Skip to content

fix: fail stalled jobs and pause invalid credentials - #38

Merged
HugoXK merged 1 commit into
mainfrom
feat/job-health
Nov 5, 2026
Merged

HugoXK merged 1 commit into
mainfrom
feat/job-health

Conversation

@HugoXK

@HugoXK HugoXK commented Nov 5, 2026

Copy link
Copy Markdown
Owner

Written-by: HugoXK

Fails stalled jobs and pauses invalid credentials.

  • Record jobs.last_callback_at on every callback and fail a running job after JOB_STALL_MINUTES (default 30) without a heartbeat.
  • Reset the failure counter on success; pause auto-iterate and notify after three consecutive failures.
  • Mark repos.credential_state = invalid on a 401/403 preflight, pause auto-iterate and show a re-authorize banner.
  • Add schema, stall, pause and credential-preflight tests; document the new env var.

Ran:

  • npm run check — pass (125 tests, build)
  • node scripts/version.mjs check — pass
  • node scripts/changelog.mjs check — pass
    Not run: production deploy.

No-issue: reliability follow-up from the production audit.

Track the last callback per job with a configurable stall timeout, pause auto-iterate after three consecutive failures, and mark a repository invalid on a 401/403 preflight with a re-authorize banner.

@HugoXK HugoXK left a comment

Copy link
Copy Markdown
Owner Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verdict: PASS at 6f05a35
Written-by: Reviewer
Posted-by: HugoXK

Ran:

  • npm run check — pass (125 tests, build)
  • node scripts/version.mjs check — pass
  • node scripts/changelog.mjs check — pass
    Not run: production deploy.

Stall timeout is configurable; auto-pause is reversible from settings.

@HugoXK
HugoXK merged commit 84f4707 into main Nov 5, 2026
18 checks passed
@HugoXK
HugoXK deleted the feat/job-health branch November 5, 2026 06:53
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant