Flaky Test: Checkpoints "after each" fails on subsequent runs

[MichaelSwigerAtBentley]

Checkpoints Test 'after each' fails due to a busy container:

  52 passing (3m)
  2 pending
  1 failing

  1) Checkpoints
       "after each" hook for "should be able to open multiple checkpoints in same container when sas expires":
     Error: detach error: container is busy: imodelblocks-b99583da-1de2-43e3-9e6e-a1ad69d1d4a6
      at Function.cleanup (/home/vsts/work/1/s/core/backend/src/CheckpointManager.ts:158:15)
      at Context.<anonymous> (src/integration/Checkpoints.test.ts:112:25)
      at processImmediate (node:internal/timers:485:21)
      at process.callbackTrampoline (node:internal/async_hooks:130:17)

This failure only occurs on subsequent reruns, the after each might not be cleaning up after the tests completely

[tcobbs-bentley]

I have investigated this, and don't know what is causing the problem. I do have some comments to make, though:

1. I don't think that it is triggered by multiple attempts on the pipeline ("subsequent reruns"). The reason for this is because the before hook for "Checkpoints" removes the Cloud SQLite cache directory. As far as I can tell, IModelJsFs.removeSync will throw an exception if it fails. Since that didn't happen, I think it's safe to say that the directory is gone before the rest of the "Checkpoint" suite runs.
2. Because of the above, I believe that something is causing the imodel to still be open at the time of the V2CheckpointManager.cleanup call.
3. The container in question (imodelblocks-b99583da-1de2-43e3-9e6e-a1ad69d1d4a6) appears to the the Cloud SQLite cloud container for "Stadium Dataset 1".

[tcobbs-bentley]

@MichaelSwigerAtBentley As far as I can tell, this hasn't failed in the last 30 days. Do you think it's ok to close this issue?

[MichaelSwigerAtBentley]

@MichaelSwigerAtBentley As far as I can tell, this hasn't failed in the last 30 days. Do you think it's ok to close this issue?

Yeah, I think so. We can always reopen the issue if we notice it again.