
+			RemoteIceCandidateType:     peerState.RemoteIceCandidateType,
+			LocalIceCandidateEndpoint:  peerState.LocalIceCandidateEndpoint,
+			RemoteIceCandidateEndpoint: peerState.RemoteIceCandidateEndpoint,
+			RelayAddress:               peerState.RelayServerAddress,
+			Fqdn:                       peerState.FQDN,
+			LastWireguardHandshake:     timestamppb.New(peerState.LastWireguardHandshake),
+			BytesRx:                    peerState.BytesRx,
+			BytesTx:                    peerState.BytesTx,
+			RosenpassEnabled:           peerState.RosenpassEnabled,
+			Networks:                   networks,
+			Latency:                    durationpb.New(peerState.Latency),
+			SshHostKey:                 peerState.SSHHostKey,
+		}
+		pbFullStatus.Peers = append(pbFullStatus.Peers, pbPeerState)
+	}
+
+	for _, relayState := range fs.Relays {
+		pbRelayState := &proto.RelayState{
+			URI:       relayState.URI,
+			Available: relayState.Err == nil,
+		}
+		if err := relayState.Err; err != nil {
+			pbRelayState.Error = err.Error()
+		}
+		pbFullStatus.Relays = append(pbFullStatus.Relays, pbRelayState)
+	}
+
+	for _, dnsState := range fs.NSGroupStates {
+		var err string
+		if dnsState.Error != nil {
+			err = dnsState.Error.Error()
+		}
+
+		var servers []string
+		for _, server := range dnsState.Servers {
+			servers = append(servers, server.String())
+		}
+
+		pbDnsState := &proto.NSGroupState{
+			Servers: servers,
+			Domains: dnsState.Domains,
+			Enabled: dnsState.Enabled,
+			Error:   err,
+		}
+		pbFullStatus.DnsServers = append(pbFullStatus.DnsServers, pbDnsState)
+	}
+
+	pbFullStatus.Events = fs.Events
+
+	return &pbFullStatus
+}

client/internal/routemanager/dnsinterceptor/handler.go

@@ -17,12 +17,13 @@ import (
 
 	nberrors "github.com/netbirdio/netbird/client/errors"
 	firewall "github.com/netbirdio/netbird/client/firewall/manager"
-	"github.com/netbirdio/netbird/client/iface/wgaddr"
 	nbdns "github.com/netbirdio/netbird/client/internal/dns"
+	"github.com/netbirdio/netbird/client/internal/dns/resutil"
 	"github.com/netbirdio/netbird/client/internal/peer"
 	"github.com/netbirdio/netbird/client/internal/peerstore"
 	"github.com/netbirdio/netbird/client/internal/routemanager/common"
 	"github.com/netbirdio/netbird/client/internal/routemanager/fakeip"
+	iface "github.com/netbirdio/netbird/client/internal/routemanager/iface"
 	"github.com/netbirdio/netbird/client/internal/routemanager/refcounter"
 	"github.com/netbirdio/netbird/route"
 	"github.com/netbirdio/netbird/shared/management/domain"
@@ -37,11 +38,6 @@ type internalDNATer interface {
 	AddInternalDNATMapping(netip.Addr, netip.Addr) error
 }
 
-type wgInterface interface {
-	Name() string
-	Address() wgaddr.Address
-}
-
 type DnsInterceptor struct {
 	mu                   sync.RWMutex
 	route                *route.Route
@@ -51,7 +47,7 @@ type DnsInterceptor struct {
 	dnsServer            nbdns.Server
 	currentPeerKey       string
 	interceptedDomains   domainMap
-	wgInterface          wgInterface
+	wgInterface          iface.WGIface
 	peerStore            *peerstore.Store
 	firewall             firewall.Manager
 	fakeIPManager        *fakeip.Manager
@@ -219,14 +215,14 @@ func (d *DnsInterceptor) RemoveAllowedIPs() error {
 
 // ServeDNS implements the dns.Handler interface
 func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
-	requestID := nbdns.GenerateRequestID()
-	logger := log.WithField("request_id", requestID)
+	logger := log.WithFields(log.Fields{
+		"request_id": resutil.GetRequestID(w),
+		"dns_id":     fmt.Sprintf("%04x", r.Id),
+	})
 
 	if len(r.Question) == 0 {
 		return
 	}
-	logger.Tracef("received DNS request for domain=%s type=%v class=%v",
-		r.Question[0].Name, r.Question[0].Qtype, r.Question[0].Qclass)
 
 	// pass if non A/AAAA query
 	if r.Question[0].Qtype != dns.TypeA && r.Question[0].Qtype != dns.TypeAAAA {
@@ -249,12 +245,6 @@ func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 		return
 	}
 
-	client, err := nbdns.GetClientPrivate(d.wgInterface.Address().IP, d.wgInterface.Name(), dnsTimeout)
-	if err != nil {
-		d.writeDNSError(w, r, logger, fmt.Sprintf("create DNS client: %v", err))
-		return
-	}
-
 	if r.Extra == nil {
 		r.MsgHdr.AuthenticatedData = true
 	}
@@ -263,32 +253,15 @@ func (d *DnsInterceptor) ServeDNS(w dns.ResponseWriter, r *dns.Msg) {
 	ctx, cancel := context.WithTimeout(context.Background(), dnsTimeout)
 	defer cancel()
 
-	startTime := time.Now()
-	reply, _, err := nbdns.ExchangeWithFallback(ctx, client, r, upstream)
-	if err != nil {
-		if errors.Is(err, context.DeadlineExceeded) {
-			elapsed := time.Since(startTime)
-			peerInfo := d.debugPeerTimeout(upstreamIP, peerKey)
-			logger.Errorf("peer DNS timeout after %v (timeout=%v) for domain=%s to peer %s (%s)%s - error: %v",
-				elapsed.Truncate(time.Millisecond), dnsTimeout, r.Question[0].Name, upstreamIP.String(), peerKey, peerInfo, err)
-		} else {
-			logger.Errorf("failed to exchange DNS request with %s (%s) for domain=%s: %v", upstreamIP.String(), peerKey, r.Question[0].Name, err)
-		}
-		if err := w.WriteMsg(&dns.Msg{MsgHdr: dns.MsgHdr{Rcode: dns.RcodeServerFailure, Id: r.Id}}); err != nil {
-			logger.Errorf("failed writing DNS response: %v", err)
-		}
+	reply := d.queryUpstreamDNS(ctx, w, r, upstream, upstreamIP, peerKey, logger)
+	if reply == nil {
 		return
 	}
 
-	var answer []dns.RR
-	if reply != nil {
-		answer = reply.Answer
-	}
-
-	logger.Tracef("upstream %s (%s) DNS response for domain=%s answers=%v", upstreamIP.String(), peerKey, r.Question[0].Name, answer)
+	resutil.SetMeta(w, "peer", peerKey)
 
 	reply.Id = r.Id
-	if err := d.writeMsg(w, reply); err != nil {
+	if err := d.writeMsg(w, reply, logger); err != nil {
 		logger.Errorf("failed writing DNS response: %v", err)
 	}
 }
@@ -324,11 +297,15 @@ func (d *DnsInterceptor) getUpstreamIP(peerKey string) (netip.Addr, error) {
 	return peerAllowedIP, nil
 }
 
-func (d *DnsInterceptor) writeMsg(w dns.ResponseWriter, r *dns.Msg) error {
+func (d *DnsInterceptor) writeMsg(w dns.ResponseWriter, r *dns.Msg, logger *log.Entry) error {
 	if r == nil {
 		return fmt.Errorf("received nil DNS message")
 	}
 
+	// Clear Zero bit from peer responses to prevent external sources from
+	// manipulating our internal fallthrough signaling mechanism
+	r.MsgHdr.Zero = false
+
 	if len(r.Answer) > 0 && len(r.Question) > 0 {
 		origPattern := ""
 		if writer, ok := w.(*nbdns.ResponseWriterChain); ok {
@@ -350,14 +327,14 @@ func (d *DnsInterceptor) writeMsg(w dns.ResponseWriter, r *dns.Msg) error {
 			case *dns.A:
 				addr, ok := netip.AddrFromSlice(rr.A)
 				if !ok {
-					log.Tracef("failed to convert A record for domain=%s ip=%v", resolvedDomain, rr.A)
+					logger.Tracef("failed to convert A record for domain=%s ip=%v", resolvedDomain, rr.A)
 					continue
 				}
 				ip = addr
 			case *dns.AAAA:
 				addr, ok := netip.AddrFromSlice(rr.AAAA)
 				if !ok {
-					log.Tracef("failed to convert AAAA record for domain=%s ip=%v", resolvedDomain, rr.AAAA)
+					logger.Tracef("failed to convert AAAA record for domain=%s ip=%v", resolvedDomain, rr.AAAA)
 					continue
 				}
 				ip = addr
@@ -370,11 +347,11 @@ func (d *DnsInterceptor) writeMsg(w dns.ResponseWriter, r *dns.Msg) error {
 		}
 
 		if len(newPrefixes) > 0 {
-			if err := d.updateDomainPrefixes(resolvedDomain, originalDomain, newPrefixes); err != nil {
-				log.Errorf("failed to update domain prefixes: %v", err)
+			if err := d.updateDomainPrefixes(resolvedDomain, originalDomain, newPrefixes, logger); err != nil {
+				logger.Errorf("failed to update domain prefixes: %v", err)
 			}
 
-			d.replaceIPsInDNSResponse(r, newPrefixes)
+			d.replaceIPsInDNSResponse(r, newPrefixes, logger)
 		}
 	}
 
@@ -386,22 +363,22 @@ func (d *DnsInterceptor) writeMsg(w dns.ResponseWriter, r *dns.Msg) error {
 }
 
 // logPrefixChanges handles the logging for prefix changes
-func (d *DnsInterceptor) logPrefixChanges(resolvedDomain, originalDomain domain.Domain, toAdd, toRemove []netip.Prefix) {
+func (d *DnsInterceptor) logPrefixChanges(resolvedDomain, originalDomain domain.Domain, toAdd, toRemove []netip.Prefix, logger *log.Entry) {
 	if len(toAdd) > 0 {
-		log.Debugf("added dynamic route(s) for domain=%s (pattern: domain=%s): %s",
+		logger.Debugf("added dynamic route(s) for domain=%s (pattern: domain=%s): %s",
 			resolvedDomain.SafeString(),
 			originalDomain.SafeString(),
 			toAdd)
 	}
 	if len(toRemove) > 0 && !d.route.KeepRoute {
-		log.Debugf("removed dynamic route(s) for domain=%s (pattern: domain=%s): %s",
+		logger.Debugf("removed dynamic route(s) for domain=%s (pattern: domain=%s): %s",
 			resolvedDomain.SafeString(),
 			originalDomain.SafeString(),
 			toRemove)
 	}
 }
 
-func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain domain.Domain, newPrefixes []netip.Prefix) error {
+func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain domain.Domain, newPrefixes []netip.Prefix, logger *log.Entry) error {
 	d.mu.Lock()
 	defer d.mu.Unlock()
 
@@ -418,9 +395,9 @@ func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain dom
 			realIP := prefix.Addr()
 			if fakeIP, err := d.fakeIPManager.AllocateFakeIP(realIP); err == nil {
 				dnatMappings[fakeIP] = realIP
-				log.Tracef("allocated fake IP %s for real IP %s", fakeIP, realIP)
+				logger.Tracef("allocated fake IP %s for real IP %s", fakeIP, realIP)
 			} else {
-				log.Errorf("Failed to allocate fake IP for %s: %v", realIP, err)
+				logger.Errorf("failed to allocate fake IP for %s: %v", realIP, err)
 			}
 		}
 	}
@@ -432,7 +409,7 @@ func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain dom
 		}
 	}
 
-	d.addDNATMappings(dnatMappings)
+	d.addDNATMappings(dnatMappings, logger)
 
 	if !d.route.KeepRoute {
 		// Remove old prefixes
@@ -448,7 +425,7 @@ func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain dom
 			}
 		}
 
-		d.removeDNATMappings(toRemove)
+		d.removeDNATMappings(toRemove, logger)
 	}
 
 	// Update domain prefixes using resolved domain as key - store real IPs
@@ -463,14 +440,14 @@ func (d *DnsInterceptor) updateDomainPrefixes(resolvedDomain, originalDomain dom
 		// Store real IPs for status (user-facing), not fake IPs
 		d.statusRecorder.UpdateResolvedDomainsStates(originalDomain, resolvedDomain, newPrefixes, d.route.GetResourceID())
 
-		d.logPrefixChanges(resolvedDomain, originalDomain, toAdd, toRemove)
+		d.logPrefixChanges(resolvedDomain, originalDomain, toAdd, toRemove, logger)
 	}
 
 	return nberrors.FormatErrorOrNil(merr)
 }
 
 // removeDNATMappings removes DNAT mappings from the firewall for real IP prefixes
-func (d *DnsInterceptor) removeDNATMappings(realPrefixes []netip.Prefix) {
+func (d *DnsInterceptor) removeDNATMappings(realPrefixes []netip.Prefix, logger *log.Entry) {
 	if len(realPrefixes) == 0 {
 		return
 	}
@@ -484,9 +461,9 @@ func (d *DnsInterceptor) removeDNATMappings(realPrefixes []netip.Prefix) {
 		realIP := prefix.Addr()
 		if fakeIP, exists := d.fakeIPManager.GetFakeIP(realIP); exists {
 			if err := dnatFirewall.RemoveInternalDNATMapping(fakeIP); err != nil {
-				log.Errorf("Failed to remove DNAT mapping for %s: %v", fakeIP, err)
+				logger.Errorf("failed to remove DNAT mapping for %s: %v", fakeIP, err)
 			} else {
-				log.Debugf("Removed DNAT mapping for: %s -> %s", fakeIP, realIP)
+				logger.Debugf("removed DNAT mapping: %s -> %s", fakeIP, realIP)
 			}
 		}
 	}
@@ -502,7 +479,7 @@ func (d *DnsInterceptor) internalDnatFw() (internalDNATer, bool) {
 }
 
 // addDNATMappings adds DNAT mappings to the firewall
-func (d *DnsInterceptor) addDNATMappings(mappings map[netip.Addr]netip.Addr) {
+func (d *DnsInterceptor) addDNATMappings(mappings map[netip.Addr]netip.Addr, logger *log.Entry) {
 	if len(mappings) == 0 {
 		return
 	}
@@ -514,9 +491,9 @@ func (d *DnsInterceptor) addDNATMappings(mappings map[netip.Addr]netip.Addr) {
 
 	for fakeIP, realIP := range mappings {
 		if err := dnatFirewall.AddInternalDNATMapping(fakeIP, realIP); err != nil {
-			log.Errorf("Failed to add DNAT mapping %s -> %s: %v", fakeIP, realIP, err)
+			logger.Errorf("failed to add DNAT mapping %s -> %s: %v", fakeIP, realIP, err)
 		} else {
-			log.Debugf("Added DNAT mapping: %s -> %s", fakeIP, realIP)
+			logger.Debugf("added DNAT mapping: %s -> %s", fakeIP, realIP)
 		}
 	}
 }
@@ -528,12 +505,12 @@ func (d *DnsInterceptor) cleanupDNATMappings() {
 	}
 
 	for _, prefixes := range d.interceptedDomains {
-		d.removeDNATMappings(prefixes)
+		d.removeDNATMappings(prefixes, log.NewEntry(log.StandardLogger()))
 	}
 }
 
