[SPARK-54165][CONNECT]Add BatchExecutePlan RPC and reattach support #52852
+1,914
−121
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
### What changes were proposed in this pull request? This PR introduces a new BatchExecutePlan RPC for Spark Connect that allows clients to submit multiple execution plans in a single batch operation, minimizing RPC overhead. It also adds reattach support for both Scala and Python clients to consume results from batch-submitted operations. Key changes: 1. New BatchExecutePlan RPC with unary request/response model 2. Support for client-provided operation IDs 3. Rollback mechanism for submission failures 4. Reattach methods in Scala and Python clients 5. Operations marked as reattachable for later result consumption ### Why are the changes needed? - **Performance**: Reduces RPC overhead when submitting multiple operations - **Flexibility**: Allows fire-and-forget execution pattern with later result retrieval - **Control**: Provides rollback capability for submission failures ### Does this PR introduce any user-facing change? Yes. New APIs added: - Scala: `SparkConnectClient.batchExecute()` and `reattach()` - Python: `SparkConnectClient.batch_execute()` and `reattach_execute()` ### How was this patch tested? - 11 server-side tests in SparkConnectBatchExecuteSuite - 11 client-side tests in SparkConnectClientBatchExecuteSuite - All tests passing with proper formatting and linting ### Key Implementation Details: - rollbackOnFailure only applies to submission failures (invalid UUID, duplicate operation ID), not execution failures - Operations are submitted sequentially and execute independently - Operations are marked as reattachable, allowing clients to consume results via reattach - Comprehensive documentation clarifying the behavior Closes #XXXXX
Member
dongjoon-hyun
left a comment
dongjoon-hyun
left a comment
There was a problem hiding this comment.
Hi, @grundprinzip .
If you don't mind, please file a JIRA issue in the ASF community repository. It will help you prevent potential accidents like the following commits.
Contributor
Author
|
Absolutely, I'll do that as soon as it's more ready to leave draft mode. |
dongjoon-hyun
previously requested changes
Nov 3, 2025
Member
dongjoon-hyun
left a comment
dongjoon-hyun
left a comment
There was a problem hiding this comment.
Got it. I'll block this PR until then. It's because the previous mistake, SPARK-52762, also had the same pattern; The incorrect Draft status remains until it became incorrect commits mistakenly.
Absolutely, I'll do that as soon as it's more ready to leave draft mode.
grundprinzip
changed the title
Contributor
Author
|
I created a JIRA for it. |
Member
|
Thank you! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
This PR introduces a new BatchExecutePlan RPC for Spark Connect that allows clients to submit multiple execution plans in a single batch operation, minimizing RPC overhead. It also adds reattach support for both Scala and Python clients to consume results from batch-submitted operations.
Key changes:
Why are the changes needed?
Does this PR introduce any user-facing change?
Yes. New APIs added:
SparkConnectClient.batchExecute()andreattach()SparkConnectClient.batch_execute()andreattach_execute()How was this patch tested?
Key Implementation Details:
Closes #XXXXX