[
https://issues.apache.org/jira/browse/SPARK-23375?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16794791#comment-16794791
]
Xiaoju Wu commented on SPARK-23375:
-----------------------------------
I think there's another case in which sort is redundant:
Sort just under an non-orderPreserving node is redundant, for example:
select count(*) from (select a1 from A order by a2);
+- Aggregate
+- Sort
+- FileScan parquet
> Optimizer should remove unneeded Sort
> -------------------------------------
>
> Key: SPARK-23375
> URL: https://issues.apache.org/jira/browse/SPARK-23375
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 2.3.0
> Reporter: Marco Gaido
> Assignee: Marco Gaido
> Priority: Minor
> Fix For: 2.4.0
>
>
> As pointed out in SPARK-23368, as of now there is no rule to remove the Sort
> operator on an already sorted plan, ie. if we have a query like:
> {code}
> SELECT b
> FROM (
> SELECT a, b
> FROM table1
> ORDER BY a
> ) t
> ORDER BY a
> {code}
> The sort is actually executed twice, even though it is not needed.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
