Files
netbird/client/android/client.go
T
Zoltan Papp 6b3cfbabd2 [client] Resolve the Android network route peer by HA unique ID instead of scanning the full status (#7705)
* [client] Build the Android network list from a single peer snapshot

Networks() called GetFullStatus() once per network to find the peer
that serves it, copying every peer state and taking eight recorder
locks each time. With 100+ peers and the UI calling Networks() from
every peer list change, this queued hundreds of callers on the status
recorder lock. Take one snapshot per call and index it by route.

* [client] Track the active route peer by HA unique ID in the status recorder

The Android network list resolved the route owner by scanning peer state
route keys. Those keys are the handler string: a prefix for static routes
and the domain pattern for dynamic ones, so the prefix-based lookup never
matched dynamic routes, and two networks sharing a prefix resolved to the
same owner.

The route watcher now records the chosen route peer under the route's HA
unique ID in the status recorder, and the Android binding looks the owner
up by that ID. The key is unique per network and independent of the
handler string format, so both anomalies are gone. The prefix-based
routeOwners helper is removed.

* [client] Record the active route peer before notifying listeners

AddPeerStateRoute and RemovePeerStateRoute fire the peer list change
callback and wake the status subscribers. The active route peer mapping
was written after those calls, so a Networks() call landing in between
found no mapping for the network and fell back to the first connected
peer, or kept showing the previous peer on removal. Nothing re-notified
after the mapping write, so the wrong peer stayed until the next peer
list change.

Write and delete the mapping before the notifying calls so a listener
reacting to the notification always reads the current owner.
2026-10-05 15:11:08 +02:00

693 lines
22 KiB
Go

//go:build android
package android
import (
"context"
"fmt"
"os"
"slices"
"strings"
"sync"
"sync/atomic"
"time"
"golang.org/x/exp/maps"
log "github.com/sirupsen/logrus"
nbAnonymize "github.com/netbirdio/netbird/client/anonymize"
"github.com/netbirdio/netbird/client/iface/device"
"github.com/netbirdio/netbird/client/internal"
"github.com/netbirdio/netbird/client/internal/debug"
"github.com/netbirdio/netbird/client/internal/dns"
"github.com/netbirdio/netbird/client/internal/listener"
"github.com/netbirdio/netbird/client/internal/peer"
"github.com/netbirdio/netbird/client/internal/profilemanager"
"github.com/netbirdio/netbird/client/internal/routemanager"
"github.com/netbirdio/netbird/client/internal/stdnet"
"github.com/netbirdio/netbird/client/net"
"github.com/netbirdio/netbird/client/netevents"
"github.com/netbirdio/netbird/client/system"
"github.com/netbirdio/netbird/formatter"
"github.com/netbirdio/netbird/route"
"github.com/netbirdio/netbird/shared/management/domain"
types "github.com/netbirdio/netbird/upload-server/types"
)
// AnonymizeLevelDefault and AnonymizeLevelStrict are the accepted
// anonymizeLevel values for DebugBundle.
const (
AnonymizeLevelDefault = nbAnonymize.LevelDefaultString
AnonymizeLevelStrict = nbAnonymize.LevelStrictString
)
// TunAdapter export internal TunAdapter for mobile
type TunAdapter interface {
device.TunAdapter
}
// IFaceDiscover export internal IFaceDiscover for mobile
type IFaceDiscover interface {
stdnet.ExternalIFaceDiscover
}
// NetworkChangeListener export internal NetworkChangeListener for mobile
type NetworkChangeListener interface {
listener.NetworkChangeListener
}
// DnsReadyListener export internal dns ReadyListener for mobile
type DnsReadyListener interface {
dns.ReadyListener
}
// TunSettings is a snapshot of the settings the TUN device is rebuilt with
type TunSettings struct {
Routes string
SearchDomains string
}
func init() {
formatter.SetLogcatFormatter(log.StandardLogger())
}
// Client struct manage the life circle of background service
type Client struct {
tunAdapter device.TunAdapter
iFaceDiscover IFaceDiscover
recorder *peer.Status
ctxCancel context.CancelFunc
ctxCancelLock *sync.Mutex
deviceName string
uiVersion string
networkChangeListener listener.NetworkChangeListener
// netMgr outlives engine restarts: it mirrors the OS connectivity, not
// the engine lifecycle. Run and RunWithoutLogin inject its state and
// sweeper into each new ConnectClient.
