KubeVirt CI Workloads Cluster Maintenance Window - 9th Jan. 12pm -> 10th Jan. 12pm (UTC)

[doll...@redhat.com]

Happy new year,

For New Year's resolutions, the KubeVirt CI maintenance team would like
to perform several (probably three) disruptive improvements to our
aging Prow instance during this quarter.

The first one will be to update the OpenShift version of the CI
workloads cluster which is reaching its end of life in April.

For this purpose, a maintenance window is scheduled to start this
Friday (9th Jan.) early in the afternoon (EMEA timezone) and to finish
during the week-end.

The update will be quite long since it will go through several steps
before catching up the latest LTS version:

OCP-v4.17 (k8s-v1.30) -> OCP-v4.18 (k8s-v1.31) -> OCP-v4.19 (k8s-
v1.32) -> OCP-v4.20 (k8s-v1.33)

The CI workloads cluster will still be available during the update but
it will experience a few hiccups.
It is expected for some ProwJobs to fail when the nodes will be evicted
to reboot in order to update the operating system.
You might need to use `/retest` more than usual.

There are countless improvements and bug fixes between OpenShift 4.17
and 4.20 but the one which might have the most impact is the switch
from `runc` to `crun`.

Unless something unexpected happens this maintenance should be quite
smooth overall.

Don't hesitate to reach us if you have any concerns.

Regards,

The KubeVirt CI maintenance team.

[Denis Ollier Pinas]

Heads up,

The maintenance is about to start.

Regards,

The KubeVirt CI maintenance team.

[Denis Ollier Pinas]

The maintenance of the workloads cluster is now complete.

It has been updated to the latest EUS version: OCP-v4.20.8.

The next maintenance will be to update the control plane cluster from OCP-v4.16 to the latest supported version.

Regards,

The KubeVirt CI maintenance team.