 // replaceIPsInDNSResponse replaces real IPs with fake IPs in the DNS response
-func (d *DnsInterceptor) replaceIPsInDNSResponse(reply *dns.Msg, realPrefixes []netip.Prefix) {
+func (d *DnsInterceptor) replaceIPsInDNSResponse(reply *dns.Msg, realPrefixes []netip.Prefix, logger *log.Entry) {
 	if _, ok := d.internalDnatFw(); !ok {
 		return
 	}
@@ -549,7 +526,7 @@ func (d *DnsInterceptor) replaceIPsInDNSResponse(reply *dns.Msg, realPrefixes []
 
 			if fakeIP, exists := d.fakeIPManager.GetFakeIP(realIP); exists {
 				rr.A = fakeIP.AsSlice()
-				log.Tracef("Replaced real IP %s with fake IP %s in DNS response", realIP, fakeIP)
+				logger.Tracef("replaced real IP %s with fake IP %s in DNS response", realIP, fakeIP)
 			}
 
 		case *dns.AAAA:
@@ -560,7 +537,7 @@ func (d *DnsInterceptor) replaceIPsInDNSResponse(reply *dns.Msg, realPrefixes []
 
 			if fakeIP, exists := d.fakeIPManager.GetFakeIP(realIP); exists {
 				rr.AAAA = fakeIP.AsSlice()
-				log.Tracef("Replaced real IP %s with fake IP %s in DNS response", realIP, fakeIP)
+				logger.Tracef("replaced real IP %s with fake IP %s in DNS response", realIP, fakeIP)
 			}
 		}
 	}
@@ -586,6 +563,44 @@ func determinePrefixChanges(oldPrefixes, newPrefixes []netip.Prefix) (toAdd, toR
 	return
 }
 
+// queryUpstreamDNS queries the upstream DNS server using netstack if available, otherwise uses regular client.
+// Returns the DNS reply on success, or nil on error (error responses are written internally).
+func (d *DnsInterceptor) queryUpstreamDNS(ctx context.Context, w dns.ResponseWriter, r *dns.Msg, upstream string, upstreamIP netip.Addr, peerKey string, logger *log.Entry) *dns.Msg {
+	startTime := time.Now()
+
+	nsNet := d.wgInterface.GetNet()
+	var reply *dns.Msg
+	var err error
+
+	if nsNet != nil {
+		reply, err = nbdns.ExchangeWithNetstack(ctx, nsNet, r, upstream)
+	} else {
+		client, clientErr := nbdns.GetClientPrivate(d.wgInterface.Address().IP, d.wgInterface.Name(), dnsTimeout)
+		if clientErr != nil {
+			d.writeDNSError(w, r, logger, fmt.Sprintf("create DNS client: %v", clientErr))
+			return nil
+		}
+		reply, _, err = nbdns.ExchangeWithFallback(ctx, client, r, upstream)
+	}
+
+	if err == nil {
+		return reply
+	}
+
+	if errors.Is(err, context.DeadlineExceeded) {
+		elapsed := time.Since(startTime)
+		peerInfo := d.debugPeerTimeout(upstreamIP, peerKey)
+		logger.Errorf("peer DNS timeout after %v (timeout=%v) for domain=%s to peer %s (%s)%s - error: %v",
+			elapsed.Truncate(time.Millisecond), dnsTimeout, r.Question[0].Name, upstreamIP.String(), peerKey, peerInfo, err)
+	} else {
+		logger.Errorf("failed to exchange DNS request with %s (%s) for domain=%s: %v", upstreamIP.String(), peerKey, r.Question[0].Name, err)
+	}
+	if err := w.WriteMsg(&dns.Msg{MsgHdr: dns.MsgHdr{Rcode: dns.RcodeServerFailure, Id: r.Id}}); err != nil {
+		logger.Errorf("failed writing DNS response: %v", err)
+	}
+	return nil
+}
+
 func (d *DnsInterceptor) debugPeerTimeout(peerIP netip.Addr, peerKey string) string {
 	if d.statusRecorder == nil {
 		return ""

client/internal/routemanager/iface/iface_common.go

@@ -4,6 +4,8 @@ import (
 	"net"
 	"net/netip"
 
+	"golang.zx2c4.com/wireguard/tun/netstack"
+
 	"github.com/netbirdio/netbird/client/iface/device"
 	"github.com/netbirdio/netbird/client/iface/wgaddr"
 )
@@ -18,4 +20,5 @@ type wgIfaceBase interface {
 	IsUserspaceBind() bool
 	GetFilter() device.PacketFilter
 	GetDevice() *device.FilteredDevice
+	GetNet() *netstack.Net
 }

client/server/debug.go

@@ -173,20 +173,9 @@ func (s *Server) SetLogLevel(_ context.Context, req *proto.SetLogLevelRequest) (
 
 	log.SetLevel(level)
 
-	if s.connectClient == nil {
-		return nil, fmt.Errorf("connect client not initialized")
+	if s.connectClient != nil {
+		s.connectClient.SetLogLevel(level)
 	}
-	engine := s.connectClient.Engine()
-	if engine == nil {
-		return nil, fmt.Errorf("engine not initialized")
-	}
-
-	fwManager := engine.GetFirewallManager()
-	if fwManager == nil {
-		return nil, fmt.Errorf("firewall manager not initialized")
-	}
-
-	fwManager.SetLogLevel(level)
 
 	log.Infof("Log level set to %s", level.String())
 

client/server/event.go

@@ -1,8 +1,6 @@
 package server
 
 import (
-	"context"
-
 	log "github.com/sirupsen/logrus"
 
 	"github.com/netbirdio/netbird/client/proto"
@@ -29,8 +27,3 @@ func (s *Server) SubscribeEvents(req *proto.SubscribeRequest, stream proto.Daemo
 		}
 	}
 }
-
-func (s *Server) GetEvents(context.Context, *proto.GetEventsRequest) (*proto.GetEventsResponse, error) {
-	events := s.statusRecorder.GetEventHistory()
-	return &proto.GetEventsResponse{Events: events}, nil
-}

client/server/server.go

@@ -13,15 +13,12 @@ import (
 	"time"
 
 	"github.com/cenkalti/backoff/v4"
-	"golang.org/x/exp/maps"
 	"golang.zx2c4.com/wireguard/wgctrl/wgtypes"
-	"google.golang.org/protobuf/types/known/durationpb"
 
 	log "github.com/sirupsen/logrus"
 	"google.golang.org/grpc/codes"
 	"google.golang.org/grpc/metadata"
 	gstatus "google.golang.org/grpc/status"
-	"google.golang.org/protobuf/types/known/timestamppb"
 
 	"github.com/netbirdio/netbird/client/internal/auth"
 	"github.com/netbirdio/netbird/client/internal/profilemanager"
@@ -1067,11 +1064,9 @@ func (s *Server) Status(
 	if msg.GetFullPeerStatus {
 		s.runProbes(msg.ShouldRunProbes)
 		fullStatus := s.statusRecorder.GetFullStatus()
-		pbFullStatus := toProtoFullStatus(fullStatus)
+		pbFullStatus := fullStatus.ToProto()
 		pbFullStatus.Events = s.statusRecorder.GetEventHistory()
-
 		pbFullStatus.SshServerState = s.getSSHServerState()
-
 		statusResponse.FullStatus = pbFullStatus
 	}
 
@@ -1600,94 +1595,6 @@ func parseEnvDuration(envVar string, defaultDuration time.Duration) time.Duratio
 	return defaultDuration
 }
 
-func toProtoFullStatus(fullStatus peer.FullStatus) *proto.FullStatus {
-	pbFullStatus := proto.FullStatus{
-		ManagementState: &proto.ManagementState{},
-		SignalState:     &proto.SignalState{},
-		LocalPeerState:  &proto.LocalPeerState{},
-		Peers:           []*proto.PeerState{},
-	}
-
-	pbFullStatus.ManagementState.URL = fullStatus.ManagementState.URL
-	pbFullStatus.ManagementState.Connected = fullStatus.ManagementState.Connected
-	if err := fullStatus.ManagementState.Error; err != nil {
-		pbFullStatus.ManagementState.Error = err.Error()
-	}
-
-	pbFullStatus.SignalState.URL = fullStatus.SignalState.URL
-	pbFullStatus.SignalState.Connected = fullStatus.SignalState.Connected
-	if err := fullStatus.SignalState.Error; err != nil {
-		pbFullStatus.SignalState.Error = err.Error()
-	}
-
-	pbFullStatus.LocalPeerState.IP = fullStatus.LocalPeerState.IP
-	pbFullStatus.LocalPeerState.PubKey = fullStatus.LocalPeerState.PubKey
-	pbFullStatus.LocalPeerState.KernelInterface = fullStatus.LocalPeerState.KernelInterface
-	pbFullStatus.LocalPeerState.Fqdn = fullStatus.LocalPeerState.FQDN
-	pbFullStatus.LocalPeerState.RosenpassPermissive = fullStatus.RosenpassState.Permissive
-	pbFullStatus.LocalPeerState.RosenpassEnabled = fullStatus.RosenpassState.Enabled
-	pbFullStatus.LocalPeerState.Networks = maps.Keys(fullStatus.LocalPeerState.Routes)
-	pbFullStatus.NumberOfForwardingRules = int32(fullStatus.NumOfForwardingRules)
-	pbFullStatus.LazyConnectionEnabled = fullStatus.LazyConnectionEnabled
-
-	for _, peerState := range fullStatus.Peers {
-		pbPeerState := &proto.PeerState{
-			IP:                         peerState.IP,
-			PubKey:                     peerState.PubKey,
-			ConnStatus:                 peerState.ConnStatus.String(),
-			ConnStatusUpdate:           timestamppb.New(peerState.ConnStatusUpdate),
-			Relayed:                    peerState.Relayed,
-			LocalIceCandidateType:      peerState.LocalIceCandidateType,
-			RemoteIceCandidateType:     peerState.RemoteIceCandidateType,
-			LocalIceCandidateEndpoint:  peerState.LocalIceCandidateEndpoint,
-			RemoteIceCandidateEndpoint: peerState.RemoteIceCandidateEndpoint,
-			RelayAddress:               peerState.RelayServerAddress,
-			Fqdn:                       peerState.FQDN,
-			LastWireguardHandshake:     timestamppb.New(peerState.LastWireguardHandshake),
-			BytesRx:                    peerState.BytesRx,
-			BytesTx:                    peerState.BytesTx,
-			RosenpassEnabled:           peerState.RosenpassEnabled,
-			Networks:                   maps.Keys(peerState.GetRoutes()),
-			Latency:                    durationpb.New(peerState.Latency),
-			SshHostKey:                 peerState.SSHHostKey,
-		}
-		pbFullStatus.Peers = append(pbFullStatus.Peers, pbPeerState)
-	}
-
-	for _, relayState := range fullStatus.Relays {
-		pbRelayState := &proto.RelayState{
-			URI:       relayState.URI,
-			Available: relayState.Err == nil,
-		}
-		if err := relayState.Err; err != nil {
-			pbRelayState.Error = err.Error()
-		}
-		pbFullStatus.Relays = append(pbFullStatus.Relays, pbRelayState)
-	}
-
-	for _, dnsState := range fullStatus.NSGroupStates {
-		var err string
-		if dnsState.Error != nil {
-			err = dnsState.Error.Error()
-		}
-
-		var servers []string
-		for _, server := range dnsState.Servers {
-			servers = append(servers, server.String())
-		}
-
-		pbDnsState := &proto.NSGroupState{
-			Servers: servers,
-			Domains: dnsState.Domains,
-			Enabled: dnsState.Enabled,
-			Error:   err,
-		}
-		pbFullStatus.DnsServers = append(pbFullStatus.DnsServers, pbDnsState)
-	}
-
-	return &pbFullStatus
-}
-
 // sendTerminalNotification sends a terminal notification message
 // to inform the user that the NetBird connection session has expired.
 func sendTerminalNotification() error {

client/status/status.go

@@ -325,61 +325,64 @@ func sortPeersByIP(peersStateDetail []PeerStateDetailOutput) {
 	}
 }
 
-func ParseToJSON(overview OutputOverview) (string, error) {
-	jsonBytes, err := json.Marshal(overview)
+// JSON returns the status overview as a JSON string.
+func (o *OutputOverview) JSON() (string, error) {
+	jsonBytes, err := json.Marshal(o)
 	if err != nil {
 		return "", fmt.Errorf("json marshal failed")
 	}
 	return string(jsonBytes), err
 }
 
-func ParseToYAML(overview OutputOverview) (string, error) {
-	yamlBytes, err := yaml.Marshal(overview)
+// YAML returns the status overview as a YAML string.
+func (o *OutputOverview) YAML() (string, error) {
+	yamlBytes, err := yaml.Marshal(o)
 	if err != nil {
 		return "", fmt.Errorf("yaml marshal failed")
 	}
 	return string(yamlBytes), nil
 }
 
-func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool, showNameServers bool, showSSHSessions bool) string {
+// GeneralSummary returns a general summary of the status overview.
+func (o *OutputOverview) GeneralSummary(showURL bool, showRelays bool, showNameServers bool, showSSHSessions bool) string {
 	var managementConnString string
-	if overview.ManagementState.Connected {
+	if o.ManagementState.Connected {
 		managementConnString = "Connected"
 		if showURL {
-			managementConnString = fmt.Sprintf("%s to %s", managementConnString, overview.ManagementState.URL)
+			managementConnString = fmt.Sprintf("%s to %s", managementConnString, o.ManagementState.URL)
 		}
 	} else {
 		managementConnString = "Disconnected"
-		if overview.ManagementState.Error != "" {
-			managementConnString = fmt.Sprintf("%s, reason: %s", managementConnString, overview.ManagementState.Error)
+		if o.ManagementState.Error != "" {
+			managementConnString = fmt.Sprintf("%s, reason: %s", managementConnString, o.ManagementState.Error)
 		}
 	}
 
 	var signalConnString string
-	if overview.SignalState.Connected {
+	if o.SignalState.Connected {
 		signalConnString = "Connected"
 		if showURL {
-			signalConnString = fmt.Sprintf("%s to %s", signalConnString, overview.SignalState.URL)
+			signalConnString = fmt.Sprintf("%s to %s", signalConnString, o.SignalState.URL)
 		}
 	} else {
 		signalConnString = "Disconnected"
-		if overview.SignalState.Error != "" {
-			signalConnString = fmt.Sprintf("%s, reason: %s", signalConnString, overview.SignalState.Error)
+		if o.SignalState.Error != "" {
+			signalConnString = fmt.Sprintf("%s, reason: %s", signalConnString, o.SignalState.Error)
 		}
 	}
 
 	interfaceTypeString := "Userspace"
-	interfaceIP := overview.IP
-	if overview.KernelInterface {
+	interfaceIP := o.IP
+	if o.KernelInterface {
 		interfaceTypeString = "Kernel"
-	} else if overview.IP == "" {
+	} else if o.IP == "" {
 		interfaceTypeString = "N/A"
 		interfaceIP = "N/A"
 	}
 
 	var relaysString string
 	if showRelays {
-		for _, relay := range overview.Relays.Details {
+		for _, relay := range o.Relays.Details {
 			available := "Available"
 			reason := ""
 
