You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@spark.apache.org by "Chao Sun (Jira)" <ji...@apache.org> on 2022/12/09 17:23:00 UTC

[jira] [Updated] (SPARK-41471) SPJ: Reduce Spark shuffle when only one side of a join is KeyGroupedPartitioning

     [ https://issues.apache.org/jira/browse/SPARK-41471?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Chao Sun updated SPARK-41471:
-----------------------------
    Summary: SPJ: Reduce Spark shuffle when only one side of a join is KeyGroupedPartitioning  (was: SPJ: reduce Spark shuffle when only one side of a join is KeyGroupedPartitioning)

> SPJ: Reduce Spark shuffle when only one side of a join is KeyGroupedPartitioning
> --------------------------------------------------------------------------------
>
>                 Key: SPARK-41471
>                 URL: https://issues.apache.org/jira/browse/SPARK-41471
>             Project: Spark
>          Issue Type: Sub-task
>          Components: SQL
>    Affects Versions: 3.3.1
>            Reporter: Chao Sun
>            Priority: Major
>
> When only one side of a SPJ (Storage-Partitioned Join) is {{{}KeyGroupedPartitioning{}}}, Spark currently needs to shuffle both sides using {{{}HashPartitioning{}}}. However, we may just need to shuffle the other side according to the partition transforms defined in {{{}KeyGroupedPartitioning{}}}. This is especially useful when the other side is relatively small.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org