[
https://issues.apache.org/jira/browse/PHOENIX-2274?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14877143#comment-14877143
]
James Taylor commented on PHOENIX-2274:
---------------------------------------
+1. Thanks south for the patch, [~maryannxue]. Please commit to master and 4.x
branches.
> Sort-merge join could not optimize out the sort on the right table
> ------------------------------------------------------------------
>
> Key: PHOENIX-2274
> URL: https://issues.apache.org/jira/browse/PHOENIX-2274
> Project: Phoenix
> Issue Type: Bug
> Affects Versions: 4.5.2
> Reporter: Maryann Xue
> Assignee: Maryann Xue
> Priority: Minor
> Attachments: PHOENIX-2274.patch
>
>
> {code}
> CREATE TABLE t (
> mypk VARCHAR(100) NOT NULL PRIMARY KEY,
> mycol VARCHAR(200));
> SELECT * FROM t t1 JOIN t t2 ON t1.mypk = t2.mypk;
> {code}
> would produce a plan like:
> {code}
> SORT-MERGE-JOIN (INNER) TABLES
> CLIENT PARALLEL 1-WAY FULL SCAN OVER T
> AND
> CLIENT PARALLEL 1-WAY FULL SCAN OVER T
> SERVER SORTED BY [MYPK]
> CLIENT MERGE SORT
> {code}
> , while it should have generated a plan like:
> {code}
> SORT-MERGE-JOIN (INNER) TABLES
> CLIENT PARALLEL 1-WAY FULL SCAN OVER T
> AND
> CLIENT PARALLEL 1-WAY FULL SCAN OVER T
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
