[
https://issues.apache.org/jira/browse/HIVE-9106?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14248564#comment-14248564
]
Pengcheng Xiong commented on HIVE-9106:
---------------------------------------
[~ashutoshc], I studied the 4 failures. 2 out of the 4,
vector_partition_diff_num_cols and vector_partition_diff_num_cols are probably
due to Sergey's enable CBO patch (and the following Vikram's patch is not
updated so that the column names are different). I run
groupby3_map_multi_distinct.q and optimize_nullscan.q, both of them passed on
my laptop. Thus, I think it is safe to check in. [~jpullokkaran], please let us
know if you have other opinions.
> improve the performance of null scan optimizer when several table scans share
> a physical path
> ---------------------------------------------------------------------------------------------
>
> Key: HIVE-9106
> URL: https://issues.apache.org/jira/browse/HIVE-9106
> Project: Hive
> Issue Type: Improvement
> Reporter: Pengcheng Xiong
> Assignee: Pengcheng Xiong
> Priority: Minor
> Attachments: HIVE-9106.00.patch
>
>
> Current fix HIVE-9053 addresses the correctness issue. The solution can be
> improved further when several table scans share a physical path.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
