[filippomc]

Closes CH-298

Implemented solution

• Create:
  • Cleans up leftovers with discardPartial, which never deletes data it can't vouch for: it refuses a mounted volume that isn't empty, and moves a file at the mountpoint aside to <mp>.discarded-<ts>.
  • Rolls back on failure.
  • Creates the mountpoint only after mkfs.
• Exports:
  • mount-all skips non-directories.
  • Create and remount succeed when the volume is in /var/lib/nfs/etab, even if exportfs -r exits 1 because of another entry. nfs-utils only warns about a bad entry, then still writes etab and flushes the cache.
• Mounting: mountOne no longer creates missing mountpoints, and the watchdog skips volumes that don't have one.
• Delete: unmounts until nothing is mounted (or fails), and keeps a regular file at the mountpoint.
• loop.Attach: serialized within the process, and retries on EBUSY/ENXIO.
• Provisioner: volumeExists also checks for <mp>.quota before skipping a delete.

How to test this PR

• Go unit and scenario tests (no root needed) in nfsvol/internal/mount/lifecycle_test.go, loop/loop_test.go and provisioner_test.go. Each fix was checked by undoing it: all 10 are caught by at least one test.
• test/failover-test.sh:
  • New 4f: export table clean after restart.
  • New 8: provisioning while another export is broken; reproduces this outage.
  • New 9: PVC delete after an interrupted nfsvol delete.
  • Extended 7: no .quota left after PVC delete; image archived when the storage class says so.
  • Not yet run on a cluster. It needs ≥3 nodes; run it on dev before prod.

Sanity checks:

• The pull request is explicitly linked to the relevant issue(s)
• The issue is well described: clearly states the problem and the general proposed solution(s)
• In this PR it is explicitly stated how to test the current change
• The issue seta the scope and the type of issue (bug, story, task, etc.)
• The relevant components are indicated in the issue (if any)
• All the automated test checks are passing
• All the linked issues are included in one Sprint
• All the linked issues are in the Review state
• All the linked issues are assigned

Breaking changes (select one):

• The present changes do not change the preexisting api in any way
• This PR and the issue are tagged as a breaking-change and the migration procedure is well described above

Possible deployment updates issues (select one):

• There is no reason why deployments based on CloudHarness may break after the current update
• This PR and the issue are tagged as alert:deployment

Test coverage (select one):

• Tests for the relevant cases are included in this pr
• The changes included in this pr are out of the current test coverage scope

Documentation (select one):

• The documentation has been updated to match the current changes
• The changes included in this PR are out of the current documentation scope

Nice to have (if relevant):

• Screenshots of the changes
• Explanatory video/animated gif