Automatically clean up Pods once they've finished running

[aditya-konarde]

The jobs leave behind failed pods. We need to make sure that there's some garbage collection happening. Ideally this needs to be done from within the service.

[aditya-konarde]

For example, found 13 pods just hanging there in error state on staging:

[akonarde@katana ~]$ oc get pods -n bayesian-preview | grep Error | grep cve | wc -l
13

[aditya-konarde]

@msrb ^