Files
netbird/client/embed/embed.go
Zoltan Papp ac9529ea8c [client] Fix engine lifecyrcle race (#6443)
* [client] always clean up on Engine.Start failure via defer

The rosenpass init paths (NewManager/Run) returned without calling
e.close(), leaking the WireGuard interface and other partially
initialized state on failure. Per-branch cleanup was easy to miss when
adding new early returns.

Convert Start to a named error return and tear down via a single defer
that calls e.close() whenever err != nil, removing the scattered
per-branch close() calls (including the redundant one in initFirewall).

* [client] make Engine single-use and guard against double Start

Create the run context once in NewEngine instead of in Start. This
keeps e.cancel valid for the engine's whole lifetime, so Stop can
cancel a Start that is blocked waiting on the network while holding
syncMsgMux: Stop now cancels before taking the lock, unblocking that
Start so it can release the mutex.

Reject re-entry into Start: a non-nil wgInterface means a prior Start
already ran (ErrEngineAlreadyStarted), and a cancelled run context
means the engine was stopped (ErrEngineAlreadyStopped). Both checks run
before the cleanup defer so a duplicate call cannot tear down the
running engine's state.

* [client] let engine context unblock WaitStreamConnected

WaitStreamConnected only watched the signal client's own context, which
derives from the parent engineCtx rather than the engine's run context.
A Start blocked here (signal stream not yet up) could therefore not be
released by Engine.Stop, since Stop only cancels the engine's run
context.

Pass a context into WaitStreamConnected and select on it too, and have
the engine pass e.ctx, so Stop cancelling e.ctx unblocks a parked Start.
Update the Client interface, the mock, and callers accordingly.

* [client] fix Start/Stop race by making the run loop own engine shutdown

ConnectClient.Stop stopped the engine directly while the run loop's
backoff cycle could still be starting an engine, so Engine.close raced
Engine.Start (e.g. firewall setup reading wgInterface while close nils
it). embed.Client.Start's rollback only avoided a deadlock by cancelling
before Stop; the race itself remained and was caught by -race.

Make the run loop the sole owner of engine shutdown: derive the run
context in NewConnectClient, and have Stop cancel it and wait for the
loop to exit (skipping the wait when the loop never ran) instead of
calling engine.Stop. The loop now always stops the engine on its way
out, dropping the unsynchronised wgInterface check it used to guard that
call. Self-calls from within the loop use runCancel to avoid waiting on
themselves.

embed keeps a defensive pre-Stop cancel(); the daemon's cleanupConnection
gets a TODO to adopt Stop() rather than stopping the engine in parallel.

* [client] init context state in engine tests

Engine tests built the engine context with context.WithCancel(
context.Background()), omitting CtxInitState. Now that the run context
is created in the constructor, the wgIfaceMonitor goroutine can reach
triggerClientRestart during teardown, which calls CtxGetState and
panics on the missing state. Real entry points (up, embed, service)
always CtxInitState; only the tests skipped it.

* [client] interrupt connect backoff on context cancel

The run loop retried with a raw ExponentialBackOff, so a backoff sleep
ignored context cancellation. Now that ConnectClient.Stop waits for the
run loop to exit, a cancel landing during a sleep would block Stop for
the full interval (up to MaxInterval). Wrap the backoff with the run
context so Retry returns promptly on cancel; the retry budget itself
(MaxElapsedTime) is unchanged.

* [client] bound WaitStreamConnected in signal client tests

The tests waited on WaitStreamConnected with context.Background() and the
client's own context was also Background, so a stream that never connects
would hang until the suite timeout. Pass a 5s timeout context and assert
StreamConnected afterwards so the tests fail fast with a clear reason.