@@ -395,18 +398,18 @@ func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool,
 			relaysString += fmt.Sprintf("\n  [%s] is %s%s", relay.URI, available, reason)
 		}
 	} else {
-		relaysString = fmt.Sprintf("%d/%d Available", overview.Relays.Available, overview.Relays.Total)
+		relaysString = fmt.Sprintf("%d/%d Available", o.Relays.Available, o.Relays.Total)
 	}
 
 	networks := "-"
-	if len(overview.Networks) > 0 {
-		sort.Strings(overview.Networks)
-		networks = strings.Join(overview.Networks, ", ")
+	if len(o.Networks) > 0 {
+		sort.Strings(o.Networks)
+		networks = strings.Join(o.Networks, ", ")
 	}
 
 	var dnsServersString string
 	if showNameServers {
-		for _, nsServerGroup := range overview.NSServerGroups {
+		for _, nsServerGroup := range o.NSServerGroups {
 			enabled := "Available"
 			if !nsServerGroup.Enabled {
 				enabled = "Unavailable"
@@ -430,25 +433,25 @@ func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool,
 			)
 		}
 	} else {
-		dnsServersString = fmt.Sprintf("%d/%d Available", countEnabled(overview.NSServerGroups), len(overview.NSServerGroups))
+		dnsServersString = fmt.Sprintf("%d/%d Available", countEnabled(o.NSServerGroups), len(o.NSServerGroups))
 	}
 
 	rosenpassEnabledStatus := "false"
-	if overview.RosenpassEnabled {
+	if o.RosenpassEnabled {
 		rosenpassEnabledStatus = "true"
-		if overview.RosenpassPermissive {
+		if o.RosenpassPermissive {
 			rosenpassEnabledStatus = "true (permissive)" //nolint:gosec
 		}
 	}
 
 	lazyConnectionEnabledStatus := "false"
-	if overview.LazyConnectionEnabled {
+	if o.LazyConnectionEnabled {
 		lazyConnectionEnabledStatus = "true"
 	}
 
 	sshServerStatus := "Disabled"
-	if overview.SSHServerState.Enabled {
-		sessionCount := len(overview.SSHServerState.Sessions)
+	if o.SSHServerState.Enabled {
+		sessionCount := len(o.SSHServerState.Sessions)
 		if sessionCount > 0 {
 			sessionWord := "session"
 			if sessionCount > 1 {
@@ -460,7 +463,7 @@ func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool,
 		}
 
 		if showSSHSessions && sessionCount > 0 {
-			for _, session := range overview.SSHServerState.Sessions {
+			for _, session := range o.SSHServerState.Sessions {
 				var sessionDisplay string
 				if session.JWTUsername != "" {
 					sessionDisplay = fmt.Sprintf("[%s@%s -> %s] %s",
@@ -484,7 +487,7 @@ func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool,
 		}
 	}
 
-	peersCountString := fmt.Sprintf("%d/%d Connected", overview.Peers.Connected, overview.Peers.Total)
+	peersCountString := fmt.Sprintf("%d/%d Connected", o.Peers.Connected, o.Peers.Total)
 
 	goos := runtime.GOOS
 	goarch := runtime.GOARCH
@@ -512,30 +515,31 @@ func ParseGeneralSummary(overview OutputOverview, showURL bool, showRelays bool,
 			"Forwarding rules: %d\n"+
 			"Peers count: %s\n",
 		fmt.Sprintf("%s/%s%s", goos, goarch, goarm),
-		overview.DaemonVersion,
+		o.DaemonVersion,
 		version.NetbirdVersion(),
-		overview.ProfileName,
+		o.ProfileName,
 		managementConnString,
 		signalConnString,
 		relaysString,
 		dnsServersString,
-		domain.Domain(overview.FQDN).SafeString(),
+		domain.Domain(o.FQDN).SafeString(),
 		interfaceIP,
 		interfaceTypeString,
 		rosenpassEnabledStatus,
 		lazyConnectionEnabledStatus,
 		sshServerStatus,
 		networks,
-		overview.NumberOfForwardingRules,
+		o.NumberOfForwardingRules,
 		peersCountString,
 	)
 	return summary
 }
 
-func ParseToFullDetailSummary(overview OutputOverview) string {
-	parsedPeersString := parsePeers(overview.Peers, overview.RosenpassEnabled, overview.RosenpassPermissive)
-	parsedEventsString := parseEvents(overview.Events)
-	summary := ParseGeneralSummary(overview, true, true, true, true)
+// FullDetailSummary returns a full detailed summary with peer details and events.
+func (o *OutputOverview) FullDetailSummary() string {
+	parsedPeersString := parsePeers(o.Peers, o.RosenpassEnabled, o.RosenpassPermissive)
+	parsedEventsString := parseEvents(o.Events)
+	summary := o.GeneralSummary(true, true, true, true)
 
 	return fmt.Sprintf(
 		"Peers detail:"+

client/status/status_test.go

@@ -268,7 +268,7 @@ func TestSortingOfPeers(t *testing.T) {
 }
 
 func TestParsingToJSON(t *testing.T) {
-	jsonString, _ := ParseToJSON(overview)
+	jsonString, _ := overview.JSON()
 
 	//@formatter:off
 	expectedJSONString := `
@@ -404,7 +404,7 @@ func TestParsingToJSON(t *testing.T) {
 }
 
 func TestParsingToYAML(t *testing.T) {
-	yaml, _ := ParseToYAML(overview)
+	yaml, _ := overview.YAML()
 
 	expectedYAML :=
 		`peers:
@@ -511,7 +511,7 @@ func TestParsingToDetail(t *testing.T) {
 	lastConnectionUpdate2 := timeAgo(overview.Peers.Details[1].LastStatusUpdate)
 	lastHandshake2 := timeAgo(overview.Peers.Details[1].LastWireguardHandshake)
 
-	detail := ParseToFullDetailSummary(overview)
+	detail := overview.FullDetailSummary()
 
 	expectedDetail := fmt.Sprintf(
 		`Peers detail:
@@ -575,7 +575,7 @@ Peers count: 2/2 Connected
 }
 
 func TestParsingToShortVersion(t *testing.T) {
-	shortVersion := ParseGeneralSummary(overview, false, false, false, false)
+	shortVersion := overview.GeneralSummary(false, false, false, false)
 
 	expectedString := fmt.Sprintf("OS: %s/%s", runtime.GOOS, runtime.GOARCH) + `
 Daemon version: 0.14.1

client/ui/debug.go

@@ -441,7 +441,7 @@ func (s *serviceClient) collectDebugData(
 	var postUpStatusOutput string
 	if postUpStatus != nil {
 		overview := nbstatus.ConvertToStatusOutputOverview(postUpStatus, params.anonymize, "", nil, nil, nil, "", profName)
-		postUpStatusOutput = nbstatus.ParseToFullDetailSummary(overview)
+		postUpStatusOutput = overview.FullDetailSummary()
 	}
 	headerPostUp := fmt.Sprintf("----- NetBird post-up - Timestamp: %s", time.Now().Format(time.RFC3339))
 	statusOutput := fmt.Sprintf("%s\n%s", headerPostUp, postUpStatusOutput)
@@ -458,7 +458,7 @@ func (s *serviceClient) collectDebugData(
 	var preDownStatusOutput string
 	if preDownStatus != nil {
 		overview := nbstatus.ConvertToStatusOutputOverview(preDownStatus, params.anonymize, "", nil, nil, nil, "", profName)
-		preDownStatusOutput = nbstatus.ParseToFullDetailSummary(overview)
+		preDownStatusOutput = overview.FullDetailSummary()
 	}
 	headerPreDown := fmt.Sprintf("----- NetBird pre-down - Timestamp: %s - Duration: %s",
 		time.Now().Format(time.RFC3339), params.duration)
@@ -595,7 +595,7 @@ func (s *serviceClient) createDebugBundle(anonymize bool, systemInfo bool, uploa
 	var statusOutput string
 	if statusResp != nil {
 		overview := nbstatus.ConvertToStatusOutputOverview(statusResp, anonymize, "", nil, nil, nil, "", profName)
-		statusOutput = nbstatus.ParseToFullDetailSummary(overview)
+		statusOutput = overview.FullDetailSummary()
 	}
 
 	request := &proto.DebugBundleRequest{

client/wasm/cmd/main.go

@@ -9,20 +9,31 @@ import (
 	"time"
 
 	log "github.com/sirupsen/logrus"
+	"google.golang.org/protobuf/encoding/protojson"
 
 	netbird "github.com/netbirdio/netbird/client/embed"
+	"github.com/netbirdio/netbird/client/proto"
 	sshdetection "github.com/netbirdio/netbird/client/ssh/detection"
+	nbstatus "github.com/netbirdio/netbird/client/status"
 	"github.com/netbirdio/netbird/client/wasm/internal/http"
 	"github.com/netbirdio/netbird/client/wasm/internal/rdp"
 	"github.com/netbirdio/netbird/client/wasm/internal/ssh"
 	"github.com/netbirdio/netbird/util"
+	"github.com/netbirdio/netbird/version"
 )
 
 const (
-	clientStartTimeout         = 30 * time.Second
-	clientStopTimeout          = 10 * time.Second
-	defaultLogLevel            = "warn"
-	defaultSSHDetectionTimeout = 20 * time.Second
+	clientStartTimeout           = 30 * time.Second
+	clientStopTimeout            = 10 * time.Second
+	pingTimeout                  = 10 * time.Second
+	defaultLogLevel              = "warn"
+	defaultSSHDetectionTimeout   = 20 * time.Second
+	defaultPeerConnectionTimeout = 60 * time.Second
+	peerConnectionPollInterval   = 500 * time.Millisecond
+
+	icmpEchoRequest = 8
+	icmpCodeEcho    = 0
+	pingBufferSize  = 1500
 )
 
 func main() {
@@ -113,18 +124,45 @@ func createStopMethod(client *netbird.Client) js.Func {
 	})
 }
 
+// validateSSHArgs validates SSH connection arguments
+func validateSSHArgs(args []js.Value) (host string, port int, username string, err js.Value) {
+	if len(args) < 2 {
+		return "", 0, "", js.ValueOf("error: requires host and port")
+	}
+
+	if args[0].Type() != js.TypeString {
+		return "", 0, "", js.ValueOf("host parameter must be a string")
+	}
+	if args[1].Type() != js.TypeNumber {
+		return "", 0, "", js.ValueOf("port parameter must be a number")
+	}
+
+	host = args[0].String()
+	port = args[1].Int()
+	username = "root"
+
+	if len(args) > 2 {
+		if args[2].Type() == js.TypeString && args[2].String() != "" {
+			username = args[2].String()
+		} else if args[2].Type() != js.TypeString {
+			return "", 0, "", js.ValueOf("username parameter must be a string")
+		}
+	}
+
+	return host, port, username, js.Undefined()
+}
+
 // createSSHMethod creates the SSH connection method
 func createSSHMethod(client *netbird.Client) js.Func {
 	return js.FuncOf(func(this js.Value, args []js.Value) any {
-		if len(args) < 2 {
-			return js.ValueOf("error: requires host and port")
-		}
-
-		host := args[0].String()
-		port := args[1].Int()
-		username := "root"
-		if len(args) > 2 && args[2].String() != "" {
-			username = args[2].String()
+		host, port, username, validationErr := validateSSHArgs(args)
+		if !validationErr.IsUndefined() {
+			if validationErr.Type() == js.TypeString && validationErr.String() == "error: requires host and port" {
+				return validationErr
+			}
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(validationErr)
+			})
 		}
 
 		var jwtToken string
@@ -133,6 +171,9 @@ func createSSHMethod(client *netbird.Client) js.Func {
 		}
 
 		return createPromise(func(resolve, reject js.Value) {
+			// Note: Don't wait for peer connection here - lazy connection manager
+			// will open the connection when Dial is called in ssh.Connect().
+
 			sshClient := ssh.NewClient(client)
 
 			if err := sshClient.Connect(host, port, username, jwtToken); err != nil {
@@ -154,6 +195,110 @@ func createSSHMethod(client *netbird.Client) js.Func {
 	})
 }
 
+func performPing(client *netbird.Client, hostname string) {
+	ctx, cancel := context.WithTimeout(context.Background(), pingTimeout)
+	defer cancel()
+
+	start := time.Now()
+	conn, err := client.Dial(ctx, "ping", hostname)
+	if err != nil {
+		js.Global().Get("console").Call("log", fmt.Sprintf("Ping to %s failed: %v", hostname, err))
+		return
+	}
+	defer func() {
+		if err := conn.Close(); err != nil {
+			log.Debugf("failed to close ping connection: %v", err)
+		}
+	}()
+
+	icmpData := make([]byte, 8)
+	icmpData[0] = icmpEchoRequest
+	icmpData[1] = icmpCodeEcho
+
+	if _, err := conn.Write(icmpData); err != nil {
+		js.Global().Get("console").Call("log", fmt.Sprintf("Ping to %s write failed: %v", hostname, err))
+		return
+	}
+
+	buf := make([]byte, pingBufferSize)
+	if _, err := conn.Read(buf); err != nil {
+		js.Global().Get("console").Call("log", fmt.Sprintf("Ping to %s read failed: %v", hostname, err))
+		return
+	}
+
+	latency := time.Since(start)
+	js.Global().Get("console").Call("log", fmt.Sprintf("Ping to %s: %dms", hostname, latency.Milliseconds()))
+}
+
+func performPingTCP(client *netbird.Client, hostname string, port int) {
+	ctx, cancel := context.WithTimeout(context.Background(), pingTimeout)
+	defer cancel()
+
+	address := fmt.Sprintf("%s:%d", hostname, port)
+	start := time.Now()
+	conn, err := client.Dial(ctx, "tcp", address)
+	if err != nil {
+		js.Global().Get("console").Call("log", fmt.Sprintf("TCP ping to %s failed: %v", address, err))
+		return
+	}
+	latency := time.Since(start)
+
+	if err := conn.Close(); err != nil {
+		log.Debugf("failed to close TCP connection: %v", err)
+	}
+
+	js.Global().Get("console").Call("log", fmt.Sprintf("TCP ping to %s succeeded: %dms", address, latency.Milliseconds()))
+}
+
+// createPingMethod creates the ping method
+func createPingMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(this js.Value, args []js.Value) any {
+		if len(args) < 1 {
+			return js.ValueOf("error: hostname required")
+		}
+
+		if args[0].Type() != js.TypeString {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("hostname parameter must be a string"))
+			})
+		}
+
+		hostname := args[0].String()
+		return createPromise(func(resolve, reject js.Value) {
+			performPing(client, hostname)
+			resolve.Invoke(js.Undefined())
+		})
+	})
+}
+
+// createPingTCPMethod creates the pingtcp method
+func createPingTCPMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(this js.Value, args []js.Value) any {
+		if len(args) < 2 {
+			return js.ValueOf("error: hostname and port required")
+		}
+
+		if args[0].Type() != js.TypeString {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("hostname parameter must be a string"))
+			})
+		}
+
+		if args[1].Type() != js.TypeNumber {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("port parameter must be a number"))
+			})
+		}
+
+		hostname := args[0].String()
+		port := args[1].Int()
+		return createPromise(func(resolve, reject js.Value) {
+			performPingTCP(client, hostname, port)
+			resolve.Invoke(js.Undefined())
+		})
+	})
+}
+
 // createProxyRequestMethod creates the proxyRequest method
 func createProxyRequestMethod(client *netbird.Client) js.Func {
 	return js.FuncOf(func(this js.Value, args []js.Value) any {
@@ -162,6 +307,11 @@ func createProxyRequestMethod(client *netbird.Client) js.Func {
 		}
 
 		request := args[0]
+		if request.Type() != js.TypeObject {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("request parameter must be an object"))
+			})
+		}
 
 		return createPromise(func(resolve, reject js.Value) {
 			response, err := http.ProxyRequest(client, request)
@@ -181,23 +331,255 @@ func createRDPProxyMethod(client *netbird.Client) js.Func {
 			return js.ValueOf("error: hostname and port required")
 		}
 
