You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "He Qi (Jira)" <ji...@apache.org> on 2023/03/08 12:46:00 UTC

[jira] [Updated] (SPARK-42719) `ShuffledRowRdd#getPreferredLocations` should respect to `spark.shuffle.reduceLocality.enabled`

     [ https://issues.apache.org/jira/browse/SPARK-42719?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

He Qi updated SPARK-42719:
--------------------------
    Description: 
Discuss as [https://github.com/apache/spark/pull/40307]

{{getPreferredLocations}} in {{ShuffledRowRDD}} should return {{Nil}} at the very beginning in case {{spark.shuffle.reduceLocality.enabled = false}} (conceptually).

This logic is pushed into MapOutputTracker though - and {{getPreferredLocationsForShuffle}} honors {{spark.shuffle.reduceLocality.enabled}} - but {{getMapLocation}} does not.

So the fix would be to fix {{getMapLocation}} to honor the parameter.

> `ShuffledRowRdd#getPreferredLocations` should respect to `spark.shuffle.reduceLocality.enabled`
> -----------------------------------------------------------------------------------------------
>
>                 Key: SPARK-42719
>                 URL: https://issues.apache.org/jira/browse/SPARK-42719
>             Project: Spark
>          Issue Type: Improvement
>          Components: Spark Core
>    Affects Versions: 3.5.0
>            Reporter: He Qi
>            Priority: Major
>
> Discuss as [https://github.com/apache/spark/pull/40307]
> {{getPreferredLocations}} in {{ShuffledRowRDD}} should return {{Nil}} at the very beginning in case {{spark.shuffle.reduceLocality.enabled = false}} (conceptually).
> This logic is pushed into MapOutputTracker though - and {{getPreferredLocationsForShuffle}} honors {{spark.shuffle.reduceLocality.enabled}} - but {{getMapLocation}} does not.
> So the fix would be to fix {{getMapLocation}} to honor the parameter.



--
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