* [client] fix WaitStreamConnected stale-channel race

The StreamConnected check and the wait-channel creation took the mutex
separately, so notifyStreamConnected could set the status and close/clear
connectedCh in between: the waiter then created a fresh channel nobody
would ever close and blocked forever. Also, the status read was unlocked
while notify wrote it under the mutex (a data race). Do the check and the
channel fetch in one locked section; drop the now-unused
getStreamStatusChan helper. Pre-existing bug, not introduced by this branch.

* [client] abort Start if context cancelled while waiting for signal stream

receiveSignalEvents blocks in WaitStreamConnected until the signal stream
connects or the context is cancelled. If Stop cancelled e.ctx while Start
was parked there, Start kept going: it started the remaining subsystems on
a cancelled context and marked a shutting-down engine as started. Return
the context error from receiveSignalEvents and propagate it from Start, so
the deferred cleanup runs and the cancellation reaches the caller.

* [client] clean up all started components on Start failure

Start's failure defer only called close(), which covers the wg interface,
firewall, rosenpass and port forwarding but leaves connMgr, srWatcher,
route/DNS/flow/state managers and the monitor goroutines running. A late
failure (e.g. the context-cancelled check after the signal stream) thus
leaked them.

Extract Stop's locked teardown into stopLocked (caller holds syncMsgMux,
does not wait on shutdownWg) and call it from both Stop and Start's defer.
The defer also cancels the run context first so goroutines started before
the failure unwind. Teardown order is unchanged.
2026-06-22 13:52:57 +02:00

