[
https://issues.apache.org/jira/browse/HIVE-11090?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14599754#comment-14599754
]
Gunther Hagleitner commented on HIVE-11090:
-------------------------------------------
* Cutting the limit off in all these cases causes the results to be another
10mb in size. Is that really necessary? I understand it's painful to do a limit
and produce stable results, but does adding the required order instead change
the test?
* In vectorization_9 you took out the a bunch of udfs and simplified one of
the queries. Why was that necessary?
* Finally - seems you need to update the spark driver tests as well
> ordering issues with windows unit test runs
> -------------------------------------------
>
> Key: HIVE-11090
> URL: https://issues.apache.org/jira/browse/HIVE-11090
> Project: Hive
> Issue Type: Bug
> Components: Hive
> Affects Versions: 1.2.1
> Reporter: Matt McCline
> Assignee: Matt McCline
> Attachments: HIVE-11090.01.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
