Gateway starts, all connections established:

2026-03-28T12:50:52.160+08:00 [INFO] feishu[default]: WebSocket client started 2026-03-28T12:50:52.169+08:00 [INFO] feishu[dagongren]: WebSocket client started 2026-03-28T12:50:52.176+08:00 [INFO] feishu[shenghuodazi]: WebSocket client started 2026-03-28T12:50:52.665+08:00 [INFO] bonjour: advertised gateway fqdn=鲜橙的Mac mini (OpenClaw)._openclaw-gw._tcp.local. host=openclaw.local. port=18789 state=announcing

4 seconds later — SIGTERM with no preceding error:

2026-03-28T12:50:56.540+08:00 [INFO] {"subsystem":"gateway"} signal SIGTERM received 2026-03-28T12:50:56.543+08:00 [INFO] {"subsystem":"gateway"} received SIGTERM; shutting down

Fix Plan

To address the restart loop issue, we'll focus on the potential causes related to the feishu WebSocket reconnection logic and the startup timeout in the restart-sentinel module. Here are the steps:

• Step 1: Modify the feishu WebSocket reconnection logic
  • Introduce a delay before considering the connections "established" to prevent premature restarts.
  • Example code (Node.js):

const ws = require('ws');

// ...

// Delay the "established" state by 10 seconds setTimeout(() => { console.log('Feishu WebSocket connections established'); // Proceed with bonjour advertisement }, 10000);

* **Step 2: Adjust the startup timeout in the restart-sentinel module**
  + Increase the timeout value to accommodate the delayed "established" state.
  + Example code (Node.js):
    ```javascript
const restartSentinel = require('restart-sentinel');

// ...

// Increase the startup timeout to 15 seconds
restartSentinel.setTimeout(15000);

• Step 3: Verify the changes
  • Restart the gateway service and monitor the logs for the restart loop.

Verification

To verify that the fix worked, check the gateway logs for the following:

• The restart loop should be broken, and the gateway should remain running without receiving SIGTERM.
• The feishu WebSocket connections should be established successfully, and the bonjour advertisement should complete without triggering a restart.

Extra Tips

• Monitor the system logs for any other potential issues that might be contributing to the restart loop.
• Consider adding additional logging or debugging statements to help identify the root cause of the issue.
• If the issue persists, try increasing the delay or timeout values to further troubleshoot the problem.