+		if args[0].Type() != js.TypeString {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("hostname parameter must be a string"))
+			})
+		}
+		if args[1].Type() != js.TypeString {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("port parameter must be a string"))
+			})
+		}
+
 		proxy := rdp.NewRDCleanPathProxy(client)
 		return proxy.CreateProxy(args[0].String(), args[1].String())
 	})
 }
 
+// getStatusOverview is a helper to get the status overview
+func getStatusOverview(client *netbird.Client) (nbstatus.OutputOverview, error) {
+	fullStatus, err := client.Status()
+	if err != nil {
+		return nbstatus.OutputOverview{}, err
+	}
+
+	pbFullStatus := fullStatus.ToProto()
+	statusResp := &proto.StatusResponse{
+		DaemonVersion: version.NetbirdVersion(),
+		FullStatus:    pbFullStatus,
+	}
+
+	return nbstatus.ConvertToStatusOutputOverview(statusResp, false, "", nil, nil, nil, "", ""), nil
+}
+
+// createStatusMethod creates the status method that returns JSON
+func createStatusMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		return createPromise(func(resolve, reject js.Value) {
+			overview, err := getStatusOverview(client)
+			if err != nil {
+				reject.Invoke(js.ValueOf(err.Error()))
+				return
+			}
+
+			jsonStr, err := overview.JSON()
+			if err != nil {
+				reject.Invoke(js.ValueOf(err.Error()))
+				return
+			}
+			jsonObj := js.Global().Get("JSON").Call("parse", jsonStr)
+			resolve.Invoke(jsonObj)
+		})
+	})
+}
+
+// createStatusSummaryMethod creates the statusSummary method
+func createStatusSummaryMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		return createPromise(func(resolve, reject js.Value) {
+			overview, err := getStatusOverview(client)
+			if err != nil {
+				reject.Invoke(js.ValueOf(err.Error()))
+				return
+			}
+
+			summary := overview.GeneralSummary(false, false, false, false)
+			js.Global().Get("console").Call("log", summary)
+			resolve.Invoke(js.Undefined())
+		})
+	})
+}
+
+// createStatusDetailMethod creates the statusDetail method
+func createStatusDetailMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		log.Info("statusDetail called")
+		return createPromise(func(resolve, reject js.Value) {
+			log.Info("statusDetail: getting overview")
+			overview, err := getStatusOverview(client)
+			if err != nil {
+				log.Errorf("statusDetail: getStatusOverview failed: %v", err)
+				reject.Invoke(js.ValueOf(err.Error()))
+				return
+			}
+
+			log.Info("statusDetail: generating full detail summary")
+			detail := overview.FullDetailSummary()
+			log.Infof("statusDetail: detail length=%d", len(detail))
+			js.Global().Get("console").Call("log", detail)
+			resolve.Invoke(js.Undefined())
+		})
+	})
+}
+
+// createWaitForPeerConnectionMethod creates a method that waits for a peer to be connected
+func createWaitForPeerConnectionMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		return createPromise(func(resolve, reject js.Value) {
+			if len(args) < 1 {
+				reject.Invoke(js.ValueOf("peer IP address required"))
+				return
+			}
+
+			peerIP := args[0].String()
+			timeoutMs := int(defaultPeerConnectionTimeout.Milliseconds())
+			if len(args) > 1 && !args[1].IsUndefined() && !args[1].IsNull() {
+				timeoutMs = args[1].Int()
+			}
+
+			timeout := time.Duration(timeoutMs) * time.Millisecond
+			ctx, cancel := context.WithTimeout(context.Background(), timeout)
+			defer cancel()
+
+			log.Infof("Waiting for peer %s to be connected (timeout: %v)", peerIP, timeout)
+
+			ticker := time.NewTicker(peerConnectionPollInterval)
+			defer ticker.Stop()
+
+			for {
+				select {
+				case <-ctx.Done():
+					reject.Invoke(js.ValueOf(fmt.Sprintf("timeout waiting for peer %s to connect", peerIP)))
+					return
+				case <-ticker.C:
+					overview, err := getStatusOverview(client)
+					if err != nil {
+						log.Debugf("Error getting status: %v", err)
+						continue
+					}
+
+					for _, peer := range overview.Peers.Details {
+						if peer.IP == peerIP && peer.Status == "Connected" {
+							log.Infof("Peer %s is now connected (type: %s)", peerIP, peer.ConnType)
+							resolve.Invoke(js.ValueOf(map[string]interface{}{
+								"ip":       peer.IP,
+								"status":   peer.Status,
+								"connType": peer.ConnType,
+							}))
+							return
+						}
+					}
+					log.Tracef("Peer %s not yet connected, waiting...", peerIP)
+				}
+			}
+		})
+	})
+}
+
+// createGetSyncResponseMethod creates the getSyncResponse method that returns the latest sync response as JSON
+func createGetSyncResponseMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		return createPromise(func(resolve, reject js.Value) {
+			syncResp, err := client.GetLatestSyncResponse()
+			if err != nil {
+				reject.Invoke(js.ValueOf(err.Error()))
+				return
+			}
+
+			options := protojson.MarshalOptions{
+				EmitUnpopulated: true,
+				UseProtoNames:   true,
+				AllowPartial:    true,
+			}
+			jsonBytes, err := options.Marshal(syncResp)
+			if err != nil {
+				reject.Invoke(js.ValueOf(fmt.Sprintf("marshal sync response: %v", err)))
+				return
+			}
+
+			jsonObj := js.Global().Get("JSON").Call("parse", string(jsonBytes))
+			resolve.Invoke(jsonObj)
+		})
+	})
+}
+
+// createSetLogLevelMethod creates the setLogLevel method to dynamically change logging level
+func createSetLogLevelMethod(client *netbird.Client) js.Func {
+	return js.FuncOf(func(_ js.Value, args []js.Value) any {
+		if len(args) < 1 {
+			return js.ValueOf("error: log level required")
+		}
+
+		if args[0].Type() != js.TypeString {
+			return createPromise(func(resolve, reject js.Value) {
+				reject.Invoke(js.ValueOf("log level parameter must be a string"))
+			})
+		}
+
+		logLevel := args[0].String()
+		return createPromise(func(resolve, reject js.Value) {
+			if err := client.SetLogLevel(logLevel); err != nil {
+				reject.Invoke(js.ValueOf(fmt.Sprintf("set log level: %v", err)))
+				return
+			}
+			log.Infof("Log level set to: %s", logLevel)
+			resolve.Invoke(js.ValueOf(true))
+		})
+	})
+}
+
 // createPromise is a helper to create JavaScript promises
 func createPromise(handler func(resolve, reject js.Value)) js.Value {
 	return js.Global().Get("Promise").New(js.FuncOf(func(_ js.Value, promiseArgs []js.Value) any {
 		resolve := promiseArgs[0]
 		reject := promiseArgs[1]
 
-		go handler(resolve, reject)
+		// Wrap reject to always log the error
+		loggingReject := js.FuncOf(func(_ js.Value, args []js.Value) any {
+			if len(args) > 0 {
+				log.Errorf("Promise rejected: %v", args[0])
+				js.Global().Get("console").Call("error", "WASM Promise rejected:", args[0])
+			}
+			reject.Invoke(args[0])
+			return nil
+		})
+
+		go handler(resolve, loggingReject.Value)
 
 		return nil
 	}))
 }
 
+// waitForPeerConnected waits for a peer with the given IP to be connected
+func waitForPeerConnected(ctx context.Context, client *netbird.Client, peerIP string) error {
+	log.Infof("Waiting for peer %s to be connected before operation", peerIP)
+
+	ticker := time.NewTicker(peerConnectionPollInterval)
+	defer ticker.Stop()
+
+	for {
+		select {
+		case <-ctx.Done():
+			return fmt.Errorf("timeout waiting for peer %s to connect", peerIP)
+		case <-ticker.C:
+			overview, err := getStatusOverview(client)
+			if err != nil {
+				log.Debugf("Error getting status while waiting for peer: %v", err)
+				continue
+			}
+
+			for _, peer := range overview.Peers.Details {
+				if peer.IP == peerIP && peer.Status == "Connected" {
+					log.Infof("Peer %s is now connected (type: %s), proceeding with operation", peerIP, peer.ConnType)
+					return nil
+				}
+			}
+			log.Tracef("Peer %s not yet connected, waiting...", peerIP)
+		}
+	}
+}
+
 // createDetectSSHServerMethod creates the SSH server detection method
 func createDetectSSHServerMethod(client *netbird.Client) js.Func {
 	return js.FuncOf(func(this js.Value, args []js.Value) any {
@@ -220,6 +602,10 @@ func createDetectSSHServerMethod(client *netbird.Client) js.Func {
 			ctx, cancel := context.WithTimeout(context.Background(), time.Duration(timeoutMs)*time.Millisecond)
 			defer cancel()
 
+			// Note: Don't wait for peer connection here - lazy connection manager
+			// will open the connection when Dial is called. Waiting would cause
+			// a deadlock since lazy connections only open on traffic.
+
 			serverType, err := sshdetection.DetectSSHServerType(ctx, client, host, port)
 			if err != nil {
 				reject.Invoke(err.Error())
@@ -237,17 +623,25 @@ func createClientObject(client *netbird.Client) js.Value {
 
 	obj["start"] = createStartMethod(client)
 	obj["stop"] = createStopMethod(client)
+	obj["ping"] = createPingMethod(client)
+	obj["pingtcp"] = createPingTCPMethod(client)
 	obj["detectSSHServerType"] = createDetectSSHServerMethod(client)
 	obj["createSSHConnection"] = createSSHMethod(client)
 	obj["proxyRequest"] = createProxyRequestMethod(client)
 	obj["createRDPProxy"] = createRDPProxyMethod(client)
+	obj["status"] = createStatusMethod(client)
+	obj["statusSummary"] = createStatusSummaryMethod(client)
+	obj["statusDetail"] = createStatusDetailMethod(client)
+	obj["waitForPeerConnection"] = createWaitForPeerConnectionMethod(client)
+	obj["getSyncResponse"] = createGetSyncResponseMethod(client)
+	obj["setLogLevel"] = createSetLogLevelMethod(client)
 
 	return js.ValueOf(obj)
 }
 
 // netBirdClientConstructor acts as a JavaScript constructor function
-func netBirdClientConstructor(this js.Value, args []js.Value) any {
-	return js.Global().Get("Promise").New(js.FuncOf(func(this js.Value, promiseArgs []js.Value) any {
+func netBirdClientConstructor(_ js.Value, args []js.Value) any {
+	return js.Global().Get("Promise").New(js.FuncOf(func(_ js.Value, promiseArgs []js.Value) any {
 		resolve := promiseArgs[0]
 		reject := promiseArgs[1]
 
@@ -263,7 +657,10 @@ func netBirdClientConstructor(this js.Value, args []js.Value) any {
 				return
 			}
 
-			if err := util.InitLog(options.LogLevel, util.LogConsole); err != nil {
+			// Force trace logging for debugging - must be set BEFORE netbird.New
+			// as New() will call logrus.SetLevel with options.LogLevel
+			options.LogLevel = "trace"
+			if err := util.InitLog("trace", util.LogConsole); err != nil {
 				log.Warnf("Failed to initialize logging: %v", err)
 			}
 

dns/dns.go

@@ -47,8 +47,8 @@ type CustomZone struct {
 	Records []SimpleRecord
 	// SearchDomainDisabled indicates whether to add match domains to a search domains list or not
 	SearchDomainDisabled bool
-	// SkipPTRProcess indicates whether a client should process PTR records from custom zones
-	SkipPTRProcess bool
+	// NonAuthoritative marks user-created zones
+	NonAuthoritative bool
 }
 
