We're having an issue with Coder leaving resources around when creation fails. Users are naturally lazy and won't try and fix this themselves.

E.g. this scenario:

kubernetes_pod.main[0]: Still creating... [4m50s elapsed]
kubernetes_pod.main[0]: Still creating... [5m0s elapsed]
kubernetes_pod.main[0]: Creation errored after 5m0s

How do we get coder to run a stop automatically when this happens so pods aren't sitting around in pending state forever plus all the rest of the supporting resources?

I'd rather not have to write our own cleanup scripts.