635 lines
18 KiB
Go

package embed
import (
"context"
"errors"
"fmt"
"io"
"net"
"net/http"
"net/netip"
"os"
"sync"
"github.com/sirupsen/logrus"
wgdevice "golang.zx2c4.com/wireguard/device"
wgnetstack "golang.zx2c4.com/wireguard/tun/netstack"
"github.com/netbirdio/netbird/client/iface"
"github.com/netbirdio/netbird/client/iface/netstack"
"github.com/netbirdio/netbird/client/internal"
"github.com/netbirdio/netbird/client/internal/auth"
"github.com/netbirdio/netbird/client/internal/peer"
"github.com/netbirdio/netbird/client/internal/profilemanager"
sshcommon "github.com/netbirdio/netbird/client/ssh"
"github.com/netbirdio/netbird/client/system"
"github.com/netbirdio/netbird/shared/management/domain"
mgmProto "github.com/netbirdio/netbird/shared/management/proto"
"github.com/netbirdio/netbird/util/capture"
)
var (
ErrClientAlreadyStarted = errors.New("client already started")
ErrClientNotStarted = errors.New("client not started")
ErrEngineNotStarted = errors.New("engine not started")
ErrConfigNotInitialized = errors.New("config not initialized")
)
const (
// PeerStatusConnected indicates the peer is in connected state.
PeerStatusConnected = peer.StatusConnected
)
// PeerConnStatus is a peer's connection status.
type PeerConnStatus = peer.ConnStatus
// Client manages a netbird embedded client instance.
type Client struct {
deviceName string
config *profilemanager.Config
mu sync.Mutex
cancel context.CancelFunc
setupKey string
jwtToken string
connect *internal.ConnectClient
recorder *peer.Status
}
// Options configures a new Client.
type Options struct {
// DeviceName is this peer's name in the network
DeviceName string
// SetupKey is used for authentication
SetupKey string
// JWTToken is used for JWT-based authentication
JWTToken string
// PrivateKey is used for direct private key authentication
PrivateKey string
// ManagementURL overrides the default management server URL
ManagementURL string
// PreSharedKey is the pre-shared key for the tunnel interface
PreSharedKey string
// LogOutput is the output destination for logs (defaults to os.Stderr if nil)
LogOutput io.Writer
// LogLevel sets the logging level (defaults to info if empty)
LogLevel string
// NoUserspace disables the userspace networking mode. Needs admin/root privileges
NoUserspace bool
// ConfigPath is the path to the netbird config file. If empty, the config will be stored in memory and not persisted.
ConfigPath string
// StatePath is the path to the netbird state file
StatePath string
// DisableClientRoutes disables the client routes
DisableClientRoutes bool
// DisableIPv6 disables IPv6 overlay addressing
DisableIPv6 bool
// BlockInbound blocks all inbound connections from peers
BlockInbound bool
// BlockLANAccess blocks the embedded peer from reaching the host's
// LAN (RFC 1918, link-local, loopback) when it's used as a routing
// peer. Mirrors profilemanager.ConfigInput.BlockLANAccess. Useful
// when the embedded client must never act as a stepping stone into
// the host's local network (e.g. the proxy's overlay peer).
BlockLANAccess bool
// WireguardPort is the port for the tunnel interface. Use 0 for a random port.
WireguardPort *int
// MTU is the MTU for the tunnel interface.
// Valid values are in the range 576..8192 bytes.
// If non-nil, this value overrides any value stored in the config file.
// If nil, the existing config MTU (if non-zero) is preserved; otherwise it defaults to 1280.
// Set to a higher value (e.g. 1400) if carrying QUIC or other protocols that require larger datagrams.
MTU *uint16
// DNSLabels defines additional DNS labels configured in the peer.
DNSLabels []string
// Performance configures the tunnel's buffer pool cap and batch size.
Performance Performance
}
// Performance configures the embedded client's tunnel memory/throughput knobs.
//
// These settings are process-global: any non-nil field also becomes the
// default for Clients constructed by later embed.New calls in the same
// process. Nil fields are ignored.
type Performance struct {
// PreallocatedBuffersPerPool caps the per-tunnel buffer pool. Zero
// leaves the pool unbounded. Lower values trade throughput for a