 // SimpleRecord provides a simple DNS record specification for CNAME, A and AAAA records

go.mod

@@ -78,8 +78,8 @@ require (
 	github.com/pion/logging v0.2.4
 	github.com/pion/randutil v0.1.0
 	github.com/pion/stun/v2 v2.0.0
-	github.com/pion/stun/v3 v3.0.0
-	github.com/pion/transport/v3 v3.0.7
+	github.com/pion/stun/v3 v3.1.0
+	github.com/pion/transport/v3 v3.1.1
 	github.com/pion/turn/v3 v3.0.1
 	github.com/pkg/sftp v1.13.9
 	github.com/prometheus/client_golang v1.23.2
@@ -241,7 +241,7 @@ require (
 	github.com/opencontainers/go-digest v1.0.0 // indirect
 	github.com/opencontainers/image-spec v1.1.0 // indirect
 	github.com/pion/dtls/v2 v2.2.10 // indirect
-	github.com/pion/dtls/v3 v3.0.7 // indirect
+	github.com/pion/dtls/v3 v3.0.9 // indirect
 	github.com/pion/mdns/v2 v2.0.7 // indirect
 	github.com/pion/transport/v2 v2.2.4 // indirect
 	github.com/pion/turn/v4 v4.1.1 // indirect
@@ -263,7 +263,7 @@ require (
 	github.com/tklauser/numcpus v0.8.0 // indirect
 	github.com/vishvananda/netns v0.0.5 // indirect
 	github.com/vmihailenco/tagparser/v2 v2.0.0 // indirect
-	github.com/wlynxg/anet v0.0.3 // indirect
+	github.com/wlynxg/anet v0.0.5 // indirect
 	github.com/yuin/goldmark v1.7.8 // indirect
 	github.com/zeebo/blake3 v0.2.3 // indirect
 	go.opentelemetry.io/auto/sdk v1.2.1 // indirect

go.sum

@@ -444,8 +444,8 @@ github.com/petermattis/goid v0.0.0-20250303134427-723919f7f203/go.mod h1:pxMtw7c
 github.com/pion/dtls/v2 v2.2.7/go.mod h1:8WiMkebSHFD0T+dIU+UeBaoV7kDhOW5oDCzZ7WZ/F9s=
 github.com/pion/dtls/v2 v2.2.10 h1:u2Axk+FyIR1VFTPurktB+1zoEPGIW3bmyj3LEFrXjAA=
 github.com/pion/dtls/v2 v2.2.10/go.mod h1:d9SYc9fch0CqK90mRk1dC7AkzzpwJj6u2GU3u+9pqFE=
-github.com/pion/dtls/v3 v3.0.7 h1:bItXtTYYhZwkPFk4t1n3Kkf5TDrfj6+4wG+CZR8uI9Q=
-github.com/pion/dtls/v3 v3.0.7/go.mod h1:uDlH5VPrgOQIw59irKYkMudSFprY9IEFCqz/eTz16f8=
+github.com/pion/dtls/v3 v3.0.9 h1:4AijfFRm8mAjd1gfdlB1wzJF3fjjR/VPIpJgkEtvYmM=
+github.com/pion/dtls/v3 v3.0.9/go.mod h1:abApPjgadS/ra1wvUzHLc3o2HvoxppAh+NZkyApL4Os=
 github.com/pion/logging v0.2.2/go.mod h1:k0/tDVsRCX2Mb2ZEmTqNa7CWsQPc+YYCB7Q+5pahoms=
 github.com/pion/logging v0.2.4 h1:tTew+7cmQ+Mc1pTBLKH2puKsOvhm32dROumOZ655zB8=
 github.com/pion/logging v0.2.4/go.mod h1:DffhXTKYdNZU+KtJ5pyQDjvOAh/GsNSyv1lbkFbe3so=
@@ -455,14 +455,14 @@ github.com/pion/randutil v0.1.0 h1:CFG1UdESneORglEsnimhUjf33Rwjubwj6xfiOXBa3mA=
 github.com/pion/randutil v0.1.0/go.mod h1:XcJrSMMbbMRhASFVOlj/5hQial/Y8oH/HVo7TBZq+j8=
 github.com/pion/stun/v2 v2.0.0 h1:A5+wXKLAypxQri59+tmQKVs7+l6mMM+3d+eER9ifRU0=
 github.com/pion/stun/v2 v2.0.0/go.mod h1:22qRSh08fSEttYUmJZGlriq9+03jtVmXNODgLccj8GQ=
-github.com/pion/stun/v3 v3.0.0 h1:4h1gwhWLWuZWOJIJR9s2ferRO+W3zA/b6ijOI6mKzUw=
-github.com/pion/stun/v3 v3.0.0/go.mod h1:HvCN8txt8mwi4FBvS3EmDghW6aQJ24T+y+1TKjB5jyU=
+github.com/pion/stun/v3 v3.1.0 h1:bS1jjT3tGWZ4UPmIUeyalOylamTMTFg1OvXtY/r6seM=
+github.com/pion/stun/v3 v3.1.0/go.mod h1:egmx1CUcfSSGJxQCOjtVlomfPqmQ58BibPyuOWNGQEU=
 github.com/pion/transport/v2 v2.2.1/go.mod h1:cXXWavvCnFF6McHTft3DWS9iic2Mftcz1Aq29pGcU5g=
 github.com/pion/transport/v2 v2.2.4 h1:41JJK6DZQYSeVLxILA2+F4ZkKb4Xd/tFJZRFZQ9QAlo=
 github.com/pion/transport/v2 v2.2.4/go.mod h1:q2U/tf9FEfnSBGSW6w5Qp5PFWRLRj3NjLhCCgpRK4p0=
 github.com/pion/transport/v3 v3.0.1/go.mod h1:UY7kiITrlMv7/IKgd5eTUcaahZx5oUN3l9SzK5f5xE0=
-github.com/pion/transport/v3 v3.0.7 h1:iRbMH05BzSNwhILHoBoAPxoB9xQgOaJk+591KC9P1o0=
-github.com/pion/transport/v3 v3.0.7/go.mod h1:YleKiTZ4vqNxVwh77Z0zytYi7rXHl7j6uPLGhhz9rwo=
+github.com/pion/transport/v3 v3.1.1 h1:Tr684+fnnKlhPceU+ICdrw6KKkTms+5qHMgw6bIkYOM=
+github.com/pion/transport/v3 v3.1.1/go.mod h1:+c2eewC5WJQHiAA46fkMMzoYZSuGzA/7E2FPrOYHctQ=
 github.com/pion/turn/v3 v3.0.1 h1:wLi7BTQr6/Q20R0vt/lHbjv6y4GChFtC33nkYbasoT8=
 github.com/pion/turn/v3 v3.0.1/go.mod h1:MrJDKgqryDyWy1/4NT9TWfXWGMC7UHT6pJIv1+gMeNE=
 github.com/pion/turn/v4 v4.1.1 h1:9UnY2HB99tpDyz3cVVZguSxcqkJ1DsTSZ+8TGruh4fc=
@@ -574,8 +574,8 @@ github.com/vmihailenco/msgpack/v5 v5.4.1 h1:cQriyiUvjTwOHg8QZaPihLWeRAAVoCpE00IU
 github.com/vmihailenco/msgpack/v5 v5.4.1/go.mod h1:GaZTsDaehaPpQVyxrf5mtQlH+pc21PIudVV/E3rRQok=
 github.com/vmihailenco/tagparser/v2 v2.0.0 h1:y09buUbR+b5aycVFQs/g70pqKVZNBmxwAhO7/IwNM9g=
 github.com/vmihailenco/tagparser/v2 v2.0.0/go.mod h1:Wri+At7QHww0WTrCBeu4J6bNtoV6mEfg5OIWRZA9qds=
-github.com/wlynxg/anet v0.0.3 h1:PvR53psxFXstc12jelG6f1Lv4MWqE0tI76/hHGjh9rg=
-github.com/wlynxg/anet v0.0.3/go.mod h1:eay5PRQr7fIVAMbTbchTnO9gG65Hg/uYGdc7mguHxoA=
+github.com/wlynxg/anet v0.0.5 h1:J3VJGi1gvo0JwZ/P1/Yc/8p63SoW98B5dHkYDmpgvvU=
+github.com/wlynxg/anet v0.0.5/go.mod h1:eay5PRQr7fIVAMbTbchTnO9gG65Hg/uYGdc7mguHxoA=
 github.com/yuin/goldmark v1.1.27/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 github.com/yuin/goldmark v1.2.1/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
 github.com/yuin/goldmark v1.3.5/go.mod h1:mwnBkeHKe2W/ZEtQ+71ViKU8L12m81fl3OWwC1Zlc8k=

idp/dex/provider.go

@@ -792,11 +792,12 @@ func (p *Provider) resolveRedirectURI(redirectURI string) string {
 // buildOIDCConnectorConfig creates config for OIDC-based connectors
 func buildOIDCConnectorConfig(cfg *ConnectorConfig, redirectURI string) ([]byte, error) {
 	oidcConfig := map[string]interface{}{
-		"issuer":       cfg.Issuer,
-		"clientID":     cfg.ClientID,
-		"clientSecret": cfg.ClientSecret,
-		"redirectURI":  redirectURI,
-		"scopes":       []string{"openid", "profile", "email"},
+		"issuer":               cfg.Issuer,
+		"clientID":             cfg.ClientID,
+		"clientSecret":         cfg.ClientSecret,
+		"redirectURI":          redirectURI,
+		"scopes":               []string{"openid", "profile", "email"},
+		"insecureEnableGroups": true,
 	}
 	switch cfg.Type {
 	case "zitadel":
@@ -806,6 +807,9 @@ func buildOIDCConnectorConfig(cfg *ConnectorConfig, redirectURI string) ([]byte,
 		oidcConfig["claimMapping"] = map[string]string{"email": "preferred_username"}
 	case "okta":
 		oidcConfig["insecureSkipEmailVerified"] = true
+		oidcConfig["scopes"] = []string{"openid", "profile", "email", "groups"}
+	case "pocketid":
+		oidcConfig["scopes"] = []string{"openid", "profile", "email", "groups"}
 	}
 	return encodeConnectorConfig(oidcConfig)
 }

infrastructure_files/getting-started.sh

@@ -270,7 +270,7 @@ AUTH_CLIENT_ID=netbird-dashboard
 AUTH_CLIENT_SECRET=
 AUTH_AUTHORITY=$NETBIRD_HTTP_PROTOCOL://$NETBIRD_DOMAIN/oauth2
 USE_AUTH0=false
-AUTH_SUPPORTED_SCOPES=openid profile email offline_access
+AUTH_SUPPORTED_SCOPES=openid profile email groups
 AUTH_REDIRECT_URI=/nb-auth
 AUTH_SILENT_REDIRECT_URI=/nb-silent-auth
 # SSL

management/cmd/management.go

@@ -143,7 +143,7 @@ func loadMgmtConfig(ctx context.Context, mgmtConfigPath string) (*nbconfig.Confi
 	applyCommandLineOverrides(loadedConfig)
 
 	// Apply EmbeddedIdP config to HttpConfig if embedded IdP is enabled
-	err := applyEmbeddedIdPConfig(loadedConfig)
+	err := applyEmbeddedIdPConfig(ctx, loadedConfig)
 	if err != nil {
 		return nil, err
 	}
@@ -177,7 +177,7 @@ func applyCommandLineOverrides(cfg *nbconfig.Config) {
 
 // applyEmbeddedIdPConfig populates HttpConfig and EmbeddedIdP storage from config when embedded IdP is enabled.
 // This allows users to only specify EmbeddedIdP config without duplicating values in HttpConfig.
-func applyEmbeddedIdPConfig(cfg *nbconfig.Config) error {
+func applyEmbeddedIdPConfig(ctx context.Context, cfg *nbconfig.Config) error {
 	if cfg.EmbeddedIdP == nil || !cfg.EmbeddedIdP.Enabled {
 		return nil
 	}
@@ -193,11 +193,6 @@ func applyEmbeddedIdPConfig(cfg *nbconfig.Config) error {
 	// Set LocalAddress for embedded IdP if enabled, used for internal JWT validation
 	cfg.EmbeddedIdP.LocalAddress = fmt.Sprintf("localhost:%d", mgmtPort)
 
-	// Ensure HttpConfig exists
-	if cfg.HttpConfig == nil {
-		cfg.HttpConfig = &nbconfig.HttpServerConfig{}
-	}
-
 	// Set storage defaults based on Datadir
 	if cfg.EmbeddedIdP.Storage.Type == "" {
 		cfg.EmbeddedIdP.Storage.Type = "sqlite3"
@@ -208,40 +203,22 @@ func applyEmbeddedIdPConfig(cfg *nbconfig.Config) error {
 
 	issuer := cfg.EmbeddedIdP.Issuer
 
-	// Set AuthIssuer from EmbeddedIdP issuer
-	if cfg.HttpConfig.AuthIssuer == "" {
-		cfg.HttpConfig.AuthIssuer = issuer
+	if cfg.HttpConfig != nil {
+		log.WithContext(ctx).Warnf("overriding HttpConfig with EmbeddedIdP config. " +
+			"HttpConfig is ignored when EmbeddedIdP is enabled. Please remove HttpConfig section from the config file")
+	} else {
+		// Ensure HttpConfig exists. We need it for backwards compatibility with the old config format.
+		cfg.HttpConfig = &nbconfig.HttpServerConfig{}
 	}
 
-	// Set AuthAudience to the dashboard client ID
-	if cfg.HttpConfig.AuthAudience == "" {
-		cfg.HttpConfig.AuthAudience = "netbird-dashboard"
-	}
-
-	// Set CLIAuthAudience to the client app client ID
-	if cfg.HttpConfig.CLIAuthAudience == "" {
-		cfg.HttpConfig.CLIAuthAudience = "netbird-cli"
-	}
-
-	// Set AuthUserIDClaim to "sub" (standard OIDC claim)
-	if cfg.HttpConfig.AuthUserIDClaim == "" {
-		cfg.HttpConfig.AuthUserIDClaim = "sub"
-	}
-
-	// Set AuthKeysLocation to the JWKS endpoint
-	if cfg.HttpConfig.AuthKeysLocation == "" {
-		cfg.HttpConfig.AuthKeysLocation = issuer + "/keys"
-	}
-
-	// Set OIDCConfigEndpoint to the discovery endpoint
-	if cfg.HttpConfig.OIDCConfigEndpoint == "" {
-		cfg.HttpConfig.OIDCConfigEndpoint = issuer + "/.well-known/openid-configuration"
-	}
-
-	// Copy SignKeyRefreshEnabled from EmbeddedIdP config
-	if cfg.EmbeddedIdP.SignKeyRefreshEnabled {
-		cfg.HttpConfig.IdpSignKeyRefreshEnabled = true
-	}
+	// Set HttpConfig values from EmbeddedIdP
+	cfg.HttpConfig.AuthIssuer = issuer
+	cfg.HttpConfig.AuthAudience = "netbird-dashboard"
+	cfg.HttpConfig.CLIAuthAudience = "netbird-cli"
+	cfg.HttpConfig.AuthUserIDClaim = "sub"
+	cfg.HttpConfig.AuthKeysLocation = issuer + "/keys"
+	cfg.HttpConfig.OIDCConfigEndpoint = issuer + "/.well-known/openid-configuration"
+	cfg.HttpConfig.IdpSignKeyRefreshEnabled = true
 
 	return nil
 }
@@ -249,7 +226,12 @@ func applyEmbeddedIdPConfig(cfg *nbconfig.Config) error {
 // applyOIDCConfig fetches and applies OIDC configuration if endpoint is specified
 func applyOIDCConfig(ctx context.Context, cfg *nbconfig.Config) error {
 	oidcEndpoint := cfg.HttpConfig.OIDCConfigEndpoint
-	if oidcEndpoint == "" || cfg.EmbeddedIdP != nil {
+	if oidcEndpoint == "" {
+		return nil
+	}
+
+	if cfg.EmbeddedIdP != nil && cfg.EmbeddedIdP.Enabled {
+		// skip OIDC config fetching if EmbeddedIdP is enabled as it is unnecessary given it is embedded
 		return nil
 	}
 

management/internals/shared/grpc/conversion.go

@@ -374,8 +374,9 @@ func shouldUsePortRange(rule *proto.FirewallRule) bool {
 // Helper function to convert nbdns.CustomZone to proto.CustomZone
 func convertToProtoCustomZone(zone nbdns.CustomZone) *proto.CustomZone {
 	protoZone := &proto.CustomZone{
-		Domain:  zone.Domain,
-		Records: make([]*proto.SimpleRecord, 0, len(zone.Records)),
+		Domain:           zone.Domain,
+		Records:          make([]*proto.SimpleRecord, 0, len(zone.Records)),
+		NonAuthoritative: zone.NonAuthoritative,
 	}
 	for _, record := range zone.Records {
 		protoZone.Records = append(protoZone.Records, &proto.SimpleRecord{

management/server/cache/idp.go

@@ -26,6 +26,8 @@ type UserDataCache interface {
 	Get(ctx context.Context, key string) (*idp.UserData, error)
 	Set(ctx context.Context, key string, value *idp.UserData, expiration time.Duration) error
 	Delete(ctx context.Context, key string) error
+	GetUsers(ctx context.Context, key string) ([]*idp.UserData, error)
+	SetUsers(ctx context.Context, key string, users []*idp.UserData, expiration time.Duration) error
 }
 
