[fix] [test] Fix flaky test MetadataStoreStatsTest.testBatchMetadataStoreMetrics

[lifepuzzlefun]

Fixes #19792

Motivation

Fix flaky test MetadataStoreStatsTest.testBatchMetadataStoreMetrics

some unit test create MetadataStore without setting name and not close will cause prometheus metric not unregisterd.
the unittest above will check promethues metric tag name not null.

Modifications

1. set metadata store name to avoid generate metric with name cause relate unit test fail.

2. auto generate metadata store name if caller not specify one for avoid this problem occur again.

Verifying this change

This change is already covered by existing tests, such as (please describe tests).

Does this pull request potentially affect one of the following parts:

If the box was checked, please highlight the changes

• Dependencies (add or upgrade a dependency)
• The public API
• The schema
• The default values of configurations
• The threading model
• The binary protocol
• The REST endpoints
• The admin CLI options
• The metrics
• Anything that affects deployment

Documentation

• doc
• doc-required
• doc-not-needed
• doc-complete

Matching PR in forked repository

PR in forked repository:

[github-actions]

@lifepuzzlefun Please add the following content to your PR description and select a checkbox:

- [ ] `doc` <!-- Your PR contains doc changes -->
- [ ] `doc-required` <!-- Your PR changes impact docs and you will update later -->
- [ ] `doc-not-needed` <!-- Your PR changes do not impact docs -->
- [ ] `doc-complete` <!-- Docs have been already added -->

[lifepuzzlefun]

the auto generate metadata store name will be like org.apache.pulsar.metadata.MetadataStoreConfigTest_testGeneCallerStackNameWhenMetadataStoreNameNotSet_11

I made some change in my local repo to capture the caller lifepuzzlefun#10

and see CI job fail message which wil print stack trace if check metric tag with name empty and make unit test fail.
https://github.com/lifepuzzlefun/pulsar/actions/runs/4391908015/jobs/7691586022

[nodece]

Is the root cause that the metadatastoreName is empty?

[lifepuzzlefun]

Is the root cause that the metadatastoreName is empty?

@nodece The root cause is create one MetadataStore with no name set and not close it.

[nodece]

I think you can throw an exception, so like:

throw new IllegalArgumentException("metadataStoreName is required");

[lifepuzzlefun]

I agree with you. the problem is there is a lot of ut didn't set name. If they don't set name but they close the object the flaky test won't occur.

[nodece]

So can we just need to close the metadatastore to fix the flaky test?

[lifepuzzlefun]

yes, other modification is for this case not happen again and supply some method to find the wrong usage.

the modification won't hurt production.

[nodece]

yes, other modification is for this case not happen again and supply some method to find the wrong usage.

Makes sense! But this is rare.

[lifepuzzlefun]

yes i agree with you. maybe we can add one comment in this unit test point to the current modification to catch up new wrong usage. and the branch will be retain for those who is block by this problem.

I'll remove the current metadataStore name generate code for this patch to merge.

[lifepuzzlefun]

@nodece @coderzc please take a look.

[nodece]

/pulsarbot rerun-failure-checks

[nodece]

/pulsarbot rerun-failure-checks

[nodece]

This fix is invalid, see https://github.com/apache/pulsar/actions/runs/4424020709/jobs/7763604452#step:11:1277