openimsdk/open-im-server · info

signal %s

Error message

signal %s

What it means

In Start, a goroutine watches SIGTERM/SIGINT; when one arrives it cancels the root context with a cause error 'signal <name>'. This error is the cancellation cause, not an internal fault — it is how the process communicates why it is shutting down (via context.Cause).

Source

Thrown at pkg/common/startrpc/start.go:171

		grpccli.GrpcClientLogger(),
		grpccli.GrpcClientContext(),
		grpccli.GrpcClientErrorConvert(),
	)
	if len(clientOptions) > 0 {
		client.AddOption(clientOptions...)
	}

	ctx, cancel := context.WithCancelCause(ctx)

	go func() {
		sigs := make(chan os.Signal, 1)
		signal.Notify(sigs, syscall.SIGTERM, syscall.SIGINT)
		select {
		case <-ctx.Done():
			return
		case val := <-sigs:
			log.ZDebug(ctx, "recv signal", "signal", val.String())
			cancel(fmt.Errorf("signal %s", val.String()))
		}
	}()

	if prometheusListenAddr != "" {
		options = append(
			options,
			prommetricsUnaryInterceptor(rpcRegisterName),
			prommetricsStreamInterceptor(rpcRegisterName),
		)
		prometheusListener, prometheusPort, err := listenTCP(prometheusListenAddr)
		if err != nil {
			return err
		}
		log.ZDebug(ctx, "prometheus start", "addr", prometheusListener.Addr(), "rpcRegisterName", rpcRegisterName)
		target, err := jsonutil.JsonMarshal(prommetrics.BuildDefaultTarget(registerIP, prometheusPort))
		if err != nil {
			return err
		}

View on GitHub (pinned to 175a7bb067)

Solutions

1. No fix needed — this is the expected graceful shutdown path.
2. Use context.Cause(ctx) after shutdown to log the reason; treat 'signal terminated/interrupt' as normal exit.
3. If shutdown is unexpected, check what sent the signal: Kubernetes events (OOM/liveness kill shows as different exit reason), systemd, or local terminal.
4. Ensure signal handling completes cleanup before exit since ctx is cancelled immediately on signal.

err := start(ctx)
if cause := context.Cause(ctx); cause != nil {
    var sigErr *fmt.wrapError // or string-match on "signal "
    if strings.HasPrefix(cause.Error(), "signal ") {
        log.Info("graceful shutdown", "cause", cause)
        return nil
    }
}

When it happens

Trigger: The process receives SIGTERM (e.g. kubectl delete pod, systemd stop, docker stop) or SIGINT (Ctrl+C) while the RPC server is running.

Common situations: Normal graceful shutdown in Kubernetes (SIGTERM on pod termination); developer pressing Ctrl+C locally; orchestrator restarts; logs/context.Cause then surface 'signal terminated' or 'signal interrupt'.

Related errors

• serve end
• listener done
• rpc %s %w

AI-assisted analysis of openimsdk/open-im-server@175a7bb067 (2026-09-04). Data as JSON: /api/errors/0dd94e91f1b1e0b9. Report an issue: GitHub.