netMgr *netevents.Manager
stateMu sync.RWMutex
connectClient *internal.ConnectClient
config *profilemanager.Config
cacheDir string
// mdmSource holds the per-Client MDM policy source and its change
// detector as one unit. Set by SetMDMPolicyFetcher (called from the
// Kotlin side). Each Run passes the loader to the resolved Config so
// applyMDMPolicy picks up the active overlay. Nil means "MDM
// enforcement off for this Client".
mdmSource atomic.Pointer[mdmSource]
// Identifies the running profile for the SSO login hint; see profile_state.go.
cfgPath string
stateChangeMu sync.Mutex
stateChangeSubID string
eventSub *peer.EventSubscription
// Closed to stop the watch goroutines from delivering buffered items to a
// listener that has been removed or replaced. See stopStateChangeWatchLocked.
stateChangeDone chan struct{}
// Latched "the server wants an interactive login": survives the engine
// restarts that replace the run loop's context state. See Client.Status.
// Guarded by loginRequiredMu together with loginCleared, which counts
// clears so a stale observation cannot re-latch over one.
loginRequiredMu sync.Mutex
loginRequired bool
loginCleared uint64
extendMu sync.Mutex
extendCancel context.CancelFunc
}
func (c *Client) setState(cfg *profilemanager.Config, cacheDir string, cfgPath string, cc *internal.ConnectClient) {
c.stateMu.Lock()
defer c.stateMu.Unlock()
c.config = cfg
c.cacheDir = cacheDir
c.cfgPath = cfgPath
c.connectClient = cc
}
func (c *Client) stateSnapshot() (*profilemanager.Config, string, *internal.ConnectClient) {
c.stateMu.RLock()
defer c.stateMu.RUnlock()
return c.config, c.cacheDir, c.connectClient
}
// authSnapshot returns the config together with the path it was loaded from, in
// one lock: the path identifies the profile whose account email backs the login
// hint, so reading it separately could pair one profile's config with another's
// hint when a profile switch lands in between.
func (c *Client) authSnapshot() (*profilemanager.Config, string, *internal.ConnectClient) {
c.stateMu.RLock()
defer c.stateMu.RUnlock()
return c.config, c.cfgPath, c.connectClient
}
func (c *Client) getConnectClient() *internal.ConnectClient {
c.stateMu.RLock()
defer c.stateMu.RUnlock()
return c.connectClient
}
// NewClient instantiate a new Client
func NewClient(androidSDKVersion int, deviceName string, uiVersion string, tunAdapter TunAdapter, iFaceDiscover IFaceDiscover, networkChangeListener NetworkChangeListener) *Client {
execWorkaround(androidSDKVersion)
net.SetAndroidProtectSocketFn(tunAdapter.ProtectSocket)
system.SetIFaceDiscover(iFaceDiscover)
recorder := peer.NewRecorder("")
return &Client{
deviceName: deviceName,
uiVersion: uiVersion,
tunAdapter: tunAdapter,
iFaceDiscover: iFaceDiscover,
recorder: recorder,
ctxCancelLock: &sync.Mutex{},
networkChangeListener: networkChangeListener,
netMgr: netevents.NewManager(recorder),
}
}
// Run start the internal client. It is a blocker function
func (c *Client) Run(platformFiles PlatformFiles, urlOpener URLOpener, isAndroidTV bool, dns *DNSList, dnsReadyListener DnsReadyListener, envList *EnvList) error {
exportEnvList(envList)
cfgFile := platformFiles.ConfigurationFilePath()
stateFile := platformFiles.StateFilePath()
cacheDir := platformFiles.CacheDir()
log.Infof("Starting client with config: %s, state: %s", cfgFile, stateFile)
cfg, err := profilemanager.UpdateOrCreateConfig(profilemanager.ConfigInput{
ConfigPath: cfgFile,
})
if err != nil {
return err
}
c.applyMDMOverlay(cfg)
c.recorder.UpdateManagementAddress(cfg.ManagementURL.String())
c.recorder.UpdateRosenpass(cfg.RosenpassEnabled, cfg.RosenpassPermissive)
