Failed Kubernetes Job alerts
This feature is part of kwatch Pro, which is free during early access. See Community vs Pro.
A Job runs work until it succeeds or reaches its failure condition. kwatch watches Jobs that fail or become suspended, helping the team distinguish a workload error from an intentional pause.
What kwatch detects
kwatch reports Jobs that fail or become suspended. There is no on/off switch for this detector: kwatch always watches every supported resource. To silence a known, intentional case use silences or a maintenance annotation (see General configuration).
An intentional suspension may not require action. Check who changed the Job and whether the pause is part of a planned operation before responding.
Investigate a failed Job
kubectl describe job <name> -n <namespace>
kubectl get pods -n <namespace>
Look at Job conditions and the Pods it created. Inspect the failing Pod's Events and logs to understand whether the problem is in the task, its image, configuration, or a cluster dependency. If the Job is suspended, confirm the expected schedule and owner.
Related
For recurring work, see the CronJob monitor.