Open
Description
Similar as we dump the cluster state with all resources in all namespaces in dump_cluster_state (fail_test), it would be good to have all the pod/container logs too (e.g. in a tar ball) and provided as an artifact in prow.
This would help on debugging CI issues