[FLINK-38131][flink-runtime] fix that JM still redistributes broadcas…

[xiangforever2014]

…t state when all subtask have reported their states.

What is the purpose of the change

fix that JM still redistributes broadcast state when all subtask have reported their states.

(For example: This pull request makes task deployment go through the blob server, rather than through RPC. That way we avoid re-transferring them on each deployment (during recovery).)

Brief change log

Fixed the logic of judging whether a Broadcast state is partially reported.

(for example:)

• The TaskInfo is stored in the blob store on job creation time as a persistent artifact
• Deployments RPC transmits only the blob storage reference
• TaskManagers retrieve the TaskInfo from the blob cache

Verifying this change

Please make sure both new and modified tests in this PR follow the conventions for tests defined in our code quality guide.

added new UT for testing partially reported logic.
org.apache.flink.runtime.checkpoint.StateAssignmentOperationTest#testPartiallyReported

(Please pick either of the following options)

This change is a trivial rework / code cleanup without any test coverage.

(or)

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

The correctness of broad cast state repartition is covered by existing tests, such as:

1. org.apache.flink.runtime.checkpoint.StateAssignmentOperationTest#testRepartitionBroadcastState
2. org.apache.flink.runtime.checkpoint.StateAssignmentOperationTest#testRepartitionBroadcastStateWithNullSubtaskState
3. org.apache.flink.runtime.checkpoint.StateAssignmentOperationTest#testRepartitionBroadcastStateWithEmptySubtaskState

(or)

This change added tests and can be verified as follows:

(example:)

• Added integration tests for end-to-end deployment with large payloads (100MB)
• Extended integration test for recovery after master (JobManager) failure
• Added test that validates that TaskInfo is transferred only once across recoveries
• Manually verified the change by running a 4 node cluster with 2 JobManagers and 4 TaskManagers, a stateful streaming program, and killing one JobManager and two TaskManagers during the execution, verifying that recovery happens correctly.

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

• Dependencies (does it add or upgrade a dependency): (yes / no)
• The public API, i.e., is any changed class annotated with @Public(Evolving): (yes / no)
• The serializers: (yes / no / don't know)
• The runtime per-record code paths (performance sensitive): (yes / no / don't know)
• Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: (yes / no / don't know)
• The S3 file system connector: (yes / no / don't know)

Documentation

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

[flinkbot]

CI report:

• 3444cf0 Azure: SUCCESS

Bot commands

The @flinkbot bot supports the following commands:

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

[xiangforever2014]

@flinkbot run azure

[xiangforever2014]

@flinkbot run azure

[Zakelly]

Thanks for the fix, basically LGTM

[xiangforever2014]

@Zakelly Comments resolved, PTAL when you're free, thx~

[Zakelly]

Thanks for the update. Merging...