mirror of
https://github.com/netbirdio/netbird.git
synced 2026-08-25 00:51:28 +02:00
072fa8143b
On a network switch (e.g. cellular to WiFi) the management, signal and relay sockets stay bound to the old network and look alive until the OS tears them down — measured at 5 seconds of dead air on Android, while the UI kept claiming Connected. The Android client papered over this with a full engine restart, paying for it with a torn-down TUN device and discarded peer state. Introduce client/netsweep: connections register on dial and deregister on close, and a sweep closes everything registered while aborting in-flight dials through sweep-cancellable dial contexts. The aborted dials matter: a relay dial started on the dying network would otherwise hold the reconnect loop hostage for the QUIC handshake timeout. After a sweep every failure surfaces as an ordinary read/write error and the existing retry loops redial immediately on the new network. The sweeper reaches the three long-lived connections through the same options that carry the netstate gate: a gRPC dial option wraps the management and signal transports (reconnects included), and the relay client wraps its connection in one place for the picker, the guard and foreign relays alike. Everything is nil-safe; platforms that inject no sweeper are untouched. Mobile clients expose the sweep as NotifyNetworkChange. Measured on Android against the engine restart it replaces: recovery in 1.6s instead of 3.2s, no Disconnected flash, and the TUN device, WireGuard config and peer state survive.
123 lines
3.2 KiB
Go
123 lines
3.2 KiB
Go
package client
|
|
|
|
import (
|
|
"context"
|
|
"errors"
|
|
"fmt"
|
|
"sync/atomic"
|
|
"time"
|
|
|
|
log "github.com/sirupsen/logrus"
|
|
|
|
"github.com/netbirdio/netbird/client/netsweep"
|
|
auth "github.com/netbirdio/netbird/shared/relay/auth/hmac"
|
|
)
|
|
|
|
const (
|
|
maxConcurrentServers = 7
|
|
defaultConnectionTimeout = 30 * time.Second
|
|
)
|
|
|
|
type connResult struct {
|
|
RelayClient *Client
|
|
Url string
|
|
Err error
|
|
}
|
|
|
|
type ServerPicker struct {
|
|
TokenStore *auth.TokenStore
|
|
ServerURLs atomic.Value
|
|
PeerID string
|
|
MTU uint16
|
|
ConnectionTimeout time.Duration
|
|
TransportFallback *transportFallback
|
|
Sweeper *netsweep.Sweeper
|
|
}
|
|
|
|
func (sp *ServerPicker) PickServer(parentCtx context.Context) (*Client, error) {
|
|
ctx, cancel := context.WithTimeout(parentCtx, sp.ConnectionTimeout)
|
|
defer cancel()
|
|
|
|
totalServers := len(sp.ServerURLs.Load().([]string))
|
|
|
|
connResultChan := make(chan connResult, totalServers)
|
|
successChan := make(chan connResult, 1)
|
|
errChan := make(chan error, 1)
|
|
concurrentLimiter := make(chan struct{}, maxConcurrentServers)
|
|
|
|
log.Debugf("pick server from list: %v", sp.ServerURLs.Load().([]string))
|
|
for _, url := range sp.ServerURLs.Load().([]string) {
|
|
// todo check if we have a successful connection so we do not need to connect to other servers
|
|
concurrentLimiter <- struct{}{}
|
|
go func(url string) {
|
|
defer func() {
|
|
<-concurrentLimiter
|
|
}()
|
|
sp.startConnection(parentCtx, connResultChan, url)
|
|
}(url)
|
|
}
|
|
|
|
go sp.processConnResults(connResultChan, successChan, errChan)
|
|
|
|
select {
|
|
case cr, ok := <-successChan:
|
|
if !ok {
|
|
return nil, <-errChan
|
|
}
|
|
log.Infof("chosen home Relay server: %s", cr.Url)
|
|
return cr.RelayClient, nil
|
|
case <-ctx.Done():
|
|
return nil, fmt.Errorf("connect to relay server: %w", ctx.Err())
|
|
}
|
|
}
|
|
|
|
func (sp *ServerPicker) startConnection(ctx context.Context, resultChan chan connResult, url string) {
|
|
log.Infof("try to connecting to relay server: %s", url)
|
|
relayClient := NewClient(url, sp.TokenStore, sp.PeerID, sp.MTU)
|
|
relayClient.SetTransportFallback(sp.TransportFallback)
|
|
relayClient.sweeper = sp.Sweeper
|
|
err := relayClient.Connect(ctx)
|
|
resultChan <- connResult{
|
|
RelayClient: relayClient,
|
|
Url: url,
|
|
Err: err,
|
|
}
|
|
}
|
|
|
|
func (sp *ServerPicker) processConnResults(resultChan chan connResult, successChan chan connResult, errChan chan error) {
|
|
var hasSuccess bool
|
|
var errs []error
|
|
for numOfResults := 0; numOfResults < cap(resultChan); numOfResults++ {
|
|
cr := <-resultChan
|
|
if cr.Err != nil {
|
|
log.Tracef("failed to connect to Relay server: %s: %v", cr.Url, cr.Err)
|
|
errs = append(errs, cr.Err)
|
|
continue
|
|
}
|
|
log.Infof("connected to Relay server: %s", cr.Url)
|
|
|
|
if hasSuccess {
|
|
log.Infof("closing unnecessary Relay connection to: %s", cr.Url)
|
|
if err := cr.RelayClient.Close(); err != nil {
|
|
log.Errorf("failed to close connection to %s: %v", cr.Url, err)
|
|
}
|
|
continue
|
|
}
|
|
|
|
hasSuccess = true
|
|
successChan <- cr
|
|
}
|
|
if !hasSuccess {
|
|
errChan <- pickErr(errs)
|
|
}
|
|
close(successChan)
|
|
}
|
|
|
|
// pickErr combines per-server connection failures into a single error.
|
|
func pickErr(errs []error) error {
|
|
if len(errs) == 0 {
|
|
return errors.New("no relay server available")
|
|
}
|
|
return errors.Join(errs...)
|
|
}
|