Runbook: K8upJobStuck

A K8up job (backup, check, prune) has been queued without starting for a long time

Why is this relevant?

If a backup cannot be started, no backup will take place.


In the namespace that is mentioned in the message annotation (NOT the one mentioned in the namespace label):

  • Check for jobs that could not be started/run

  • Check backup, check or prune objects (kubectl get backup,check,prune)

The alert will persist, even if subsequent jobs succeed. If subsequent jobs of the same type were successful, delete the problematic object to resolve the monitoring alert.