Backport of CSI: ensure initial unpublish state is checkpointed into release/1.3.x

[hc-github-team-nomad-core]

Backport

This PR is auto-generated from #14675 to be assessed for backporting due to the inclusion of the label backport/1.3.x.

The below text is copied from the body of the original PR.

---

A test flake revealed a bug in the CSI unpublish workflow, where an unpublish
that comes from a client that's successfully done the node-unpublish step will
not have the claim checkpointed if the controller-unpublish step fails. This
will result in a delay in releasing the volume claim until the next GC.

This changeset also ensures we're using a new snapshot after each write to raft,
and fixes two timing issues in test where either the volume watcher can
unpublish before the unpublish RPC is sent or we don't wait long enough in
resource-restricted environments like GHA.

---

Two notes for reviewers:

• This is somewhat related to CSI: failed allocation should not block its own controller unpublish #14484 but is really a distinct bug, so I've given it its own changelog and it'll need backports.
• This doesn't fix the longstanding issue of Nomad thumbing its nose at the concurrency requirements of the CSI spec, inasmuch that the volumewatcher can kick off after we've checkpointed (which doesn't actually result in bad behavior so long as plugins are mostly spec-compliant themselves, but...). I've got some thoughts about fixing that, but it's not going to land in 1.4.0.

[github-actions]

I'm going to lock this pull request because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active contributions.
If you have found a problem that seems related to this change, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.