Major
Resolved
Worker pool outage
02:25 UTC
+0 min
Investigating
Background Jobs is unavailable. Requests are queued but not processed.
03:11 UTC
+46 min
Identified
The worker pool failed to restart after a scheduled node upgrade. We are rolling the upgrade back.
04:05 UTC
+1 h 40 min
Monitoring
Workers are back online and processing the queue.
04:59 UTC
+2 h 34 min
Resolved
Background Jobs has fully recovered and the queue is empty.
Root cause
A scheduled node upgrade changed the container runtime version, which was incompatible with the worker pool's startup probe. Upgrades now run against a staging pool first.