// tighter memory ceiling. May also be changed on a running Client via
// Client.SetPerformance, provided this field was nonzero at construction.
PreallocatedBuffersPerPool *uint32
// MaxBatchSize overrides the number of packets the tunnel reads or
// writes per syscall, which also bounds eager buffer allocation per
// worker. Zero uses the platform default. Applied at construction
// only; ignored by Client.SetPerformance.
MaxBatchSize *uint32
}
// validateCredentials checks that exactly one credential type is provided
func (opts *Options) validateCredentials() error {
credentialsProvided := 0
if opts.SetupKey != "" {
credentialsProvided++
}
if opts.JWTToken != "" {
credentialsProvided++
}
if opts.PrivateKey != "" {
credentialsProvided++
}
if credentialsProvided == 0 {
return fmt.Errorf("one of SetupKey, JWTToken, or PrivateKey must be provided")
}
if credentialsProvided > 1 {
return fmt.Errorf("only one of SetupKey, JWTToken, or PrivateKey can be specified")
}
return nil
}
// New creates a new netbird embedded client.
func New(opts Options) (*Client, error) {
if err := opts.validateCredentials(); err != nil {
return nil, err
}
if opts.MTU != nil {
if err := iface.ValidateMTU(*opts.MTU); err != nil {
return nil, fmt.Errorf("invalid MTU: %w", err)
}
}
if opts.LogOutput != nil {
logrus.SetOutput(opts.LogOutput)
}
if opts.LogLevel != "" {
level, err := logrus.ParseLevel(opts.LogLevel)
if err != nil {
return nil, fmt.Errorf("parse log level: %w", err)
}
logrus.SetLevel(level)
}
if !opts.NoUserspace {
if err := os.Setenv(netstack.EnvUseNetstackMode, "true"); err != nil {
return nil, fmt.Errorf("setenv: %w", err)
}
if err := os.Setenv(netstack.EnvSkipProxy, "true"); err != nil {
return nil, fmt.Errorf("setenv: %w", err)
}
}
if opts.StatePath != "" {
// TODO: Disable state if path not provided
if err := os.Setenv("NB_DNS_STATE_FILE", opts.StatePath); err != nil {
return nil, fmt.Errorf("setenv: %w", err)
}
}
var err error
var parsedLabels domain.List
if parsedLabels, err = domain.FromStringList(opts.DNSLabels); err != nil {
return nil, fmt.Errorf("invalid dns labels: %w", err)
}
t := true
var config *profilemanager.Config
input := profilemanager.ConfigInput{
ConfigPath: opts.ConfigPath,
ManagementURL: opts.ManagementURL,
PreSharedKey: &opts.PreSharedKey,
DisableServerRoutes: &t,
DisableClientRoutes: &opts.DisableClientRoutes,
DisableIPv6: &opts.DisableIPv6,
BlockInbound: &opts.BlockInbound,
BlockLANAccess: &opts.BlockLANAccess,
WireguardPort: opts.WireguardPort,
MTU: opts.MTU,
DNSLabels: parsedLabels,
}
if opts.ConfigPath != "" {
config, err = profilemanager.UpdateOrCreateConfig(input)
} else {
config, err = profilemanager.CreateInMemoryConfig(input)
}
if err != nil {
return nil, fmt.Errorf("create config: %w", err)
}
if opts.PrivateKey != "" {
config.PrivateKey = opts.PrivateKey
}
if opts.Performance.PreallocatedBuffersPerPool != nil {
wgdevice.SetPreallocatedBuffersPerPool(*opts.Performance.PreallocatedBuffersPerPool)
}
if opts.Performance.MaxBatchSize != nil {
wgdevice.SetMaxBatchSizeOverride(*opts.Performance.MaxBatchSize)
}
return &Client{
deviceName: opts.DeviceName,
setupKey: opts.SetupKey,
jwtToken: opts.JWTToken,
config: config,
recorder: peer.NewRecorder(config.ManagementURL.String()),
}, nil
}
// Start begins client operation and blocks until the engine has been started successfully or a startup error occurs.
// Pass a context with a deadline to limit the time spent waiting for the engine to start.
func (c *Client) Start(startCtx context.Context) error {
c.mu.Lock()
defer c.mu.Unlock()
if c.connect != nil {
return ErrClientAlreadyStarted
}
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)
authClient, err := auth.NewAuth(ctx, c.config.PrivateKey, c.config.ManagementURL, c.config)
if err != nil {
return fmt.Errorf("create auth client: %w", err)
}
defer authClient.Close()
if err, _ := authClient.Login(ctx, c.setupKey, c.jwtToken); err != nil {
return fmt.Errorf("login: %w", err)
}
client := internal.NewConnectClient(ctx, c.config, c.recorder)
client.SetSyncResponsePersistence(true)
// either startup error (permanent backoff err) or nil err (successful engine up)
// TODO: make after-startup backoff err available
run := make(chan struct{})
clientErr := make(chan error, 1)
go func() {
if err := client.Run(run, ""); err != nil {
clientErr <- err
}
}()
select {
case <-startCtx.Done():
// ConnectClient.Stop now cancels its own run context and waits for the
// run loop to tear the engine down, so this cancel() is no longer
// required to break the deadlock and could be removed. It is kept as a
// defensive belt-and-suspenders: cancelling the parent context first
// guarantees the run loop is unblocked even if Stop's contract regresses.
cancel()
if stopErr := client.Stop(); stopErr != nil {
return fmt.Errorf("stop error after context done. Stop error: %w. Context done: %w", stopErr, startCtx.Err())
}
return startCtx.Err()
case err := <-clientErr:
return fmt.Errorf("startup: %w", err)
case <-run:
}
c.connect = client
c.cancel = cancel
return nil
}
// Stop gracefully stops the client.
// Pass a context with a deadline to limit the time spent waiting for the engine to stop.
func (c *Client) Stop(ctx context.Context) error {
c.mu.Lock()
defer c.mu.Unlock()
if c.connect == nil {
return ErrClientNotStarted
}
if c.cancel != nil {
c.cancel()
c.cancel = nil
}
done := make(chan error, 1)
connect := c.connect
go func() {
done <- connect.Stop()
}()
select {
case <-ctx.Done():
c.connect = nil
return ctx.Err()
case err := <-done:
c.connect = nil
if err != nil {
return fmt.Errorf("stop: %w", err)
}
return nil
}
}
// GetConfig returns a copy of the internal client config.
func (c *Client) GetConfig() (profilemanager.Config, error) {
c.mu.Lock()
defer c.mu.Unlock()
if c.config == nil {
return profilemanager.Config{}, ErrConfigNotInitialized
}
return *c.config, nil
}
// Dial dials a network address in the netbird network.
// Not applicable if the userspace networking mode is disabled.
func (c *Client) Dial(ctx context.Context, network, address string) (net.Conn, error) {
engine, err := c.getEngine()
if err != nil {
return nil, err
}
nsnet, err := engine.GetNet()
if err != nil {
return nil, fmt.Errorf("get net: %w", err)
}
return nsnet.DialContext(ctx, network, address)
}
// DialContext dials a network address in the netbird network with context
func (c *Client) DialContext(ctx context.Context, network, address string) (net.Conn, error) {
return c.Dial(ctx, network, address)
}
// ListenTCP listens on the given address in the netbird network.
// Not applicable if the userspace networking mode is disabled.
func (c *Client) ListenTCP(address string) (net.Listener, error) {
nsnet, addr, err := c.getNet()
if err != nil {
return nil, err
}
_, port, err := net.SplitHostPort(address)
if err != nil {
return nil, fmt.Errorf("split host port: %w", err)
}
listenAddr := net.JoinHostPort(addr.String(), port)
tcpAddr, err := net.ResolveTCPAddr("tcp", listenAddr)
if err != nil {
return nil, fmt.Errorf("resolve: %w", err)
}
return nsnet.ListenTCP(tcpAddr)
}
// ListenUDP listens on the given address in the netbird network.
// Not applicable if the userspace networking mode is disabled.
func (c *Client) ListenUDP(address string) (net.PacketConn, error) {
nsnet, addr, err := c.getNet()
if err != nil {
return nil, err
}
_, port, err := net.SplitHostPort(address)
if err != nil {
return nil, fmt.Errorf("split host port: %w", err)
}
listenAddr := net.JoinHostPort(addr.String(), port)
udpAddr, err := net.ResolveUDPAddr("udp", listenAddr)
if err != nil {
return nil, fmt.Errorf("resolve: %w", err)
}
return nsnet.ListenUDP(udpAddr)
}
// NewHTTPClient returns a configured http.Client that uses the netbird network for requests.
// Not applicable if the userspace networking mode is disabled.
func (c *Client) NewHTTPClient() *http.Client {
transport := &http.Transport{
DialContext: c.Dial,
}
return &http.Client{
Transport: transport,
}
}
// Expose exposes a local service via the NetBird reverse proxy, making it accessible through a public URL.
// It returns an ExposeSession. Call Wait on the session to keep it alive.
func (c *Client) Expose(ctx context.Context, req ExposeRequest) (*ExposeSession, error) {
engine, err := c.getEngine()
if err != nil {
return nil, err
}
mgr := engine.GetExposeManager()
if mgr == nil {
return nil, fmt.Errorf("expose manager not available")
}
