[FLINK-22886][State Backends] fix Thread leak in RocksDBStateUploader

[mayuehappy]

What is the purpose of the change

fix the thread leak in RocksDBStateUploader

Brief change log

• name the thread group in RocksDBStateDataTransfer's ExecutorService
• add cleanUp() in RocksDBSnapshotStrategyBase
• invoke checkpointSnapshotStrategy.cleanUp() when RocksDBKeyedStateBackend dispose
• add test case testDisposeCleanUpRocksDBStateDataTransferThread to Detected rocksDBStateDataTransfer thread leak

Verifying this change

This change added tests and can be verified as follows:
testDisposeCleanUpRocksDBStateDataTransferThread

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

• Dependencies (does it add or upgrade a dependency): (no)
• The public API, i.e., is any changed class annotated with @Public(Evolving): (yes)
• The serializers: (no )
• The runtime per-record code paths (performance sensitive): (no )
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn/Mesos, ZooKeeper: (no)
• The S3 file system connector: (no)

Documentation

• Does this pull request introduce a new feature? ( no)
• If yes, how is the feature documented? (not applicable

[flinkbot]

Thanks a lot for your contribution to the Apache Flink project. I'm the @flinkbot. I help the community
to review your pull request. We will use this comment to track the progress of the review.

Automated Checks

Last check on commit 5d04238 (Sat Aug 28 12:19:18 UTC 2021)

Warnings:

• No documentation files were touched! Remember to keep the Flink docs up to date!

_{Mention the bot in a comment to re-run the automated checks.}

Review Progress

• ❓ 1. The [description] looks good.
• ❓ 2. There is [consensus] that the contribution should go into to Flink.
• ❓ 3. Needs [attention] from.
• ❓ 4. The change fits into the overall [architecture].
• ❓ 5. Overall code [quality] is good.

Please see the Pull Request Review Guide for a full explanation of the review process.

The Bot is tracking the review progress through labels. Labels are applied according to the order of the review items. For consensus, approval by a Flink committer of PMC member is required

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot approve description to approve one or more aspects (aspects: description, consensus, architecture and quality)
• @flinkbot approve all to approve all aspects
• @flinkbot approve-until architecture to approve everything until architecture
• @flinkbot attention @username1 [@username2 ..] to require somebody's attention
• @flinkbot disapprove architecture to remove an approval you gave earlier

[flinkbot]

CI report:

• 3cd40d1 UNKNOWN
• 5d04238 Azure: SUCCESS

Bot commands

The @flinkbot bot supports the following commands:

• @flinkbot run travis re-run the last Travis build
• @flinkbot run azure re-run the last Azure build

[mayuehappy]

@flinkbot run azure

[rkhachatryan]

Thanks for the PR @mayuehappy .
I noticed that CI failed after adding commits from two PRs. So I think it was a bad idea and it's better to have two separate PRs (#16168 with production changes and this one for tests). Sorry for the confusion.
Could you please drop its commit (363c912)?

I also have some minor remarks regarding the test, PTAL.

[mayuehappy]

@rkhachatryan Thanks again for your review .
i have drop the commit (363c912) and the new RocksDBStateUploader was passed in TestCase (although it took some minor changes).
After every use of RocksDBKeyedStateBackend that contains this RocksDBStateUploader, it will check whether RocksDBStateUploader has been closed
can you re-review it if you got free time?

[rkhachatryan]

Thanks for updating your PR @mayuehappy.
I've left some comments, PTAL.

[mayuehappy]

@rkhachatryan Thx very much for your patience, I have updated the commit, PTAL

[rkhachatryan]

Thanks a lot for updating the PR @mayuehappy !
I have a couple of more comments (sorry for nit-picking).

[rkhachatryan]

Should we also have a similar here as in setNumberOfTransferingThreads?

[rkhachatryan]

I think there is no need to introduce a setter in RocksIncrementalSnapshotStrategy.
stateUploader can be passed to its constructor and be a final field there (in fact, I think one of the benefits of having a builder is immutability of the object it builds).
WDYT?

[mayuehappy]

Hahahaha. . i have updated again @rkhachatryan ( Perhaps this is why flink can always guarantee a relatively high code quality 👍 )

[rkhachatryan]

Thanks for updating the PR @mayuehappy :) LGTM.

Could you please squash the commits (and change the component label in the commit message to tests or state).

I'd then merge it once the CI passes.

[mayuehappy]

Thanks again , i have squashed the commits and changed the name @rkhachatryan : ）