# Background job queue backlog: first 30 minutes checklist

Source: https://releasefailures.com/incidents/background-job-queue-backlog/
Content version: 1.0.0
Verified on: 2026-10-21
Licence: CC BY 4.0

Definition: A job change slows processing and the queue grows faster than it drains.
- [ ] FM-01 Read queue depth and the age of the oldest pending task and compare them with the normal baseline.
- [ ] FM-02 Check whether recurring and scheduled work is still being enqueued, even when worker processes report healthy.
- [ ] FM-03 Find long-running instances of any task deployed in the release and cancel their queries.
- [ ] FM-04 Disable the new periodic task, or pin the upgraded library back to its previous version, whichever matches the stall.
- [ ] FM-05 Roll back the release if worker throughput stays below the previous release after the disable or the pin.
- [ ] FM-06 Scale up workers or public-facing routers by hand while the backlog drains.
- [ ] FM-07 Post a status update that names the delayed work, such as scheduled builds, emails or cache purges.
- [ ] FM-08 Keep the incident open until queue depth returns to the normal baseline.