[
https://issues.apache.org/jira/browse/SPARK-37652?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
mcdull_zhang updated SPARK-37652:
---------------------------------
Description:
`OptimizeSkewedJoin` rule will take effect only when the plan has two
ShuffleQueryStageExec。
With `Union`, it might break the assumption. For example, the following plans
*scenes 1*
{noformat}
Union
SMJ
ShuffleQueryStage
ShuffleQueryStage
SMJ
ShuffleQueryStage
ShuffleQueryStage
{noformat}
*scenes 2*
{noformat}
Union
SMJ
ShuffleQueryStage
ShuffleQueryStage
HashAggregate
{noformat}
when one or more of the SMJ data in the above plan is skewed, it cannot be
processed at present.
It's better to support partial optimize with Union.
was:
`OptimizeSkewedJoin` rule will take effect only when the plan has two
ShuffleQueryStageExec。
With `Union`, it might break the assumption. For example, the following plans
{code: none}
Union
SMJ
SMJ
{code}
> Support optimize skewed join through union
> ------------------------------------------
>
> Key: SPARK-37652
> URL: https://issues.apache.org/jira/browse/SPARK-37652
> Project: Spark
> Issue Type: Sub-task
> Components: SQL
> Affects Versions: 3.2.0
> Reporter: mcdull_zhang
> Priority: Minor
>
> `OptimizeSkewedJoin` rule will take effect only when the plan has two
> ShuffleQueryStageExec。
> With `Union`, it might break the assumption. For example, the following plans
> *scenes 1*
> {noformat}
> Union
> SMJ
> ShuffleQueryStage
> ShuffleQueryStage
> SMJ
> ShuffleQueryStage
> ShuffleQueryStage
> {noformat}
> *scenes 2*
> {noformat}
> Union
> SMJ
> ShuffleQueryStage
> ShuffleQueryStage
> HashAggregate
> {noformat}
> when one or more of the SMJ data in the above plan is skewed, it cannot be
> processed at present.
> It's better to support partial optimize with Union.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