var ctx context.Context
//nolint
ctxWithValues := context.WithValue(context.Background(), system.DeviceNameCtxKey, c.deviceName)
//nolint
ctxWithValues = context.WithValue(ctxWithValues, system.UiVersionCtxKey, c.uiVersion)
c.ctxCancelLock.Lock()
ctx, c.ctxCancel = context.WithCancel(ctxWithValues)
defer c.ctxCancel()
c.ctxCancelLock.Unlock()
auth := NewAuthWithConfig(ctx, cfg, cfgFile)
err = auth.login(urlOpener, isAndroidTV)
if err != nil {
return err
}
// todo do not throw error in case of cancelled context
ctx = internal.CtxInitState(ctx)
connectClient := internal.NewConnectClient(ctx, cfg, c.recorder,
internal.WithNetEvents(c.netMgr))
c.setState(cfg, cacheDir, cfgFile, connectClient)
connectClient.SetSyncResponsePersistence(true)
// This path runs the interactive SSO flow, so reaching here means the peer
// is authenticated again — release the latch Status() reports from. Clear
// only once the fresh connect client is installed: until then Status()
// still reads the previous run's context state, which holds the NeedsLogin
// that prompted this login, and would re-latch what was just cleared.
c.clearLoginRequired()
return connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
}
// RunWithoutLogin we apply this type of run function when the backed has been started without UI (i.e. after reboot).
// In this case make no sense handle registration steps.
func (c *Client) RunWithoutLogin(platformFiles PlatformFiles, dns *DNSList, dnsReadyListener DnsReadyListener, envList *EnvList) error {
exportEnvList(envList)
cfgFile := platformFiles.ConfigurationFilePath()
stateFile := platformFiles.StateFilePath()
cacheDir := platformFiles.CacheDir()
log.Infof("Starting client without login with config: %s, state: %s", cfgFile, stateFile)
cfg, err := profilemanager.UpdateOrCreateConfig(profilemanager.ConfigInput{
ConfigPath: cfgFile,
})
if err != nil {
return err
}
c.applyMDMOverlay(cfg)
c.recorder.UpdateManagementAddress(cfg.ManagementURL.String())
c.recorder.UpdateRosenpass(cfg.RosenpassEnabled, cfg.RosenpassPermissive)
var ctx context.Context
//nolint
ctxWithValues := context.WithValue(context.Background(), system.DeviceNameCtxKey, c.deviceName)
c.ctxCancelLock.Lock()
ctx, c.ctxCancel = context.WithCancel(ctxWithValues)
defer c.ctxCancel()
c.ctxCancelLock.Unlock()
// todo do not throw error in case of cancelled context
ctx = internal.CtxInitState(ctx)
connectClient := internal.NewConnectClient(ctx, cfg, c.recorder,
internal.WithNetEvents(c.netMgr))
c.setState(cfg, cacheDir, cfgFile, connectClient)
connectClient.SetSyncResponsePersistence(true)
return connectClient.RunOnAndroid(c.tunAdapter, c.iFaceDiscover, c.networkChangeListener, slices.Clone(dns.items), dnsReadyListener, stateFile, cacheDir)
}
// Stop the internal client and free the resources
func (c *Client) Stop() {
c.ctxCancelLock.Lock()
defer c.ctxCancelLock.Unlock()
if c.ctxCancel == nil {
return
}
c.ctxCancel()
}
func (c *Client) RenewTun(fd int) error {
cc := c.getConnectClient()
if cc == nil {
return fmt.Errorf("engine not running")
}
e := cc.Engine()
if e == nil {
return fmt.Errorf("engine not initialized")
}
return e.RenewTun(fd)
}
func (c *Client) GetTunSettings() (*TunSettings, error) {
cc := c.getConnectClient()
if cc == nil {
return nil, fmt.Errorf("engine not running")
}
e := cc.Engine()
if e == nil {
return nil, fmt.Errorf("engine not initialized")
}
routes, searchDomains := e.TunSettings()
return &TunSettings{
Routes: strings.Join(routes, ";"),
SearchDomains: strings.Join(searchDomains, ";"),
}, nil
}
// SetNetworkAvailable feeds OS-reported network availability into the client.
