Uh oh!
There was an error while loading. Please reload this page.
[SPARK-55619][SQL][4.1] Fix custom metrics in case of coalesced partitions - #54407
Closed
viirya wants to merge 1 commit into
Closed
[SPARK-55619][SQL][4.1] Fix custom metrics in case of coalesced partitions#54407viirya wants to merge 1 commit into
viirya wants to merge 1 commit into
Conversation
viirya
commented
Feb 20, 2026
MemberAuthor
szehon-ho
approved these changes
Feb 21, 2026
dongjoon-hyun
commented
Feb 21, 2026
Member
Could you re-trigger the failed test pipeline? |
viirya
commented
Feb 21, 2026
MemberAuthor
Re-triggered it but failed again. Re-triggered now. |
viiryaforce-pushed
the
SPARK-55619-branch-4.1
branch
2 times, most recently
from
February 21, 2026 18:43
22ea03f to
84809b7Comparepeter-toth
approved these changes
Feb 21, 2026
…tions Replace PartitionMetricCallback with a ConcurrentHashMap keyed by task attempt ID to correctly track reader state across multiple compute() calls when DataSourceRDD is coalesced. The completion listener is registered only once per task attempt, and metrics are flushed and carried forward between readers as partitions are advanced. Co-Authored-By: Peter Toth <peter.toth@gmail.com> Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
viiryaforce-pushed
the
SPARK-55619-branch-4.1
branch
from
February 21, 2026 21:35
84809b7 to
5425638Compareviirya added a commit
that referenced
this pull request
Feb 22, 2026
…tions ### What changes were proposed in this pull request? Replace `PartitionMetricCallback` with a `ConcurrentHashMap` keyed by task attempt ID to correctly track reader state across multiple `compute()` calls when `DataSourceRDD` is coalesced. The completion listener is registered only once per task attempt, and metrics are flushed and carried forward between readers as partitions are advanced. ### Why are the changes needed? When `DataSourceRDD` is coalesced (e.g., via `.coalesce(1)`), `compute()` gets called multiple times per task, which causes the custom metrics incorrect. ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? Unit test ### Was this patch authored or co-authored using generative AI tooling? Generated-by: Claude Sonnet 4.6 Closes#54407 from viirya/SPARK-55619-branch-4.1. Authored-by: Liang-Chi Hsieh <viirya@gmail.com> Signed-off-by: Liang-Chi Hsieh <viirya@gmail.com>
viirya
commented
Feb 22, 2026
MemberAuthor
Thank you @dongjoon-hyun@szehon-ho@peter-toth |
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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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?
Replace
PartitionMetricCallbackwith aConcurrentHashMapkeyed by task attempt ID to correctly track reader state across multiplecompute()calls whenDataSourceRDDis coalesced. The completion listener is registered only once per task attempt, and metrics are flushed and carried forward between readers as partitions are advanced.Why are the changes needed?
When
DataSourceRDDis coalesced (e.g., via.coalesce(1)),compute()gets called multiple times per task, which causes the custom metrics incorrect.Does this PR introduce any user-facing change?
No
How was this patch tested?
Unit test
Was this patch authored or co-authored using generative AI tooling?
Generated-by: Claude Sonnet 4.6