I was under the impression that the IBM tests were a subset of the jenkins e2e tests yet the ibm test found a bug in #481 while e2e nor prow did

ps: initially thought it's a k8s version difference but it apparently is not