 // UserDataCacheImpl is a struct that implements the UserDataCache interface.
@@ -51,6 +53,29 @@ func (u *UserDataCacheImpl) Delete(ctx context.Context, key string) error {
 	return u.cache.Delete(ctx, key)
 }
 
+func (u *UserDataCacheImpl) GetUsers(ctx context.Context, key string) ([]*idp.UserData, error) {
+	var users []*idp.UserData
+	v, err := u.cache.Get(ctx, key, &users)
+	if err != nil {
+		return nil, err
+	}
+
+	switch v := v.(type) {
+	case []*idp.UserData:
+		return v, nil
+	case *[]*idp.UserData:
+		return *v, nil
+	case []byte:
+		return unmarshalUserData(v)
+	}
+
+	return nil, fmt.Errorf("unexpected type: %T", v)
+}
+
+func (u *UserDataCacheImpl) SetUsers(ctx context.Context, key string, users []*idp.UserData, expiration time.Duration) error {
+	return u.cache.Set(ctx, key, users, store.WithExpiration(expiration))
+}
+
 // NewUserDataCache creates a new UserDataCacheImpl object.
 func NewUserDataCache(store store.StoreInterface) *UserDataCacheImpl {
 	simpleCache := cache.New[any](store)

management/server/http/middleware/auth_middleware.go

@@ -178,7 +178,7 @@ func (m *AuthMiddleware) checkPATFromRequest(r *http.Request, authHeaderParts []
 		m.patUsageTracker.IncrementUsage(token)
 	}
 
-	if m.rateLimiter != nil {
+	if m.rateLimiter != nil && !isTerraformRequest(r) {
 		if !m.rateLimiter.Allow(token) {
 			return r, status.Errorf(status.TooManyRequests, "too many requests")
 		}
@@ -214,6 +214,11 @@ func (m *AuthMiddleware) checkPATFromRequest(r *http.Request, authHeaderParts []
 	return nbcontext.SetUserAuthInRequest(r, userAuth), nil
 }
 
+func isTerraformRequest(r *http.Request) bool {
+	ua := strings.ToLower(r.Header.Get("User-Agent"))
+	return strings.Contains(ua, "terraform")
+}
+
 // getTokenFromJWTRequest is a "TokenExtractor" that takes auth header parts and extracts
 // the JWT token from the Authorization header.
 func getTokenFromJWTRequest(authHeaderParts []string) (string, error) {

management/server/http/middleware/auth_middleware_test.go

@@ -508,6 +508,103 @@ func TestAuthMiddleware_RateLimiting(t *testing.T) {
 		handler.ServeHTTP(rec, req)
 		assert.Equal(t, http.StatusTooManyRequests, rec.Code, "Second request after cleanup should be rate limited again")
 	})
+
+	t.Run("Terraform User Agent Not Rate Limited", func(t *testing.T) {
+		rateLimitConfig := &RateLimiterConfig{
+			RequestsPerMinute: 1,
+			Burst:             1,
+			CleanupInterval:   5 * time.Minute,
+			LimiterTTL:        10 * time.Minute,
+		}
+
+		authMiddleware := NewAuthMiddleware(
+			mockAuth,
+			func(ctx context.Context, userAuth nbauth.UserAuth) (string, string, error) {
+				return userAuth.AccountId, userAuth.UserId, nil
+			},
+			func(ctx context.Context, userAuth nbauth.UserAuth) error {
+				return nil
+			},
+			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
+				return &types.User{}, nil
+			},
+			rateLimitConfig,
+			nil,
+		)
+
+		handler := authMiddleware.Handler(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+			w.WriteHeader(http.StatusOK)
+		}))
+
+		// Test various Terraform user agent formats
+		terraformUserAgents := []string{
+			"Terraform/1.5.0",
+			"terraform/1.0.0",
+			"Terraform-Provider/2.0.0",
+			"Mozilla/5.0 (compatible; Terraform/1.3.0)",
+		}
+
+		for _, userAgent := range terraformUserAgents {
+			t.Run("UserAgent: "+userAgent, func(t *testing.T) {
+				successCount := 0
+				for i := 0; i < 10; i++ {
+					req := httptest.NewRequest("GET", "http://testing/test", nil)
+					req.Header.Set("Authorization", "Token "+PAT)
+					req.Header.Set("User-Agent", userAgent)
+					rec := httptest.NewRecorder()
+
+					handler.ServeHTTP(rec, req)
+					if rec.Code == http.StatusOK {
+						successCount++
+					}
+				}
+
+				assert.Equal(t, 10, successCount, "All Terraform user agent requests should succeed (not rate limited)")
+			})
+		}
+	})
+
+	t.Run("Non-Terraform User Agent With PAT Is Rate Limited", func(t *testing.T) {
+		rateLimitConfig := &RateLimiterConfig{
+			RequestsPerMinute: 1,
+			Burst:             1,
+			CleanupInterval:   5 * time.Minute,
+			LimiterTTL:        10 * time.Minute,
+		}
+
+		authMiddleware := NewAuthMiddleware(
+			mockAuth,
+			func(ctx context.Context, userAuth nbauth.UserAuth) (string, string, error) {
+				return userAuth.AccountId, userAuth.UserId, nil
+			},
+			func(ctx context.Context, userAuth nbauth.UserAuth) error {
+				return nil
+			},
+			func(ctx context.Context, userAuth nbauth.UserAuth) (*types.User, error) {
+				return &types.User{}, nil
+			},
+			rateLimitConfig,
+			nil,
+		)
+
+		handler := authMiddleware.Handler(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+			w.WriteHeader(http.StatusOK)
+		}))
+
+		req := httptest.NewRequest("GET", "http://testing/test", nil)
+		req.Header.Set("Authorization", "Token "+PAT)
+		req.Header.Set("User-Agent", "curl/7.68.0")
+		rec := httptest.NewRecorder()
+		handler.ServeHTTP(rec, req)
+		assert.Equal(t, http.StatusOK, rec.Code, "First request should succeed")
+
+		req = httptest.NewRequest("GET", "http://testing/test", nil)
+		req.Header.Set("Authorization", "Token "+PAT)
+		req.Header.Set("User-Agent", "curl/7.68.0")
+		rec = httptest.NewRecorder()
+		handler.ServeHTTP(rec, req)
+		assert.Equal(t, http.StatusTooManyRequests, rec.Code, "Second request should be rate limited")
+	})
 }
 
 func TestAuthMiddleware_Handler_Child(t *testing.T) {

management/server/idp/embedded.go

@@ -20,7 +20,7 @@ const (
 	staticClientCLI        = "netbird-cli"
 	defaultCLIRedirectURL1 = "http://localhost:53000/"
 	defaultCLIRedirectURL2 = "http://localhost:54000/"
-	defaultScopes          = "openid profile email offline_access"
+	defaultScopes          = "openid profile email"
 	defaultUserIDClaim     = "sub"
 )
 

management/server/user.go

@@ -911,10 +911,12 @@ func (am *DefaultAccountManager) GetUsersFromAccount(ctx context.Context, accoun
 	accountUsers := []*types.User{}
 	switch {
 	case allowed:
+		start := time.Now()
 		accountUsers, err = am.Store.GetAccountUsers(ctx, store.LockingStrengthNone, accountID)
 		if err != nil {
 			return nil, err
 		}
+		log.WithContext(ctx).Tracef("Got %d users from account %s after %s", len(accountUsers), accountID, time.Since(start))
 	case user != nil && user.AccountID == accountID:
 		accountUsers = append(accountUsers, user)
 	default:
@@ -933,23 +935,40 @@ func (am *DefaultAccountManager) BuildUserInfosForAccount(ctx context.Context, a
 	if !isNil(am.idpManager) && !IsEmbeddedIdp(am.idpManager) {
 		users := make(map[string]userLoggedInOnce, len(accountUsers))
 		usersFromIntegration := make([]*idp.UserData, 0)
+		filtered := make(map[string]*idp.UserData, len(accountUsers))
+		log.WithContext(ctx).Tracef("Querying users from IDP for account %s", accountID)
+		start := time.Now()
+
+		integrationKeys := make(map[string]struct{})
 		for _, user := range accountUsers {
 			if user.Issued == types.UserIssuedIntegration {
-				key := user.IntegrationReference.CacheKey(accountID, user.Id)
-				info, err := am.externalCacheManager.Get(am.ctx, key)
-				if err != nil {
-					log.WithContext(ctx).Infof("Get ExternalCache for key: %s, error: %s", key, err)
-					users[user.Id] = true
-					continue
-				}
-				usersFromIntegration = append(usersFromIntegration, info)
+				integrationKeys[user.IntegrationReference.CacheKey(accountID)] = struct{}{}
 				continue
 			}
 			if !user.IsServiceUser {
 				users[user.Id] = userLoggedInOnce(!user.GetLastLogin().IsZero())
 			}
 		}
+
+		for key := range integrationKeys {
+			usersData, err := am.externalCacheManager.GetUsers(am.ctx, key)
+			if err != nil {
+				log.WithContext(ctx).Debugf("GetUsers from ExternalCache for key: %s, error: %s", key, err)
+				continue
+			}
+			for _, ud := range usersData {
+				filtered[ud.ID] = ud
+			}
+		}
+
+		for _, ud := range filtered {
+			usersFromIntegration = append(usersFromIntegration, ud)
+		}
+
+		log.WithContext(ctx).Tracef("Got user info from external cache after %s", time.Since(start))
+		start = time.Now()
 		queriedUsers, err = am.lookupCache(ctx, users, accountID)
+		log.WithContext(ctx).Tracef("Got user info from cache for %d users after %s", len(queriedUsers), time.Since(start))
 		if err != nil {
 			return nil, err
 		}

management/server/user_test.go

@@ -1086,8 +1086,12 @@ func TestDefaultAccountManager_ExternalCache(t *testing.T) {
 	assert.NoError(t, err)
 
 	cacheManager := am.GetExternalCacheManager()
-	cacheKey := externalUser.IntegrationReference.CacheKey(mockAccountID, externalUser.Id)
-	err = cacheManager.Set(context.Background(), cacheKey, &idp.UserData{ID: externalUser.Id, Name: "Test User", Email: "user@example.com"}, time.Minute)
+	tud := &idp.UserData{ID: externalUser.Id, Name: "Test User", Email: "user@example.com"}
+	cacheKeyUser := externalUser.IntegrationReference.CacheKey(mockAccountID, externalUser.Id)
+	err = cacheManager.Set(context.Background(), cacheKeyUser, tud, time.Minute)
+	assert.NoError(t, err)
+	cacheKeyAccount := externalUser.IntegrationReference.CacheKey(mockAccountID)
+	err = cacheManager.SetUsers(context.Background(), cacheKeyAccount, []*idp.UserData{tud}, time.Minute)
 	assert.NoError(t, err)
 
 	infos, err := am.GetUsersFromAccount(context.Background(), mockAccountID, mockUserID)

relay/cmd/root.go

@@ -6,6 +6,7 @@ import (
 	"crypto/tls"
 	"errors"
 	"fmt"
+	"net"
 	"net/http"
 	"os"
 	"os/signal"
@@ -22,6 +23,7 @@ import (
 	"github.com/netbirdio/netbird/relay/server"
 	"github.com/netbirdio/netbird/shared/relay/auth"
 	"github.com/netbirdio/netbird/signal/metrics"
+	"github.com/netbirdio/netbird/stun"
 	"github.com/netbirdio/netbird/util"
 )
 
@@ -43,6 +45,10 @@ type Config struct {
 	LogLevel                 string
 	LogFile                  string
 	HealthcheckListenAddress string
+	// STUN server configuration
+	EnableSTUN   bool
+	STUNPorts    []int
+	STUNLogLevel string
 }
 
 func (c Config) Validate() error {
@@ -52,6 +58,25 @@ func (c Config) Validate() error {
 	if c.AuthSecret == "" {
 		return fmt.Errorf("auth secret is required")
 	}
+
+	// Validate STUN configuration
+	if c.EnableSTUN {
+		if len(c.STUNPorts) == 0 {
+			return fmt.Errorf("--stun-ports is required when --enable-stun is set")
+		}
+
+		seen := make(map[int]bool)
+		for _, port := range c.STUNPorts {
+			if port <= 0 || port > 65535 {
+				return fmt.Errorf("invalid STUN port %d: must be between 1 and 65535", port)
+			}
+			if seen[port] {
+				return fmt.Errorf("duplicate STUN port %d", port)
+			}
+			seen[port] = true
+		}
+	}
+
 	return nil
 }
 
@@ -91,6 +116,9 @@ func init() {
 	rootCmd.PersistentFlags().StringVar(&cobraConfig.LogLevel, "log-level", "info", "log level")
 	rootCmd.PersistentFlags().StringVar(&cobraConfig.LogFile, "log-file", "console", "log file")
 	rootCmd.PersistentFlags().StringVarP(&cobraConfig.HealthcheckListenAddress, "health-listen-address", "H", ":9000", "listen address of healthcheck server")
+	rootCmd.PersistentFlags().BoolVar(&cobraConfig.EnableSTUN, "enable-stun", false, "enable embedded STUN server")
+	rootCmd.PersistentFlags().IntSliceVar(&cobraConfig.STUNPorts, "stun-ports", []int{3478}, "ports for the embedded STUN server (can be specified multiple times or comma-separated)")
+	rootCmd.PersistentFlags().StringVar(&cobraConfig.STUNLogLevel, "stun-log-level", "info", "log level for STUN server (panic, fatal, error, warn, info, debug, trace)")
 
 	setFlagsFromEnvVars(rootCmd)
 }
@@ -119,21 +147,14 @@ func execute(cmd *cobra.Command, args []string) error {
 		return fmt.Errorf("failed to initialize log: %s", err)
 	}
 
+	// Resource creation phase (fail fast before starting any goroutines)
+
 	metricsServer, err := metrics.NewServer(cobraConfig.MetricsPort, "")
 	if err != nil {
 		log.Debugf("setup metrics: %v", err)
 		return fmt.Errorf("setup metrics: %v", err)
 	}
 
-	wg.Add(1)
-	go func() {
-		defer wg.Done()
-		log.Infof("running metrics server: %s%s", metricsServer.Addr, metricsServer.Endpoint)
-		if err := metricsServer.ListenAndServe(); !errors.Is(err, http.ErrServerClosed) {
-			log.Fatalf("Failed to start metrics server: %v", err)
-		}
-	}()
-
 	srvListenerCfg := server.ListenerConfig{
 		Address: cobraConfig.ListenAddress,
 	}
@@ -145,6 +166,12 @@ func execute(cmd *cobra.Command, args []string) error {
 	}
 	srvListenerCfg.TLSConfig = tlsConfig
 
+	// Create STUN listeners early to fail fast
+	stunListeners, err := createSTUNListeners()
+	if err != nil {
+		return err
+	}
+
 	hashedSecret := sha256.Sum256([]byte(cobraConfig.AuthSecret))
 	authenticator := auth.NewTimedHMACValidator(hashedSecret[:], 24*time.Hour)
 
