openclaw - 💡(How to fix) Fix macOS launchd Gateway still restarts via gateway-update/update.run; document and honor auto-update kill-switch

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…

On macOS headless Gateway services managed by launchd, the Gateway can keep requesting full process restarts via the Gateway update path even when the Sparkle/UI updater is not the trigger. This makes the Gateway intermittently unavailable to control-plane health checks and plugin heartbeats.

Root Cause

On macOS headless Gateway services managed by launchd, the Gateway can keep requesting full process restarts via the Gateway update path even when the Sparkle/UI updater is not the trigger. This makes the Gateway intermittently unavailable to control-plane health checks and plugin heartbeats.

Code Example

{
  "kind": "gateway-supervisor-restart-handoff",
  "version": 1,
  "reason": "update.run",
  "source": "gateway-update",
  "restartKind": "update-process",
  "supervisorMode": "launchd"
}
RAW_BUFFERClick to expand / collapse

Summary

On macOS headless Gateway services managed by launchd, the Gateway can keep requesting full process restarts via the Gateway update path even when the Sparkle/UI updater is not the trigger. This makes the Gateway intermittently unavailable to control-plane health checks and plugin heartbeats.

Evidence

A live launchd-managed Gateway wrote this restart handoff file under the OpenClaw state directory:

{
  "kind": "gateway-supervisor-restart-handoff",
  "version": 1,
  "reason": "update.run",
  "source": "gateway-update",
  "restartKind": "update-process",
  "supervisorMode": "launchd"
}

The process then restarted under launchd. This is separate from Sparkle auto-update checks: disabling the Sparkle/defaults path reduced one trigger, but this gateway-update path still requested restarts.

Impact

For long-running/headless Gateway deployments, repeated update.run restarts cause recurring windows where HTTP/RPC health checks and plugin work become unreliable. In the same incident window, another already-filed pre-warm issue caused long event-loop stalls; this issue is specifically about the restart trigger/control path, not the provider auth pre-warm stall.

Related existing issue for the event-loop stall: #85999.

Expected behavior

Operators should have a documented, reliable way to disable Gateway package/background update restarts for supervised/headless deployments, preferably honoring OPENCLAW_NO_AUTO_UPDATE=1 in the launchd service environment before any gateway-update restart handoff is created.

Actual behavior

The launchd Gateway wrote a gateway-supervisor-restart-handoff.json with source=gateway-update, reason=update.run, and restartKind=update-process, then restarted the Gateway process.

Notes

The changelog mentions OPENCLAW_NO_AUTO_UPDATE=1 as a startup kill-switch for configured background package auto-updates, but this is hard to discover operationally and should be verified for the Gateway launchd path that writes the handoff file above.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

FAQ

Expected behavior

Operators should have a documented, reliable way to disable Gateway package/background update restarts for supervised/headless deployments, preferably honoring OPENCLAW_NO_AUTO_UPDATE=1 in the launchd service environment before any gateway-update restart handoff is created.

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING