release-22.2.17-rc: importer: only check import *atomicity* in TestImportWorkerFailure

[yuzefovich]

Backport 1/1 commits from #108626.

/cc @cockroachdb/release

---

Five years ago, in #26881, we changed import to retry on worker failures, which made imports much more resilient to transient failures like nodes going down. As part of this work we created TestImportWorkerFailure which shuts down one node during an import, and checks that the import succeeded. Unfortunately, this test was checked-in skipped, because though imports were much more resilient to node failures, they were not completely resilient in every possible scenario, making the test flakey.

Two months ago, in #105712, we unskipped this test and discovered that in some cases the import statement succeeded but only imported a partial dataset. This non-atomicity seems like a bigger issue than whether the import is able to succeed in every possible transient failure scenario, and is tracked separately in #108547.

This PR changes TestImportWorkerFailure to remove successful import as a necessary condition for test success. Instead, the test now only checks whether the import was atomic; that is, whether a successful import imported all data or a failed import imported none. This is more in line with what we can guarantee about imports today.

Fixes: #102839

Release note: None

Release justification: test-only change.

[blathers-crl]

Thanks for opening a backport.

Please check the backport criteria before merging:

• Backports should only be created for serious
  issues or test-only changes.
• Backports should not break backwards-compatibility.
• Backports should change as little code as possible.
• Backports should not change on-disk formats or node communication protocols.
• Backports should not add new functionality (except as defined
  here).
• Backports must not add, edit, or otherwise modify cluster versions; or add version gates.
• All backports must be reviewed by the owning areas TL and one additional
  TL. For more information as to how that review should be conducted, please consult the backport
  policy.

If your backport adds new functionality, please ensure that the following additional criteria are satisfied:

• There is a high priority need for the functionality that cannot wait until the next release and is difficult to address in another way.
• The new functionality is additive-only and only runs for clusters which have specifically “opted in” to it (e.g. by a cluster setting).
• New code is protected by a conditional check that is trivial to verify and ensures that it only runs for opt-in clusters. State changes must be further protected such that nodes running old binaries will not be negatively impacted by the new state (with a mixed version test added).
• The PM and TL on the team that owns the changed code have signed off that the change obeys the above rules.
• Your backport must be accompanied by a post to the appropriate Slack
  channel (#db-backports-point-releases or #db-backports-XX-X-release) for awareness and discussion.

Also, please add a brief release justification to the body of your PR to justify this
backport.

[cockroach-teamcity]

This change is 

[michae2]

Does this need to be part of the .17 release branch? Once we cut the .17 release we won't be testing the release branch any more.

[yuzefovich]

We did see a failure on this branch #113844, so I don't see any downsides to merging this to prevent further pointless failures that might occur before the branch is deleted. It's a test-only change after all, so it bears no risk.

[michae2]

Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status: complete! 1 of 0 LGTMs obtained (waiting on @DrewKimball and @mgartner)