[sig-node] nodes should not go unready after being upgraded and go unready only once is failing frequently in CI and the problem appears related to Azure, see: https://sippy.ci.openshift.org/sippy-ng/tests/4.10/analysis?test=%5Bsig-node%5D%20nodes%20should%20not%20go%20unready%20after%20being%20upgraded%20and%20go%20unready%20only%20once Sample prow job failure: https://prow.ci.openshift.org/view/gcs/origin-ci-test/logs/periodic-ci-openshift-release-master-ci-4.10-upgrade-from-stable-4.9-e2e-azure-upgrade/1446060099507326976
I think this is a test bug, thanks to Harshal Patil analysis ruling out a node problem. I am preparing a fix and will explain more in that PR.
deads has requested we expand on this as there are still going to be tests that will fail if we do not have the expected capacity. He suggests waiting for nodes to be ready, and possibly for machines to be ready first in a third pr. Will hopefully get this completed this week.
In https://github.com/openshift/release/pull/22928 support was added to wait for number of nodes to equal number of machines before proceeding with tests.