bump: k8s to 1.23

[damemi]

In preparation for the 1.23 release, bumping to the 1.23 release candidate for soak time

ref #641

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: damemi

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [damemi]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[damemi]

/retest

[damemi]

/cc @ingvagabund
helm test is failing, but it's not clear to me why. Maybe @stevehipwell could look too?

see https://prow.k8s.io/view/gs/kubernetes-jenkins/pr-logs/pull/kubernetes-sigs_descheduler/667/pull-descheduler-helm-test/1467877755264700416

[stevehipwell]

@damemi haven't PSPs been removed in K8s v1.23?

[seanmalloy]

@damemi haven't PSPs been removed in K8s v1.23?

If I recall correctly they are still deprecated. They will be removed in v1.25.

[seanmalloy]

@damemi haven't PSPs been removed in K8s v1.23?

If I recall correctly they are still deprecated. They will be removed in v1.25.

https://kubernetes.io/blog/2021/04/06/podsecuritypolicy-deprecation-past-present-and-future/

[stevehipwell]

Thanks @seanmalloy, that would have been too easy.

@damemi are there any API version overrides in the values being used? Maybe for the cronjob? I'm not in front of my computer, so just guessing but might be able to take a proper look tomorrow.

[damemi]

@stevehipwell oh yeah! probably #577

[damemi]

Hm, actually looks like we should still be good with v1beta1 until 1.25 https://kubernetes.io/docs/reference/using-api/deprecation-guide/#cronjob-v125
/retest
maybe it's just a flake?

[damemi]

Now that 1.23 has released, I'm just updating this to use the GA tag. Still need to sort out why Helm test is failing, we have some ideas above.

In the meantime, we also need to make the regular updates to test-infra to run against k8s 1.23 and add a test for the release-1.23 branch. If there are any volunteers to make that change please feel free!

[damemi]

Ran this locally and saw this error in the pod logs:

$ kubectl logs pod/descheduler-ci-27319252-5lpdg -n kube-system
Error: unknown flag: --v

which looks like we need the change to initialize klog, similar to this. I'll add that and report back

it's also weird that helm-test is the only one that caught this...

[damemi]

Green now, so this should be ready to merge

[ingvagabund]

Running locally with -v=4 I don't see the descheduler logs

[damemi]

Running locally with -v=4 I don't see the descheduler logs

well that's not good...
/hold

[ingvagabund]

Klog seems to work, it's just the --v option that's not properly wired to it probably

[ingvagabund]

Because kubernetes/kubernetes#106090
See change in kubernetes/kubernetes@3948cb8#diff-465167b08358906be13f9641d4798c6e8ad0790395e045af8ace4d08223fa922R78

[ingvagabund]

/lgtm
/hold cancel

[damemi]

Thanks again @ingvagabund for solving that problem!