You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I initially thought this wasn't necessary since we tear down the clusters after each e2e test run. However, the plan for presubmits is to not recycle the clusters after each test run to avoid cluster bring up and tear down times. If we retain the clusters we have to clean up the kube-dns ConfigMap entries we create in the clusters to have a clean start at the beginning of each run. kubefed unjoin is learning this functionality (PR #39338) and that's the easiest way to cleanup kube-dns entries. So we need to run kubefed unjoin in each cluster in federation-down.sh
I initially thought this wasn't necessary since we tear down the clusters after each e2e test run. However, the plan for presubmits is to not recycle the clusters after each test run to avoid cluster bring up and tear down times. If we retain the clusters we have to clean up the
kube-dns
ConfigMap entries we create in the clusters to have a clean start at the beginning of each run.kubefed unjoin
is learning this functionality (PR #39338) and that's the easiest way to cleanupkube-dns
entries. So we need to runkubefed unjoin
in each cluster infederation-down.sh
cc @kubernetes/sig-federation-bugs @shashidharatd
The text was updated successfully, but these errors were encountered: