Fix subpath tests not to fail in namespace deletion

[mkimuram]

What this PR does / why we need it:
This PR fixes below subpath test not to fail in namespace deletion

• subPath should support restarting containers using directory as subpath
• subPath should support restarting containers using file as subpath

Which issue(s) this PR fixes (optional, in fixes #<issue number>(, fixes #<issue_number>, ...) format, will close the issue(s) when PR gets merged):
Fixes: #68073

Special notes for your reviewer:
/sig storage
/sig testing

Release note:

NONE

[mkimuram]

@msau42

PTAL

[davidz627]

/ok-to-test
/assign

[davidz627]

/assign @msau42

[davidz627]

/test pull-kubernetes-e2e-gce

[tpepper]

/milestone v1.12
/priority critical-urgent

[msau42]

/lgtm

This is fine for quickly fixing the release blocking tests.

Let's still investigate why the AfterEach cleanup is not working.

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

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

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:

• test/e2e/storage/OWNERS [msau42]

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

[davidz627]

I looked up the error message we're seeing and found in the code:

// but they were all undergoing deletion (kubelet is probably culprit, check NodeLost)
return fmt.Errorf("namespace %v was not deleted with limit: %v, pods remaining: %v", namespace, err, remainingPods)

meaning the pods already had deletion timestamp but they were not deleted in time.

[msau42]

That's from the namespace deletion that happens at the very end of the tests. But in the case of nfs and gluster, where the storage backends are pods themselves, this is too late because the storage backend pods get deleted at the end of the test. So if the client Pod gets deleted after the storage backend Pod, then the client Pod deletion will hang forever.

[k8s-github-robot]

Automatic merge from submit-queue (batch tested with PRs 67368, 59930, 68074). If you want to cherry-pick this change to another branch, please follow the instructions here: https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md.

[mkimuram]

@davidz627 @msau42

Let's still investigate why the AfterEach cleanup is not working.

I confirmed that after #68569 is applied, even deleting this commit doesn't make error.
So, the root cause of this issue seems failure in updating resource that will be used in AfterEach. And it is fixed there.

Although this commit isn't necessarily needed now, but it isn't harmful and is consistent with other test functions. Therefore, I think that we can keep this commit as it is.

I think that we can regard that issue #68073 is now resolved completely.