[
https://issues.apache.org/jira/browse/DRILL-791?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14003951#comment-14003951
]
Aman Sinha commented on DRILL-791:
----------------------------------
Uploaded patch and tested with a few queries.. one example below:
0: jdbc:drill:zk=local> select min(employee_id) as x, max(employee_id) as y,
min(employee_id) as z from cp.`employee.json`;
+------------+------------+------------+
| x | y | z |
+------------+------------+------------+
| 1 | 1156 | 1 |
+------------+------------+------------+
> Using two aggregate functions in a query causes wrong results.
> --------------------------------------------------------------
>
> Key: DRILL-791
> URL: https://issues.apache.org/jira/browse/DRILL-791
> Project: Apache Drill
> Issue Type: Bug
> Reporter: Mehant Baid
> Assignee: Aman Sinha
> Attachments:
> 0001-Fix-DRILL-791-In-Phase-1of2-use-the-agg-expr-ordinal.patch
>
>
> select min(employee_id), max(employee_id) from cp.`employee.json`;
> +------------+------------+
> | EXPR$0 | EXPR$1 |
> +------------+------------+
> | 1 | 1 |
> +------------+------------+
> In the above query we return the same result for both Min and Max. This seems
> to be related to multi phase aggregate not picking the correct ordinal.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
