Closed mboersma closed 3 years ago
@chewong, @CecileRobertMichon, @jackfrancis what is the purpose of running the aks-engine vmas disk test if we are also running the CAPZ test? Do we need this test?
does capz use availability sets for vms? (is the availability set aspect of this test important?)
we actually already disabled the periodic job for azure file + aks-engine, in favor of the azure file + capz one (https://testgrid.k8s.io/provider-azure-1.21-signal#capz-azure-file).
This is the presubmit job, for which there is no capz equivalent yet.
Let's convert all aks-engine-based PR jobs to CAPZ-based now that we have decent test signal from the periodic jobs. What do you guys think?
I'm on it.
removed aks-engine job
should we do this for other jobs as well? do we need a separate issue to track it?
Unfortunately, the only details included here are below. We do not have any logs from the actual VM, so we'll need to try to reproduce the failure and ssh into the machine to determine what failed during provisioning.
https://prow.k8s.io/view/gs/kubernetes-jenkins/pr-logs/directory/pull-kubernetes-e2e-aks-engine-azure-disk-vmas/1393436578520502272 https://prow.k8s.io/view/gs/kubernetes-jenkins/pr-logs/directory/pull-kubernetes-e2e-aks-engine-azure-disk-vmas/1394218517963739136