openclaw - 💡(How to fix) Fix Bonjour plugin crash-loops gateway on headless VPS (v2026.4.24) [1 comments, 2 participants]

[plugins] bonjour: watchdog detected non-announced service; attempting re-advertise
  (gateway fqdn=openclaw._openclaw-gw._tcp.local. host=openclaw.local. port=18789 state=probing)
[plugins] bonjour: restarting advertiser
  (service stuck in probing for 21303ms)
[openclaw] Unhandled promise rejection: CIAO PROBING CANCELLED

systemd restarts the gateway, which crash-loops every ~30 seconds. Stability bundles written to ~/.openclaw/logs/stability/.

TL;DR

Disable the bonjour plugin in openclaw.json to prevent crash-loops on headless VPS environments.

Guidance

• The bonjour plugin's probing mechanism is not designed to handle headless environments without mDNS responders, leading to unhandled promise rejections and gateway crashes.
• To verify the issue, check the logs for CIAO PROBING CANCELLED errors and observe the gateway's crash-loop behavior.
• As a workaround, disable the bonjour plugin by setting "enabled": false in the openclaw.json configuration file, which will prevent the probing mechanism from causing crashes.
• Consider reporting this issue to the @homebridge/ciao plugin maintainers to request a more robust handling of headless environments.

Example

{
  "plugins": {
    "entries": {
      "bonjour": { "enabled": false }
    }
  }
}

This configuration change will disable the bonjour plugin and prevent the gateway from crash-looping.

Notes

This workaround may not be suitable for environments where bonjour discovery is necessary. The root cause of the issue lies in the bonjour plugin's handling of headless environments, and a more permanent fix would require changes to the plugin itself.

Recommendation

Apply the workaround by disabling the bonjour plugin, as it is a simple and effective solution to prevent crash-loops on headless VPS environments.