Skip to content

Bonjour advertiser failure cascades into iMessage channel teardown #65393

@adminfedres

Description

@adminfedres

Summary

When the Bonjour/mDNS advertiser fails (stuck in probing/announcing), the gateway tears down the iMessage channel as a side effect. These subsystems should be independent.

Evidence

  • Log pattern: [bonjour] restarting advertiser (service stuck in probing for 17-20s) immediately followed by [imessage] monitor failed: Error: imsg rpc timeout (watch.subscribe) and [imessage] [default] channel exited
  • Frequency: 33 occurrences over 4 weeks
  • Root cause of Bonjour failure: two gateways on the same LAN advertising _openclaw-gw._tcp (confirmed via dns-sd -B)

Requested changes

  1. Decouple Bonjour lifecycle from iMessage channel lifecycle
  2. Add a bonjour.enabled config key in openclaw.json (currently only OPENCLAW_DISABLE_BONJOUR env var exists, discovered in server.impl-WjqjRArz.js:9341)

Workaround

OPENCLAW_DISABLE_BONJOUR=1 environment variable disables Bonjour entirely. Gateway is accessed via loopback; discovery is not needed.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type
No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions