[
https://issues.apache.org/jira/browse/SPARK-37567?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
junbiao chen updated SPARK-37567:
---------------------------------
Description:
use case：query2 in TPC-DS.There are three exchange subquery will scan the same
table "store_sales" in logical plan,these subqueries meet exchange reuse rule.I
confirm that the exchange use rule work in physical plan.But when spark
execute the physical plan,I find out
exchange reuse failed,reused exchange has been executed twice.
physical plan：
!physical plan-query2.png!
execution stages：
!execution stage-query2.png!
!execution stage(1)-query2.png!
was:
use case：query2 in TPC-DS.There are three exchange subquery will scan the same
table "store_sales" in logical plan,these subqueries meet exchange reuse rule.I
confirm that the exchange use rule work in physical plan.But when spark
execute the physical plan,I find out
exchange reuse failed,reused exchange has been executed twice.
> reuse Exchange failed
> ----------------------
>
> Key: SPARK-37567
> URL: https://issues.apache.org/jira/browse/SPARK-37567
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 3.1.1
> Reporter: junbiao chen
> Priority: Major
> Labels: performance
> Attachments: execution stage(1)-query2.png, execution
> stage-query2.png, physical plan-query2.png
>
>
> use case：query2 in TPC-DS.There are three exchange subquery will scan the
> same table "store_sales" in logical plan,these subqueries meet exchange reuse
> rule.I confirm that the exchange use rule work in physical plan.But when
> spark execute the physical plan,I find out
> exchange reuse failed,reused exchange has been executed twice.
> physical plan：
> !physical plan-query2.png!
>
> execution stages：
> !execution stage-query2.png!
>
> !execution stage(1)-query2.png!
--
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
