A number of improvements have been made to the upgrade tests in 4.5 that must be back ported to earlier releases: * Allow multiple sequential upgrades to be performed as a unit for 4.1-4.2-4.3-4.4 upgrade testing (https://github.com/openshift/origin/pull/24570) * Report router downtime (via the console and oauth routes) during upgrade like we do for the core platform and service load balancers (https://github.com/openshift/origin/pull/24580) * Ensure AWS service LB availability test is tuned consistently with GCP and Azure (https://github.com/openshift/origin/pull/24584) * Minor interspersed fixes to naming (https://github.com/openshift/origin/pull/24557) Backport all of these to 4.3 as one unit to avoid future cherry-picking issues. We need to go back to 4.3 to verify upgrade downtime as fixes are rolled out.
All fixes have been delivered to 4.5 except a full fix for the AWS SLB failing after upgrade starts (due to passing, then failing) which I am currently working on fixing. This does not block back porting these changes to 4.4 now.
https://prow.svc.ci.openshift.org/view/gcs/origin-ci-test/logs/release-openshift-origin-installer-e2e-aws-upgrade/21768 All green. A few remaining fixes are in progress.
Removing from the Test Infrastructure component since it's ... not :)
Checked this issue already in the CI job. so move this bug to 'verified'
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.5 image release advisory), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:2409