[
https://issues.apache.org/jira/browse/SPARK-36823?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
XiDuo You updated SPARK-36823:
------------------------------
Description:
For the join if one side is small and other side is large, the shuffle overhead
is also very big. Due to the
bhj limitation, we can only broadcast right side for left join and left side
for right join. So for the other case, we can try to use
`BroadcastNestedLoopJoin` as the join strategy.
was:
For the join if one side is small and other side is large, the shuffle overhead
is also very big. Due to the limitation, we can only broadcast right side for
left join and left side for right join. So for the other case, we can try to
use `BroadcastNestedLoopJoin` as the join strategy.
> Support broadcast nested loop join hint for equi-join
> -----------------------------------------------------
>
> Key: SPARK-36823
> URL: https://issues.apache.org/jira/browse/SPARK-36823
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 3.3.0
> Reporter: XiDuo You
> Priority: Major
>
> For the join if one side is small and other side is large, the shuffle
> overhead is also very big. Due to the
> bhj limitation, we can only broadcast right side for left join and left side
> for right join. So for the other case, we can try to use
> `BroadcastNestedLoopJoin` as the join strategy.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
