[jira] Commented: (HIVE-168) join on a subquery with a group by does not work

["Ashish Thusoo (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-168?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12656151#action_12656151 ] 

Ashish Thusoo commented on HIVE-168:
------------------------------------

A couple of minor comments:

1. mergeTask function is actually also doing a splitTask, so we should name it appropriately.
2. setCurrTasks is done twice (inside mergeTask and also in the callers). We should just get rid of the calls in the callers.\

Also please file a separate JIRA to make the explain output better (so that the tasks and seen in the order of dependencies instead of the order of names).


> join on a subquery with a group by does not work
> ------------------------------------------------
>
>                 Key: HIVE-168
>                 URL: https://issues.apache.org/jira/browse/HIVE-168
>             Project: Hadoop Hive
>          Issue Type: Bug
>          Components: Query Processor
>            Reporter: Namit Jain
>            Assignee: Namit Jain
>         Attachments: 168.patch, patch2
>
>
>  SELECT a.key, a.value, b.key, b.value
>  FROM 
>   (
>   SELECT src1.key as key, count(src1.value) AS value FROM src src1 group by src1.key
>   ) a
>  FULL OUTER JOIN 
>  (
>   SELECT src2.key as key, count(distinct(src2.value)) AS value 
>   FROM src1 src2 group by src2.key
>  ) b 
>  ON (a.key = b.key);
> does not generate the plan correctly

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.