[ https://issues.apache.org/jira/browse/SPARK-36423?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17393646#comment-17393646 ]
Apache Spark commented on SPARK-36423: -------------------------------------- User 'Victsm' has created a pull request for this issue: https://github.com/apache/spark/pull/33649 > Randomize blocks within a push request before pushing to improve block merge > ratio > ---------------------------------------------------------------------------------- > > Key: SPARK-36423 > URL: https://issues.apache.org/jira/browse/SPARK-36423 > Project: Spark > Issue Type: Sub-task > Components: Shuffle, Spark Core > Affects Versions: 3.2.0 > Reporter: Min Shen > Priority: Major > > On the client side, we are currently randomizing the order of push requests > before processing each request. In addition we canĀ further randomize the > order of blocks within each push request before pushing them. > In our benchmark, this has resulted in a 60%-70% reduction of blocks that > fail to be merged due to bock collision (the existing block merge ratio is > already pretty good in general, and this further improves it). -- This message was sent by Atlassian Jira (v8.3.4#803005) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org