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

Posted to issues@spark.apache.org by "Hyukjin Kwon (JIRA)" <ji...@apache.org> on 2019/05/21 04:25:27 UTC

[jira] [Updated] (SPARK-11708) 20-25% performance regression in TeraSort

     [ https://issues.apache.org/jira/browse/SPARK-11708?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Hyukjin Kwon updated SPARK-11708:
---------------------------------
    Labels: bulk-closed  (was: )

> 20-25% performance regression in TeraSort
> -----------------------------------------
>
>                 Key: SPARK-11708
>                 URL: https://issues.apache.org/jira/browse/SPARK-11708
>             Project: Spark
>          Issue Type: Bug
>          Components: Shuffle
>    Affects Versions: 1.6.0
>            Reporter: Nishkam Ravi
>            Priority: Major
>              Labels: bulk-closed
>
> There is a 20-25% performance regression for TeraSort variant with hash-based shuffle. This was introduced by 'remove hash file shuffle consolidation' on Sept 18th. Hash-based shuffle is faster than sort-based shuffle for some of our workloads including this one. cc  [~joshrosen]



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

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