[
https://issues.apache.org/jira/browse/SPARK-37023?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17437041#comment-17437041
]
Apache Spark commented on SPARK-37023:
--------------------------------------
User 'rmcyang' has created a pull request for this issue:
https://github.com/apache/spark/pull/34461
> Avoid fetching merge status when shuffleMergeEnabled is false for a
> shuffleDependency during retry
> --------------------------------------------------------------------------------------------------
>
> Key: SPARK-37023
> URL: https://issues.apache.org/jira/browse/SPARK-37023
> Project: Spark
> Issue Type: Sub-task
> Components: Shuffle
> Affects Versions: 3.2.0
> Reporter: Ye Zhou
> Priority: Major
>
> The assertion below in MapOutoutputTracker.getMapSizesByExecutorId is not
> guaranteed
> {code:java}
> assert(mapSizesByExecutorId.enableBatchFetch == true){code}
> The reason is during some stage retry cases, the
> shuffleDependency.shuffleMergeEnabled is set to false, but there will be
> mergeStatus since the Driver has collected the merged status for its shuffle
> dependency. If this is the case, the current implementation would set the
> enableBatchFetch to false, since there are mergeStatus.
> Details can be found here:
> [https://github.com/apache/spark/blob/master/core/src/main/scala/org/apache/spark/MapOutputTracker.scala#L1492]
> We should improve the implementation here.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
