@ialidzhikov: This cherry pick PR is for a release branch and has not yet been approved by Release Managers.
Adding the do-not-merge/cherry-pick-not-approved label.

To merge this cherry pick, it must first be approved (/lgtm + /approve) by the relevant OWNERS.

AFTER it has been approved by code owners, please ping the kubernetes/release-managers team in a comment to request a cherry pick review.

(For details on the patch release process and schedule, see the Patch Releases page.)

@ialidzhikov: This issue is currently awaiting triage.

If a SIG or subproject determines this is a relevant issue, they will accept it by applying the triage/accepted label and provide further guidance.

The triage/accepted label can be added by org members by writing /triage accepted in a comment.

Hi @ialidzhikov. Thanks for your PR.

I'm waiting for a kubernetes member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

/area storage
/kind bug

This PR is a fix for a regression that was introduced in 1.18.16 with #98224.
/kind regression

/cc @Jiawei0227 @msau42

PS. I really know that 1.18 is out of support and there will be an exceptional last patch release for 1.18 (https://groups.google.com/g/kubernetes-dev/c/KuF8s2zueFs/m/GWGAACgKAgAJ). But I think the bug fix #101737 deserves a cherry-pick.

@ialidzhikov: The label(s) area/storage cannot be applied, because the repository doesn't have them.

/sig storage

/cc @puerco

Hi @ialidzhikov I'm sorry, 1.18 is now EOL so we cannot approve the cherry-pick anymore :(

Hi @ialidzhikov I'm sorry, 1.18 is now EOL so we cannot approve the cherry-pick anymore :(

According to https://groups.google.com/g/kubernetes-dev/c/KuF8s2zueFs/m/GWGAACgKAgAJ it seems that there will be one final patch release. I know that 1.18 is already out of support but I believe that this fix is critical enough to be included as well. The regression itself was introduced in 1.18.16. @Jiawei0227 @msau42 I believe you would agree with me.

/ok-to-test

There is another PR that is related to this which is: https://github.com/kubernetes/kubernetes/pull/102176/files

I believe if you really want to backport it in 1.18 we should also cherrypick this.

I synced with the PR author @timebertt few days back on this topic. We agreed that the change is not that critical because it only fixes the garbage collection logic for VolumeAttachments to consider CSI migrated volumes as well. It is nice to have of course but I don't want to bundle it into this PR.
On the other side #101737 fixes an issue where kube-controller-manager can potentially wrongly detach CSI migrated volumes. This issue is already reported in kubernetes/cloud-provider-openstack#1531, kubernetes-sigs/aws-ebs-csi-driver#896 and we see it on our landscapes as well.

/test pull-kubernetes-node-e2e

All checks are passing now. @Jiawei0227 @msau42 @gnufied can you check again?

PS: I guess the final patch release for 1.18 is scheduled for 2021-06-16.

/lgtm

/approve

This fixes a regression that causes disks using CSI migration to be prematurely detached from nodes while pods are still running, causing IO errors to workloads if kube-controller-manger restarts (like on an upgrade). I think this is a serious enough bug to be considered for the final patch release.

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ialidzhikov, msau42

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

The pull request process is described here

/priority critical-urgent

/cc @kubernetes/release-managers

/close

@ialidzhikov: Closed this PR.

Slack thread: https://kubernetes.slack.com/archives/CJH2GBF7Y/p1623453145304800