Izeren opened a new pull request, #26821:
URL: https://github.com/apache/flink/pull/26821
## What is the purpose of the change
Reduces risks of pekko framesize exceeded errors during task deployment
caused by large TaskRestore through asynchronous offloading to the BlobStore.
## Brief change log
* Modifies `Execution#deploy()` to offload large TaskRestore values using
BlobWriter (BlobWriter decides on threshold).
* Offloading of TaskRestore is done asynchronously to avoid blocks the main
thread.
* As `deploy` method now returns earlier, additional
`CompletableFuture<Void>` is exported from Execution for testing purposes.
* Existing impacted tests are adjusted for new behaviour of
`Execution#deploy`.
* Added new tests to verify new code path.
## Verifying this change
* Existing tests.
* Additional unit test to verify new code path:
`ExecutionTest#testExecutionCancelledDuringTaskRestoreOffload`
## 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)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
