[ 
https://issues.apache.org/jira/browse/SPARK-29002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17343740#comment-17343740
 ] 

Penglei Shi edited comment on SPARK-29002 at 5/13/21, 6:58 AM:
---------------------------------------------------------------

When changing a SMJ to BHJ, there will be an additional BroadcastQueryStageExec 
for build side, the new broadcast query stage contains a 
CustomShuffleReaderExec after applying queryStageOptimizerRules, and the 
CoalesceShufflePartitions inĀ finalStageOptimizerRules will not work. Am i 
correct? And on the other hand, i think it is notĀ proper to coalesce partitions 
for probe side in BHJ.

!image-2021-05-13-12-12-23-530.png|width=524,height=376!


was (Author: penglei shi):
When changing a SMJ to BHJ, there will be an additional BroadcastQueryStageExec 
for build side, the new broadcast query stage contains a 
CustomShuffleReaderExec after applying queryStageOptimizerRules, and the 
CoalesceShufflePartitions inĀ finalStageOptimizerRules will not work. Am i 
correct?

!image-2021-05-13-12-12-23-530.png|width=524,height=376!

> Avoid changing SMJ to BHJ if the build side has a high ratio of empty 
> partitions
> --------------------------------------------------------------------------------
>
>                 Key: SPARK-29002
>                 URL: https://issues.apache.org/jira/browse/SPARK-29002
>             Project: Spark
>          Issue Type: Sub-task
>          Components: SQL
>    Affects Versions: 3.0.0
>            Reporter: Wei Xue
>            Assignee: Wei Xue
>            Priority: Major
>             Fix For: 3.0.0
>
>         Attachments: image-2021-05-13-12-12-23-530.png
>
>




--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to