Summary

Gateway enters infinite restart loop when configuration changes require restart, with zombie process holding port.

Steps to reproduce

1. Gateway running normally
2. Make config change requiring restart (e.g., gateway.controlUi.allowedOrigins)
3. Config reload triggers SIGUSR1 signal
4. New process fails to start: 'port 18789 already in use'
5. Systemd retries indefinitely, old PID still holding port

Expected behavior

Gateway cleanly stops old process before starting new one

Actual behavior

30+ restart attempts, all failing with 'port already in use'
Old PID remains listening on 18789
Only recoverable via manual pkill -9 openclaw-gateway

Environment

• OpenClaw 2026.3.2
• Ubuntu 24.04 LTS
• Port 18789 (default)

Logs

Error message: 'Gateway failed to start: another gateway instance is already listening on ws://127.0.0.1:18789'
Systemd restart counter reaches 30+

Related issues

See #28134 (Gateway restart failed - similar startup issues)