Skip to content
Latchkey

GitHub Actions concurrency: pending job canceled

A concurrency group allows one running and one pending run per group. A newer run can cancel the pending (or in-progress, with cancel-in-progress) run, which can surprise deploys that must not be canceled.

What this error means

A queued or running job is canceled when another run enters the same concurrency group, sometimes aborting a deployment mid-flight.

github-actions
Canceling since a higher priority waiting request for 'deploy-main' exists.

Common causes

cancel-in-progress cancels active runs

Setting cancel-in-progress: true cancels a running job when a new one joins the group.

Pending run superseded

A newer queued run replaces the prior pending run in the same group.

How to fix it

Scope the group and choose cancellation carefully

  1. Include a per-ref key in the group so unrelated runs do not collide.
  2. For deploys, set cancel-in-progress: false so they are not aborted.
  3. For PR CI, cancel-in-progress: true to save minutes is usually fine.
.github/workflows/deploy.yml
concurrency:
  group: deploy-${{ github.ref }}
  cancel-in-progress: false   # do not cancel an in-flight deploy

How to prevent it

  • Pick concurrency group keys that isolate independent runs.
  • Disable cancel-in-progress for jobs that must finish (deploys, releases).

Related guides

Tired of flaky CI? Latchkey auto-heals failed jobs and retries them for you. Start free →