// While unavailable, the internal reconnect loops suspend their attempts and
// the connection listener reports NoNetwork instead of Connecting; when
// availability returns, the loops resume immediately with a fresh backoff.
// Losing the last network also sweeps the registered connections: nothing can
// redial while offline, so the stale sockets would otherwise stay silently
// "connected" until their own timeouts and the client would keep reporting
// Connected with no network at all.
func (c *Client) SetNetworkAvailable(available bool) {
c.netMgr.SetNetworkAvailable(available)
}
// NotifyNetworkChange marks the management, signal and relay connections
// stale after the OS switched networks and schedules a sweep that cuts
// whatever has not redialed on the new network by then. The engine and the
// TUN device stay untouched.
func (c *Client) NotifyNetworkChange() {
c.netMgr.NotifyNetworkChange()
}
// DebugBundle generates a debug bundle, uploads it, and returns the upload key.
// It works both with and without a running engine. anonymizeLevel is "default"
// or "strict"; strict also anonymizes internal IP ranges, peer names, and
// WireGuard public keys, and implies anonymize.
func (c *Client) DebugBundle(platformFiles PlatformFiles, anonymize bool, anonymizeLevel string) (string, error) {
return c.debugBundle(platformFiles, anonymize, anonymizeLevel, true)
}
// DebugBundleFile generates a debug bundle and returns the path of the zip in
// the cache directory instead of uploading it, so the app can hand the file to
// the user for inspection. The caller owns the file and removes it once done;
// the stale-bundle cleanup of later runs removes it only after a day.
// anonymize and anonymizeLevel behave as in DebugBundle.
func (c *Client) DebugBundleFile(platformFiles PlatformFiles, anonymize bool, anonymizeLevel string) (string, error) {
return c.debugBundle(platformFiles, anonymize, anonymizeLevel, false)
}
func (c *Client) debugBundle(platformFiles PlatformFiles, anonymize bool, anonymizeLevel string, upload bool) (string, error) {
cfg, cacheDir, cc := c.stateSnapshot()
// If the engine hasn't been started, load config from disk
if cfg == nil {
var err error
cfg, err = profilemanager.UpdateOrCreateConfig(profilemanager.ConfigInput{
ConfigPath: platformFiles.ConfigurationFilePath(),
})
if err != nil {
return "", fmt.Errorf("load config: %w", err)
}
c.applyMDMOverlay(cfg)
cacheDir = platformFiles.CacheDir()
}
// Clear what an interrupted earlier run may have left in the cache before
// adding to it. Remote debug jobs write to the same directory, so anything
// younger than an hour is treated as possibly still in use.
debug.RemoveStaleBundles(cacheDir, time.Hour)
deps := debug.GeneratorDependencies{
InternalConfig: cfg,
StatusRecorder: c.recorder,
TempDir: cacheDir,
StatePath: platformFiles.StateFilePath(),
}
if cc != nil {
resp, err := cc.GetLatestSyncResponse()
if err != nil {
log.Warnf("get latest sync response: %v", err)
}
deps.SyncResponse = resp
if e := cc.Engine(); e != nil {
deps.RefreshStatus = func() {
e.RunHealthProbes(context.Background(), true)
}
if cm := e.GetClientMetrics(); cm != nil {
deps.ClientMetrics = cm
}
}
}
bundleGenerator := debug.NewBundleGenerator(
deps,
debug.BundleConfig{
Anonymize: anonymize,
AnonymizeLevel: nbAnonymize.ParseLevel(anonymizeLevel),
IncludeSystemInfo: true,
},
)
path, err := bundleGenerator.Generate()
if err != nil {
return "", fmt.Errorf("generate debug bundle: %w", err)
}
if !upload {
return debug.ExportBundle(path)
}
defer func() {
if err := os.Remove(path); err != nil {
log.Errorf("failed to remove debug bundle file: %v", err)
}
}()
uploadCtx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
defer cancel()
key, err := debug.UploadDebugBundle(uploadCtx, types.DefaultBundleURL, cfg.ManagementURL.String(), path, false)
if err != nil {
return "", fmt.Errorf("upload debug bundle: %w", err)
}
log.Infof("debug bundle uploaded with key %s", key)
return key, nil
}
// SetTraceLogLevel configure the logger to trace level
func (c *Client) SetTraceLogLevel() {
log.SetLevel(log.TraceLevel)
}
// SetInfoLogLevel configure the logger to info level
func (c *Client) SetInfoLogLevel() {
log.SetLevel(log.InfoLevel)
}
// PeersList return with the list of the PeerInfos
func (c *Client) PeersList() *PeerInfoArray {
// The recorder only caches transfer counters and handshake times; nothing
// refreshes them on its own, so without this they read as zero. The desktop
// daemon does the same before serving a full peer status.
if err := c.recorder.RefreshWireGuardStats(); err != nil {
log.Debugf("failed to refresh WireGuard stats: %v", err)
}
fullStatus := c.recorder.GetFullStatus()
peerInfos := make([]PeerInfo, len(fullStatus.Peers))
for n, p := range fullStatus.Peers {
pi := PeerInfo{
IP: p.IP,
IPv6: p.IPv6,
FQDN: p.FQDN,
ConnStatus: int(p.ConnStatus),
Routes: PeerRoutes{routes: maps.Keys(p.GetRoutes())},
PubKey: p.PubKey,
Latency: formatDuration(p.Latency),
LatencyMs: p.Latency.Milliseconds(),
BytesRx: p.BytesRx,
BytesTx: p.BytesTx,
ConnStatusUpdate: formatTime(p.ConnStatusUpdate),
Relayed: p.Relayed,
RosenpassEnabled: p.RosenpassEnabled,
LastWireguardHandshake: formatTime(p.LastWireguardHandshake),
LocalIceCandidateType: p.LocalIceCandidateType,
RemoteIceCandidateType: p.RemoteIceCandidateType,
LocalIceCandidateEndpoint: p.LocalIceCandidateEndpoint,
RemoteIceCandidateEndpoint: p.RemoteIceCandidateEndpoint,
}
peerInfos[n] = pi
}
return &PeerInfoArray{items: peerInfos}
}
func (c *Client) Networks() *NetworkArray {
cc := c.getConnectClient()
if cc == nil {
log.Error("not connected")
return nil
}
engine := cc.Engine()
if engine == nil {
log.Error("could not get engine")
return nil
}
routeManager := engine.GetRouteManager()
if routeManager == nil {
log.Error("could not get route manager")
return nil
}
routeSelector := routeManager.GetRouteSelector()
if routeSelector == nil {
log.Error("could not get route selector")
return nil
}
routesMap := routeManager.GetClientRoutesWithNetID()
v6Merged := route.V6ExitMergeSet(routesMap)
resolvedDomains := c.recorder.GetResolvedDomainsStates()
activeRoutePeers := c.recorder.GetActiveRoutePeers()
networkArray := &NetworkArray{
items: make([]Network, 0),
}
for id, routes := range routesMap {
if len(routes) == 0 {
continue
}
if _, skip := v6Merged[id]; skip {
continue
}
network := c.buildNetwork(id, routes, routeSelector.IsSelected(id), resolvedDomains, v6Merged, activeRoutePeers)
if network == nil {
continue
}
networkArray.Add(*network)
}
return networkArray
}
func (c *Client) buildNetwork(id route.NetID, routes []*route.Route, selected bool, resolvedDomains map[domain.Domain]peer.ResolvedDomainInfo, v6Merged map[route.NetID]struct{}, activeRoutePeers map[route.HAUniqueID]string) *Network {
r := routes[0]
netStr := r.Network.String()
if r.IsDynamic() {
netStr = r.Domains.SafeString()
}
routePeer, err := c.findBestRoutePeer(routes, activeRoutePeers)
if err != nil {
log.Errorf("could not get peer info for route %s: %v", id, err)
return nil
}
network := &Network{
Name: string(id),
Network: netStr,
Peer: routePeer.FQDN,
Status: routePeer.ConnStatus.String(),
IsSelected: selected,
Domains: c.getNetworkDomainsFromRoute(r, resolvedDomains),
}
if route.IsV4DefaultRoute(r.Network) && route.HasV6ExitPair(id, v6Merged) {
network.Network = "0.0.0.0/0, ::/0"
}
return network
}
// findBestRoutePeer returns the peer actively routing traffic for the given
// HA route group. Falls back to the first connected peer, then the first peer.
func (c *Client) findBestRoutePeer(routes []*route.Route, activeRoutePeers map[route.HAUniqueID]string) (peer.State, error) {
if peerKey, ok := activeRoutePeers[routes[0].GetHAUniqueID()]; ok {
if p, err := c.recorder.GetPeer(peerKey); err == nil {
return p, nil
}
}
for _, r := range routes {
p, err := c.recorder.GetPeer(r.Peer)
if err != nil {
continue
}
if p.ConnStatus == peer.StatusConnected {
return p, nil
}
}
return c.recorder.GetPeer(routes[0].Peer)
}
// OnUpdatedHostDNS update the DNS servers addresses for root zones
func (c *Client) OnUpdatedHostDNS(list *DNSList) error {
dnsServer, err := dns.GetServerDns()
if err != nil {
return err
}
dnsServer.OnUpdatedHostDNSServer(slices.Clone(list.items))
return nil
}
// SetConnectionListener set the network connection listener
func (c *Client) SetConnectionListener(listener ConnectionListener) {
if listener == nil {
c.recorder.RemoveConnectionListener()
return
}
c.recorder.SetConnectionListener(connectionListenerAdapter{listener})
}
// RemoveConnectionListener remove connection listener
func (c *Client) RemoveConnectionListener() {
c.recorder.RemoveConnectionListener()
}
func (c *Client) getRouteManager() (routemanager.Manager, error) {
client := c.getConnectClient()
if client == nil {
return nil, fmt.Errorf("not connected")
}
engine := client.Engine()
if engine == nil {
return nil, fmt.Errorf("engine is not running")
}
manager := engine.GetRouteManager()
if manager == nil {
return nil, fmt.Errorf("could not get route manager")
}
return manager, nil
}
func (c *Client) SelectRoute(id string) error {
manager, err := c.getRouteManager()
if err != nil {
return err
}
return manager.SelectRoutes([]route.NetID{route.NetID(id)}, true)
}
func (c *Client) DeselectRoute(id string) error {
manager, err := c.getRouteManager()
if err != nil {
return err
}
return manager.DeselectRoutes([]route.NetID{route.NetID(id)})
}
// getNetworkDomainsFromRoute extracts domains from a route and enriches each domain
// with its resolved IP addresses from the provided resolvedDomains map.
func (c *Client) getNetworkDomainsFromRoute(route *route.Route, resolvedDomains map[domain.Domain]peer.ResolvedDomainInfo) NetworkDomains {
domains := NetworkDomains{}
for _, d := range route.Domains {
networkDomain := NetworkDomain{
Address: d.SafeString(),
}
if info, exists := resolvedDomains[d]; exists {
for _, prefix := range info.Prefixes {
networkDomain.addResolvedIP(prefix.Addr().String())
}
}
domains.Add(&networkDomain)
}
return domains
}
func exportEnvList(list *EnvList) {
if list == nil {
return
}
for k, v := range list.AllItems() {
if err := os.Setenv(k, v); err != nil {
log.Errorf("could not set env variable %s: %v", k, err)
}
}
}
// formatDuration renders a duration for display, trimming the fractional part
// to two digits so latencies read as "12.34ms" rather than "12.345678ms".
func formatDuration(d time.Duration) string {
ds := d.String()
dotIndex := strings.Index(ds, ".")
if dotIndex == -1 {
return ds
}
endIndex := min(dotIndex+3, len(ds))
// Skip the remaining digits so only the unit suffix is appended back.
unitStart := endIndex
for unitStart < len(ds) && ds[unitStart] >= '0' && ds[unitStart] <= '9' {
unitStart++
}
return ds[:endIndex] + ds[unitStart:]
}
// formatTime renders a timestamp in UTC using a fixed layout. The zero time is
// passed through as-is so the UI can recognise it and show "never" instead.
func formatTime(t time.Time) string {
return t.UTC().Format("2006-01-02 15:04:05")
}