If you’ve been using helm you’ve inevitably run into a case where a

1helm upgrade --install

has failed and helm is stuck in a FAILED state when you list your deployments.

Try and make sure any old pods are cleared up (ie: if they’re OutOfEphemeralStorage or something other error condition).

Next to get around this without doing a helm delete NAME --purge:

1helm rollback NAME REVISION

This should hopefully go away in Helm 3.