resp, err := mgr.Expose(ctx, req)
if err != nil {
return nil, fmt.Errorf("expose: %w", err)
}
return &ExposeSession{
Domain: resp.Domain,
ServiceName: resp.ServiceName,
ServiceURL: resp.ServiceURL,
mgr: mgr,
}, nil
}
// IdentityForIP looks up a remote peer by its tunnel IP using the
// embedded client's status recorder. Returns the peer's WireGuard public
// key and FQDN. ok=false means the IP doesn't belong to an active peer
// — offline roster peers are treated as unknown, same as foreign IPs.
func (c *Client) IdentityForIP(ip netip.Addr) (pubKey, fqdn string, ok bool) {
if !ip.IsValid() || c.recorder == nil {
return "", "", false
}
state, found := c.recorder.PeerStateByIP(ip.String())
if !found {
return "", "", false
}
return state.PubKey, state.FQDN, true
}
// Status returns the current status of the client.
func (c *Client) Status() (peer.FullStatus, error) {
c.mu.Lock()
connect := c.connect
c.mu.Unlock()
if connect != nil {
engine := connect.Engine()
if engine != nil {
_ = engine.RunHealthProbes(false)
}
}
return c.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
}
// 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()
if connect != nil {
connect.SetLogLevel(level)
}
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.
func (c *Client) VerifySSHHostKey(peerAddress string, key []byte) error {
engine, err := c.getEngine()
if err != nil {
return err
}
storedKey, found := engine.GetPeerSSHKey(peerAddress)
if !found {
return sshcommon.ErrPeerNotFound
}
return sshcommon.VerifyHostKey(storedKey, key, peerAddress)
}
// SetPerformance retunes a running Client. Only PreallocatedBuffersPerPool
// takes effect, and only when it was nonzero at construction;
// MaxBatchSize is construction-only and returns an error if set here.
//
// Returns ErrClientNotStarted / ErrEngineNotStarted if the Client is not
// running yet.
func (c *Client) SetPerformance(t Performance) error {
if t.MaxBatchSize != nil {
return errors.New("MaxBatchSize is construction-only and cannot be changed at runtime")
}
engine, err := c.getEngine()
if err != nil {
return err
}
return engine.SetPerformance(internal.Performance{
PreallocatedBuffersPerPool: t.PreallocatedBuffersPerPool,
})
}
// StartCapture begins capturing packets on this client's tunnel device.
// Only one capture can be active at a time; starting a new one stops the previous.
// Call StopCapture (or CaptureSession.Stop) to end it.
func (c *Client) StartCapture(opts CaptureOptions) (*CaptureSession, error) {
engine, err := c.getEngine()
if err != nil {
return nil, err
}
var matcher capture.Matcher
if opts.Filter != "" {
m, err := capture.ParseFilter(opts.Filter)
if err != nil {
return nil, fmt.Errorf("parse filter: %w", err)
}
matcher = m
}
sess, err := capture.NewSession(capture.Options{
Output: opts.Output,
TextOutput: opts.TextOutput,
Matcher: matcher,
Verbose: opts.Verbose,
ASCII: opts.ASCII,
})
if err != nil {
return nil, fmt.Errorf("create capture session: %w", err)
}
if err := engine.SetCapture(sess); err != nil {
sess.Stop()
return nil, fmt.Errorf("set capture: %w", err)
}
return &CaptureSession{sess: sess, engine: engine}, nil
}
// StopCapture stops the active capture session if one is running.
func (c *Client) StopCapture() error {
engine, err := c.getEngine()
if err != nil {
return err
}
return engine.SetCapture(nil)
}
// getEngine safely retrieves the engine from the client with proper locking.
// Returns ErrClientNotStarted if the client is not started.
// Returns ErrEngineNotStarted if the engine is not available.
func (c *Client) getEngine() (*internal.Engine, error) {
c.mu.Lock()
connect := c.connect
c.mu.Unlock()
if connect == nil {
return nil, ErrClientNotStarted
}
engine := connect.Engine()
if engine == nil {
return nil, ErrEngineNotStarted
}
return engine, nil
}
func (c *Client) getNet() (*wgnetstack.Net, netip.Addr, error) {
engine, err := c.getEngine()
if err != nil {
return nil, netip.Addr{}, err
}
addr, err := engine.Address()
if err != nil {
return nil, netip.Addr{}, fmt.Errorf("engine address: %w", err)
}
nsnet, err := engine.GetNet()
if err != nil {
return nil, netip.Addr{}, fmt.Errorf("get net: %w", err)
}
return nsnet, addr, nil
}