@@ -155,60 +182,145 @@ func execute(cmd *cobra.Command, args []string) error {
 		TLSSupport:     tlsSupport,
 	}
 
-	srv, err := server.NewServer(cfg)
+	srv, err := createRelayServer(cfg)
 	if err != nil {
-		log.Debugf("failed to create relay server: %v", err)
-		return fmt.Errorf("failed to create relay server: %v", err)
+		cleanupSTUNListeners(stunListeners)
+		return err
 	}
+
+	hCfg := healthcheck.Config{
+		ListenAddress:  cobraConfig.HealthcheckListenAddress,
+		ServiceChecker: srv,
+	}
+	httpHealthcheck, err := createHealthCheck(hCfg)
+	if err != nil {
+		cleanupSTUNListeners(stunListeners)
+		return err
+	}
+
+	var stunServer *stun.Server
+	if len(stunListeners) > 0 {
+		stunServer = stun.NewServer(stunListeners, cobraConfig.STUNLogLevel)
+	}
+
+	// Start all servers (only after all resources are successfully created)
+	startServers(&wg, metricsServer, srv, srvListenerCfg, httpHealthcheck, stunServer)
+
+	waitForExitSignal()
+
+	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer cancel()
+
+	err = shutdownServers(ctx, metricsServer, srv, httpHealthcheck, stunServer)
+	wg.Wait()
+	return err
+}
+
+func startServers(wg *sync.WaitGroup, metricsServer *metrics.Metrics, srv *server.Server, srvListenerCfg server.ListenerConfig, httpHealthcheck *healthcheck.Server, stunServer *stun.Server) {
+	wg.Add(1)
+	go func() {
+		defer wg.Done()
+		log.Infof("running metrics server: %s%s", metricsServer.Addr, metricsServer.Endpoint)
+		if err := metricsServer.ListenAndServe(); !errors.Is(err, http.ErrServerClosed) {
+			log.Fatalf("failed to start metrics server: %v", err)
+		}
+	}()
+
 	instanceURL := srv.InstanceURL()
 	log.Infof("server will be available on: %s", instanceURL.String())
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
 		if err := srv.Listen(srvListenerCfg); err != nil {
-			log.Fatalf("failed to bind server: %s", err)
+			log.Fatalf("failed to bind relay server: %s", err)
 		}
 	}()
 
-	hCfg := healthcheck.Config{
-		ListenAddress:  cobraConfig.HealthcheckListenAddress,
-		ServiceChecker: srv,
-	}
-	httpHealthcheck, err := healthcheck.NewServer(hCfg)
-	if err != nil {
-		log.Debugf("failed to create healthcheck server: %v", err)
-		return fmt.Errorf("failed to create healthcheck server: %v", err)
-	}
 	wg.Add(1)
 	go func() {
 		defer wg.Done()
 		if err := httpHealthcheck.ListenAndServe(); !errors.Is(err, http.ErrServerClosed) {
-			log.Fatalf("Failed to start healthcheck server: %v", err)
+			log.Fatalf("failed to start healthcheck server: %v", err)
 		}
 	}()
 
-	// it will block until exit signal
-	waitForExitSignal()
+	if stunServer != nil {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			if err := stunServer.Listen(); err != nil {
+				if errors.Is(err, stun.ErrServerClosed) {
+					return
+				}
+				log.Errorf("STUN server error: %v", err)
+			}
+		}()
+	}
+}
 
-	ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
-	defer cancel()
+func shutdownServers(ctx context.Context, metricsServer *metrics.Metrics, srv *server.Server, httpHealthcheck *healthcheck.Server, stunServer *stun.Server) error {
+	var errs error
 
-	var shutDownErrors error
 	if err := httpHealthcheck.Shutdown(ctx); err != nil {
-		shutDownErrors = multierror.Append(shutDownErrors, fmt.Errorf("failed to close healthcheck server: %v", err))
+		errs = multierror.Append(errs, fmt.Errorf("failed to close healthcheck server: %w", err))
+	}
+
+	if stunServer != nil {
+		if err := stunServer.Shutdown(); err != nil {
+			errs = multierror.Append(errs, fmt.Errorf("failed to close STUN server: %w", err))
+		}
 	}
 
 	if err := srv.Shutdown(ctx); err != nil {
-		shutDownErrors = multierror.Append(shutDownErrors, fmt.Errorf("failed to close server: %s", err))
+		errs = multierror.Append(errs, fmt.Errorf("failed to close relay server: %w", err))
 	}
 
 	log.Infof("shutting down metrics server")
 	if err := metricsServer.Shutdown(ctx); err != nil {
-		shutDownErrors = multierror.Append(shutDownErrors, fmt.Errorf("failed to close metrics server: %v", err))
+		errs = multierror.Append(errs, fmt.Errorf("failed to close metrics server: %w", err))
 	}
 
-	wg.Wait()
-	return shutDownErrors
+	return errs
+}
+
+func createHealthCheck(hCfg healthcheck.Config) (*healthcheck.Server, error) {
+	httpHealthcheck, err := healthcheck.NewServer(hCfg)
+	if err != nil {
+		log.Debugf("failed to create healthcheck server: %v", err)
+		return nil, fmt.Errorf("failed to create healthcheck server: %v", err)
+	}
+	return httpHealthcheck, nil
+}
+
+func createRelayServer(cfg server.Config) (*server.Server, error) {
+	srv, err := server.NewServer(cfg)
+	if err != nil {
+		return nil, fmt.Errorf("failed to create relay server: %v", err)
+	}
+	return srv, nil
+}
+
+func cleanupSTUNListeners(stunListeners []*net.UDPConn) {
+	for _, l := range stunListeners {
+		_ = l.Close()
+	}
+}
+
+func createSTUNListeners() ([]*net.UDPConn, error) {
+	var stunListeners []*net.UDPConn
+	if cobraConfig.EnableSTUN {
+		for _, port := range cobraConfig.STUNPorts {
+			listener, err := net.ListenUDP("udp", &net.UDPAddr{Port: port})
+			if err != nil {
+				// Close already opened listeners on failure
+				cleanupSTUNListeners(stunListeners)
+				log.Debugf("failed to create STUN listener on port %d: %v", port, err)
+				return nil, fmt.Errorf("failed to create STUN listener on port %d: %v", port, err)
+			}
+			stunListeners = append(stunListeners, listener)
+		}
+	}
+	return stunListeners, nil
 }
 
 func handleTLSConfig(cfg *Config) (*tls.Config, bool, error) {

release_files/ui-post-install.sh

@@ -8,6 +8,10 @@ pid="$(pgrep -x -f /usr/bin/netbird-ui || true)"
 if [ -n "${pid}" ]
 then
   uid="$(cat /proc/"${pid}"/loginuid)"
+  # loginuid can be 4294967295 (-1) if not set, fall back to process uid
+  if [ "${uid}" = "4294967295" ] || [ "${uid}" = "-1" ]; then
+    uid="$(stat -c '%u' /proc/"${pid}")"
+  fi
   username="$(id -nu "${uid}")"
   # Only re-run if it was already running
   pkill -x -f /usr/bin/netbird-ui >/dev/null 2>&1

shared/management/proto/management.pb.go

@@ -2873,7 +2873,7 @@ type CustomZone struct {
 	Domain               string          `protobuf:"bytes,1,opt,name=Domain,proto3" json:"Domain,omitempty"`
 	Records              []*SimpleRecord `protobuf:"bytes,2,rep,name=Records,proto3" json:"Records,omitempty"`
 	SearchDomainDisabled bool            `protobuf:"varint,3,opt,name=SearchDomainDisabled,proto3" json:"SearchDomainDisabled,omitempty"`
-	SkipPTRProcess       bool            `protobuf:"varint,4,opt,name=SkipPTRProcess,proto3" json:"SkipPTRProcess,omitempty"`
+	NonAuthoritative     bool            `protobuf:"varint,4,opt,name=NonAuthoritative,proto3" json:"NonAuthoritative,omitempty"`
 }
 
 func (x *CustomZone) Reset() {
@@ -2929,9 +2929,9 @@ func (x *CustomZone) GetSearchDomainDisabled() bool {
 	return false
 }
 
-func (x *CustomZone) GetSkipPTRProcess() bool {
+func (x *CustomZone) GetNonAuthoritative() bool {
 	if x != nil {
-		return x.SkipPTRProcess
+		return x.NonAuthoritative
 	}
 	return false
 }

shared/management/proto/management.proto

@@ -464,7 +464,9 @@ message CustomZone {
   string Domain = 1;
   repeated SimpleRecord Records = 2;
   bool SearchDomainDisabled = 3;
-  bool SkipPTRProcess = 4;
+  // NonAuthoritative indicates this is a user-created zone (not the built-in peer DNS zone).
+  // Non-authoritative zones will fallthrough to lower-priority handlers on NXDOMAIN and skip PTR processing.
+  bool NonAuthoritative = 4;
 }
 
 // SimpleRecord represents a dns.SimpleRecord

stun/server.go

@@ -0,0 +1,170 @@
+// Package stun provides an embedded STUN server for NAT traversal discovery.
+package stun
+
+import (
+	"errors"
+	"fmt"
+	"net"
+	"sync"
+
+	"github.com/hashicorp/go-multierror"
+	nberrors "github.com/netbirdio/netbird/client/errors"
+	log "github.com/sirupsen/logrus"
+
+	"github.com/netbirdio/netbird/formatter"
+	"github.com/pion/stun/v3"
+)
+
+// ErrServerClosed is returned by Listen when the server is shut down gracefully.
+var ErrServerClosed = errors.New("stun: server closed")
+
+// ErrNoListeners is returned by Listen when no UDP connections were provided.
+var ErrNoListeners = errors.New("stun: no listeners configured")
+
+// Server implements a STUN server that responds to binding requests
+// with the client's reflexive transport address.
+type Server struct {
+	conns    []*net.UDPConn
+	logger   *log.Entry
+	logLevel log.Level
+
+	wg sync.WaitGroup
+}
+
+// NewServer creates a new STUN server with the given UDP listeners.
+// The caller is responsible for creating and providing the listeners.
+// logLevel can be: panic, fatal, error, warn, info, debug, trace
+func NewServer(conns []*net.UDPConn, logLevel string) *Server {
+	level, err := log.ParseLevel(logLevel)
+	if err != nil {
+		level = log.InfoLevel
+	}
+
+	// Create a separate logger with its own level setting
+	// This allows --stun-log-level to work independently of --log-level
+	stunLogger := log.New()
+	stunLogger.SetOutput(log.StandardLogger().Out)
+	stunLogger.SetLevel(level)
+	// Use the formatter package to set up formatter, ReportCaller, and context hook
+	formatter.SetTextFormatter(stunLogger)
+
+	logger := stunLogger.WithField("component", "stun-server")
+	logger.Infof("STUN server log level set to: %s", level.String())
+
+	return &Server{
+		conns:    conns,
+		logger:   logger,
+		logLevel: level,
+	}
+}
+
+// Listen starts the STUN server and blocks until the server is shut down.
+// Returns ErrServerClosed when shut down gracefully via Shutdown.
+// Returns ErrNoListeners if no UDP connections were provided.
+func (s *Server) Listen() error {
+	if len(s.conns) == 0 {
+		return ErrNoListeners
+	}
+
+	// Start a read loop for each listener
+	for _, conn := range s.conns {
+		s.logger.Infof("STUN server listening on %s", conn.LocalAddr())
+		s.wg.Add(1)
+		go s.readLoop(conn)
+	}
+
+	s.wg.Wait()
+	return ErrServerClosed
+}
+
+// readLoop continuously reads UDP packets and handles STUN requests.
+func (s *Server) readLoop(conn *net.UDPConn) {
+	defer s.wg.Done()
+	buf := make([]byte, 1500) // Standard MTU size
+	for {
+		n, remoteAddr, err := conn.ReadFromUDP(buf)
+
+		if err != nil {
+			// Check if the connection was closed externally
+			if errors.Is(err, net.ErrClosed) {
+				s.logger.Info("UDP connection closed, stopping read loop")
+				return
+			}
+			s.logger.Warnf("failed to read UDP packet: %v", err)
+			continue
+		}
+
+		// Handle packet in the same goroutine to avoid complexity
+		// STUN responses are small and fast
+		s.handlePacket(conn, buf[:n], remoteAddr)
+	}
+}
+
+// handlePacket processes a STUN request and sends a response.
+func (s *Server) handlePacket(conn *net.UDPConn, data []byte, addr *net.UDPAddr) {
+	localPort := conn.LocalAddr().(*net.UDPAddr).Port
+
+	s.logger.Debugf("[port:%d] received %d bytes from %s", localPort, len(data), addr)
+
+	// Check if it's a STUN message
+	if !stun.IsMessage(data) {
+		s.logger.Debugf("[port:%d] not a STUN message (first bytes: %x)", localPort, data[:min(len(data), 8)])
+		return
+	}
+
+	// Parse the STUN message
+	msg := &stun.Message{Raw: data}
+	if err := msg.Decode(); err != nil {
+		s.logger.Warnf("[port:%d] failed to decode STUN message from %s: %v", localPort, addr, err)
+		return
+	}
+
+	s.logger.Debugf("[port:%d] received STUN %s from %s (tx=%x)", localPort, msg.Type, addr, msg.TransactionID[:8])
+
+	// Only handle binding requests
+	if msg.Type != stun.BindingRequest {
+		s.logger.Debugf("[port:%d] ignoring non-binding request: %s", localPort, msg.Type)
+		return
+	}
+
+	// Build the response
+	response, err := stun.Build(
+		stun.NewTransactionIDSetter(msg.TransactionID),
+		stun.BindingSuccess,
+		&stun.XORMappedAddress{
+			IP:   addr.IP,
+			Port: addr.Port,
+		},
+		stun.Fingerprint,
+	)
+	if err != nil {
+		s.logger.Errorf("[port:%d] failed to build STUN response: %v", localPort, err)
+		return
+	}
+
+	// Send the response on the same connection it was received on
+	n, err := conn.WriteToUDP(response.Raw, addr)
+	if err != nil {
+		s.logger.Errorf("[port:%d] failed to send STUN response to %s: %v", localPort, addr, err)
+		return
+	}
+
+	s.logger.Debugf("[port:%d] sent STUN BindingSuccess to %s (%d bytes) with XORMappedAddress %s:%d", localPort, addr, n, addr.IP, addr.Port)
+}
+
+// Shutdown gracefully stops the STUN server.
+func (s *Server) Shutdown() error {
+	s.logger.Info("shutting down STUN server")
+
+	var merr *multierror.Error
+
+	for _, conn := range s.conns {
+		if err := conn.Close(); err != nil && !errors.Is(err, net.ErrClosed) {
+			merr = multierror.Append(merr, fmt.Errorf("close STUN UDP connection: %w", err))
+		}
+	}
+
+	// Wait for all readLoops to finish
+	s.wg.Wait()
+	return nberrors.FormatErrorOrNil(merr)
+}

