[opt](exec) skip result serialization for dry run queries#63356
Open
Mryange wants to merge 1 commit into
Open
Conversation
Contributor
Author
|
/review |
Contributor
|
Thank you for your contribution to Apache Doris. Please clearly describe your PR:
|
Contributor
Author
|
run buildall |
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
No blocking issues found in the actual PR diff. The dry-run fast path now still evaluates output expressions, counts written rows, and skips result serialization/enqueue for both MySQL and Arrow Flight writers, which matches the stated goal and avoids the previous serialization cost. The MySQL unit test covers both normal and dry-run queue behavior; Arrow Flight has analogous logic but no dedicated test in this PR, which is a residual coverage gap rather than a blocker.\n\nCritical checkpoint conclusions: goal is clear and implemented for the touched writers; the change is small and focused. No new concurrency, lifecycle, persistence, storage-format, or dynamic-config concerns were found. Parallel writer paths were both updated. Error propagation remains unchanged around expression evaluation. Test coverage exists for the MySQL path, but Arrow Flight dry-run behavior is not directly tested. No additional user-provided focus points were supplied.
Contributor
TPC-H: Total hot run time: 31404 ms |
Contributor
BE Regression && UT Coverage ReportIncrement line coverage Increment coverage report
|
Contributor
TPC-DS: Total hot run time: 169944 ms |
Mryange
force-pushed
the
skip-result-serialization-for-dry-run
branch
from
176fa8a to
cf8e0d3
Compare
Contributor
Author
|
/review |
Contributor
There was a problem hiding this comment.
Review result: no blocking issues found in the live PR diff. The change keeps dry-run MySQL result writes executing output expressions while skipping serialization/enqueue, and the added unit path verifies no result batch is queued while written rows are still counted. Critical checkpoints: goal/test coverage satisfied for the dry-run skip behavior; scope is small and focused; no new concurrency, lifecycle, config, compatibility, transaction, persistence, or observability concerns identified; no parallel path requiring updates found; test coverage is targeted. User focus: no additional user-provided review focus was supplied.
Contributor
Author
|
run buildall |
Contributor
TPC-H: Total hot run time: 30667 ms |
Contributor
TPC-DS: Total hot run time: 169209 ms |
Contributor
BE Regression && UT Coverage ReportIncrement line coverage Increment coverage report
|
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
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 problem does this PR solve?
Issue Number: N/A
Related PR: None
Problem Summary:
When dry_run_query is enabled, FE only needs the returned row count, but BE still spends most of PhysicalResultSink time serializing MySQL result rows In a local dry-run case against numbers("number"="1000000"), the profile showed AppendBatchTime = 77.689ms, TupleConvertTime = 68.650ms, and ResultSendTime = 2.702us, which means the dry-run path was still paying almost the full result sink conversion cost.
This change keeps output expr evaluation intact, but returns early in the MySQL result writers once the output block is produced in dry-run mode. That preserves returned row accounting while skipping result serialization, block copy, and sink enqueue work that dry-run queries never consume.
Release note
None
Check List (For Author)
Test
Behavior changed:
Does this need documentation?
Check List (For Reviewer who merge this PR)