Currently, we don't have any regular e2e coverage for AWS pod-identity and GCP workload-identity. There have been issues with new components written in a way where they don't work in these configurations. We should create informing periodics to cover these configurations.

Currently, there is a 1-to-1 mapping of a prow job to a team, so the hive team will have to receive these failures and triage/forward them to the failing component owner(s).

Discussion on adding the periodics can be seen at https://coreos.slack.com/archives/C01CQA76KMX/p1641394597251400