Kopia backup jobs should be killed if they run too long #2013
Description
Describe the bug
A current issue in k8s is causing kopia-backup jobs to stall/take forever.
As a result, the previous cron-job is still running when a new one is started.
The new one fails with Containers with incomplete status: [postgres-dump] presumably, because there's already a job running.
That's just confusing noise.
We should probably just kill a job if it takes longer than X (12?) hours.
We'll still get loads of alerts, because we're alerted if a job takes longer than 1h to complete.