stun/server_test.go

@@ -0,0 +1,479 @@
+package stun
+
+import (
+	"errors"
+	"fmt"
+	"math/rand"
+	"net"
+	"os"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	"github.com/pion/stun/v3"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+// createTestServer creates a STUN server listening on a random port for testing.
+// Returns the server, the listener connection (caller must close), and the server address.
+func createTestServer(t testing.TB) (*Server, *net.UDPConn, *net.UDPAddr) {
+	t.Helper()
+	conn, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.ParseIP("127.0.0.1"), Port: 0})
+	require.NoError(t, err)
+	server := NewServer([]*net.UDPConn{conn}, "debug")
+	return server, conn, conn.LocalAddr().(*net.UDPAddr)
+}
+
+// waitForServerReady polls the server with STUN binding requests until it responds.
+// This avoids flaky tests on slow CI machines that relied on time.Sleep.
+func waitForServerReady(t testing.TB, serverAddr *net.UDPAddr, timeout time.Duration) {
+	t.Helper()
+	deadline := time.Now().Add(timeout)
+	retryInterval := 10 * time.Millisecond
+
+	clientConn, err := net.DialUDP("udp", nil, serverAddr)
+	require.NoError(t, err)
+	defer clientConn.Close()
+
+	buf := make([]byte, 1500)
+	for time.Now().Before(deadline) {
+		msg, err := stun.Build(stun.TransactionID, stun.BindingRequest)
+		require.NoError(t, err)
+
+		_, err = clientConn.Write(msg.Raw)
+		require.NoError(t, err)
+
+		_ = clientConn.SetReadDeadline(time.Now().Add(retryInterval))
+		n, err := clientConn.Read(buf)
+		if err != nil {
+			// Timeout or other error, retry
+			continue
+		}
+
+		response := &stun.Message{Raw: buf[:n]}
+		if err := response.Decode(); err != nil {
+			continue
+		}
+
+		if response.Type == stun.BindingSuccess {
+			return // Server is ready
+		}
+	}
+
+	t.Fatalf("server did not become ready within %v", timeout)
+}
+
+func TestServer_BindingRequest(t *testing.T) {
+	// Start the STUN server on a random port
+	server, listener, serverAddr := createTestServer(t)
+
+	// Start server in background
+	serverErrCh := make(chan error, 1)
+	go func() {
+		serverErrCh <- server.Listen()
+	}()
+
+	// Wait for server to be ready
+	waitForServerReady(t, serverAddr, 2*time.Second)
+
+	// Create a UDP client
+	clientConn, err := net.DialUDP("udp", nil, serverAddr)
+	require.NoError(t, err)
+	defer clientConn.Close()
+
+	// Build a STUN binding request
+	msg, err := stun.Build(stun.TransactionID, stun.BindingRequest)
+	require.NoError(t, err)
+
+	// Send the request
+	_, err = clientConn.Write(msg.Raw)
+	require.NoError(t, err)
+
+	// Read the response
+	buf := make([]byte, 1500)
+	_ = clientConn.SetReadDeadline(time.Now().Add(2 * time.Second))
+	n, err := clientConn.Read(buf)
+	require.NoError(t, err)
+
+	// Parse the response
+	response := &stun.Message{Raw: buf[:n]}
+	err = response.Decode()
+	require.NoError(t, err)
+
+	// Verify it's a binding success
+	assert.Equal(t, stun.BindingSuccess, response.Type)
+
+	// Extract the XOR-MAPPED-ADDRESS
+	var xorAddr stun.XORMappedAddress
+	err = xorAddr.GetFrom(response)
+	require.NoError(t, err)
+
+	// Verify the address matches our client's local address
+	clientAddr := clientConn.LocalAddr().(*net.UDPAddr)
+	assert.Equal(t, clientAddr.IP.String(), xorAddr.IP.String())
+	assert.Equal(t, clientAddr.Port, xorAddr.Port)
+
+	// Close listener first to unblock readLoop, then shutdown
+	_ = listener.Close()
+	err = server.Shutdown()
+	require.NoError(t, err)
+}
+
+func TestServer_IgnoresNonSTUNPackets(t *testing.T) {
+	server, listener, serverAddr := createTestServer(t)
+
+	go func() {
+		_ = server.Listen()
+	}()
+
+	waitForServerReady(t, serverAddr, 2*time.Second)
+
+	clientConn, err := net.DialUDP("udp", nil, serverAddr)
+	require.NoError(t, err)
+	defer clientConn.Close()
+
+	// Send non-STUN data
+	_, err = clientConn.Write([]byte("hello world"))
+	require.NoError(t, err)
+
+	// Try to read response (should timeout since server ignores non-STUN)
+	buf := make([]byte, 1500)
+	_ = clientConn.SetReadDeadline(time.Now().Add(200 * time.Millisecond))
+	_, err = clientConn.Read(buf)
+	assert.Error(t, err) // Should be a timeout error
+
+	// Close listener first to unblock readLoop, then shutdown
+	_ = listener.Close()
+	_ = server.Shutdown()
+}
+
+func TestServer_Shutdown(t *testing.T) {
+	server, listener, serverAddr := createTestServer(t)
+
+	serverDone := make(chan struct{})
+	go func() {
+		err := server.Listen()
+		assert.True(t, errors.Is(err, ErrServerClosed))
+		close(serverDone)
+	}()
+
+	waitForServerReady(t, serverAddr, 2*time.Second)
+
+	// Close listener first to unblock readLoop, then shutdown
+	_ = listener.Close()
+
+	err := server.Shutdown()
+	require.NoError(t, err)
+
+	// Wait for Listen to return
+	select {
+	case <-serverDone:
+		// Success
+	case <-time.After(3 * time.Second):
+		t.Fatal("server did not shutdown in time")
+	}
+}
+
+func TestServer_MultipleRequests(t *testing.T) {
+	server, listener, serverAddr := createTestServer(t)
+
+	go func() {
+		_ = server.Listen()
+	}()
+
+	waitForServerReady(t, serverAddr, 2*time.Second)
+
+	// Create multiple clients and send requests
+	for i := 0; i < 5; i++ {
+		func() {
+			clientConn, err := net.DialUDP("udp", nil, serverAddr)
+			require.NoError(t, err)
+			defer clientConn.Close()
+
+			msg, err := stun.Build(stun.TransactionID, stun.BindingRequest)
+			require.NoError(t, err)
+
+			_, err = clientConn.Write(msg.Raw)
+			require.NoError(t, err)
+
+			buf := make([]byte, 1500)
+			_ = clientConn.SetReadDeadline(time.Now().Add(2 * time.Second))
+			n, err := clientConn.Read(buf)
+			require.NoError(t, err)
+
+			response := &stun.Message{Raw: buf[:n]}
+			err = response.Decode()
+			require.NoError(t, err)
+
+			assert.Equal(t, stun.BindingSuccess, response.Type)
+		}()
+	}
+
+	// Close listener first to unblock readLoop, then shutdown
+	_ = listener.Close()
+	_ = server.Shutdown()
+}
+
+func TestServer_ConcurrentClients(t *testing.T) {
+	numClients := 100
+	requestsPerClient := 5
+	maxStartDelay := 100 * time.Millisecond   // Random delay before client starts
+	maxRequestDelay := 500 * time.Millisecond // Random delay between requests
+
+	// Remote server to test against via env var STUN_TEST_SERVER
+	// Example: STUN_TEST_SERVER=example.netbird.io:3478 go test -v ./stun/... -run ConcurrentClients
+	remoteServer := os.Getenv("STUN_TEST_SERVER")
+
+	var serverAddr *net.UDPAddr
+	var server *Server
+	var listener *net.UDPConn
+
+	if remoteServer != "" {
+		// Use remote server
+		var err error
+		serverAddr, err = net.ResolveUDPAddr("udp", remoteServer)
+		require.NoError(t, err)
+		t.Logf("Testing against remote server: %s", remoteServer)
+	} else {
+		// Start local server
+		server, listener, serverAddr = createTestServer(t)
+		go func() {
+			_ = server.Listen()
+		}()
+		waitForServerReady(t, serverAddr, 2*time.Second)
+		t.Logf("Testing against local server: %s", serverAddr)
+	}
+
+	var wg sync.WaitGroup
+	errorz := make(chan error, numClients*requestsPerClient)
+	successCount := make(chan int, numClients)
+
+	startTime := time.Now()
+
+	for i := 0; i < numClients; i++ {
+		wg.Add(1)
+		go func(clientID int) {
+			defer wg.Done()
+
+			// Random delay before starting
+			time.Sleep(time.Duration(rand.Int63n(int64(maxStartDelay))))
+
+			clientConn, err := net.DialUDP("udp", nil, serverAddr)
+			if err != nil {
+				errorz <- fmt.Errorf("client %d: failed to dial: %w", clientID, err)
+				return
+			}
+			defer clientConn.Close()
+
+			success := 0
+			for j := 0; j < requestsPerClient; j++ {
+				// Random delay between requests
+				if j > 0 {
+					time.Sleep(time.Duration(rand.Int63n(int64(maxRequestDelay))))
+				}
+
+				msg, err := stun.Build(stun.TransactionID, stun.BindingRequest)
+				if err != nil {
+					errorz <- fmt.Errorf("client %d: failed to build request: %w", clientID, err)
+					continue
+				}
+
+				_, err = clientConn.Write(msg.Raw)
+				if err != nil {
+					errorz <- fmt.Errorf("client %d: failed to write: %w", clientID, err)
+					continue
+				}
+
+				buf := make([]byte, 1500)
+				_ = clientConn.SetReadDeadline(time.Now().Add(5 * time.Second))
+				n, err := clientConn.Read(buf)
+				if err != nil {
+					errorz <- fmt.Errorf("client %d: failed to read: %w", clientID, err)
+					continue
+				}
+
+				response := &stun.Message{Raw: buf[:n]}
+				if err := response.Decode(); err != nil {
+					errorz <- fmt.Errorf("client %d: failed to decode: %w", clientID, err)
+					continue
+				}
+
+				if response.Type != stun.BindingSuccess {
+					errorz <- fmt.Errorf("client %d: unexpected response type: %s", clientID, response.Type)
+					continue
+				}
+
+				success++
+			}
+			successCount <- success
+		}(i)
+	}
+
+	wg.Wait()
+	close(errorz)
+	close(successCount)
+
+	elapsed := time.Since(startTime)
+
+	totalSuccess := 0
+	for count := range successCount {
+		totalSuccess += count
+	}
+
+	var errs []error
+	for err := range errorz {
+		errs = append(errs, err)
+	}
+
+	totalRequests := numClients * requestsPerClient
+	t.Logf("Completed %d/%d requests in %v (%.2f req/s)",
+		totalSuccess, totalRequests, elapsed,
+		float64(totalSuccess)/elapsed.Seconds())
+
+	if len(errs) > 0 {
+		t.Logf("Errors (%d):", len(errs))
+		for i, err := range errs {
+			if i < 10 { // Only show first 10 errors
+				t.Logf("  - %v", err)
+			}
+		}
+	}
+
+	// Require at least 95% success rate
+	successRate := float64(totalSuccess) / float64(totalRequests)
+	require.GreaterOrEqual(t, successRate, 0.95, "success rate too low: %.2f%%", successRate*100)
+
+	// Cleanup local server if used
+	if server != nil {
+		// Close listener first to unblock readLoop, then shutdown
+		_ = listener.Close()
+		_ = server.Shutdown()
+	}
+}
+
+func TestServer_MultiplePorts(t *testing.T) {
+	// Create listeners on two random ports
+	conn1, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.ParseIP("127.0.0.1"), Port: 0})
+	require.NoError(t, err)
+	conn2, err := net.ListenUDP("udp", &net.UDPAddr{IP: net.ParseIP("127.0.0.1"), Port: 0})
+	require.NoError(t, err)
+
+	addr1 := conn1.LocalAddr().(*net.UDPAddr)
+	addr2 := conn2.LocalAddr().(*net.UDPAddr)
+
+	server := NewServer([]*net.UDPConn{conn1, conn2}, "debug")
+
+	go func() {
+		_ = server.Listen()
+	}()
+
+	// Wait for server to be ready (checking first port is sufficient)
+	waitForServerReady(t, addr1, 2*time.Second)
+
+	// Test requests on both ports
+	for _, serverAddr := range []*net.UDPAddr{addr1, addr2} {
+		func() {
+			clientConn, err := net.DialUDP("udp", nil, serverAddr)
+			require.NoError(t, err)
+			defer clientConn.Close()
+
+			msg, err := stun.Build(stun.TransactionID, stun.BindingRequest)
+			require.NoError(t, err)
+
+			_, err = clientConn.Write(msg.Raw)
+			require.NoError(t, err)
+
+			buf := make([]byte, 1500)
+			_ = clientConn.SetReadDeadline(time.Now().Add(2 * time.Second))
+			n, err := clientConn.Read(buf)
+			require.NoError(t, err)
+
+			response := &stun.Message{Raw: buf[:n]}
+			err = response.Decode()
+			require.NoError(t, err)
+
+			assert.Equal(t, stun.BindingSuccess, response.Type)
+
+			var xorAddr stun.XORMappedAddress
+			err = xorAddr.GetFrom(response)
+			require.NoError(t, err)
+
+			clientAddr := clientConn.LocalAddr().(*net.UDPAddr)
+			assert.Equal(t, clientAddr.Port, xorAddr.Port)
+		}()
+	}
+
+	// Close listeners first to unblock readLoops, then shutdown
+	_ = conn1.Close()
+	_ = conn2.Close()
+	_ = server.Shutdown()
+}
+
+// BenchmarkSTUNServer benchmarks the STUN server with concurrent clients
+func BenchmarkSTUNServer(b *testing.B) {
+	server, listener, serverAddr := createTestServer(b)
+
+	go func() {
+		_ = server.Listen()
+	}()
+
+	waitForServerReady(b, serverAddr, 2*time.Second)
+
+	// Capture first error atomically - b.Fatal cannot be called from worker goroutines
+	var firstErr atomic.Pointer[error]
+	setErr := func(err error) {
+		firstErr.CompareAndSwap(nil, &err)
+	}
+
+	b.ResetTimer()
+	b.RunParallel(func(pb *testing.PB) {
+		// Stop work if an error has occurred
+		if firstErr.Load() != nil {
+			return
+		}
+
+		clientConn, err := net.DialUDP("udp", nil, serverAddr)
+		if err != nil {
+			setErr(err)
+			return
+		}
+		defer clientConn.Close()
+
+		buf := make([]byte, 1500)
+
+		for pb.Next() {
+			if firstErr.Load() != nil {
+				return
+			}
+
+			msg, _ := stun.Build(stun.TransactionID, stun.BindingRequest)
+			_, _ = clientConn.Write(msg.Raw)
+
+			_ = clientConn.SetReadDeadline(time.Now().Add(2 * time.Second))
+			n, err := clientConn.Read(buf)
+			if err != nil {
+				setErr(err)
+				return
+			}
+
+			response := &stun.Message{Raw: buf[:n]}
+			if err := response.Decode(); err != nil {
+				setErr(err)
+				return
+			}
+		}
+	})
+
+	b.StopTimer()
+
+	// Fail after RunParallel completes
+	if errPtr := firstErr.Load(); errPtr != nil {
+		b.Fatal(*errPtr)
+	}
+
+	// Close listener first to unblock readLoop, then shutdown
+	_ = listener.Close()
+	_ = server.Shutdown()
+}