[
https://issues.apache.org/jira/browse/HIVE-16467?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15972450#comment-15972450
]
Peter Vary commented on HIVE-16467:
-----------------------------------
[~mmccline], the vector_order_null query file is added by HIVE-16448, and the
only problem with that the source data is:
{code}
insert into src_null values (1, 'A');
insert into src_null values (null, null);
insert into src_null values (3, null);
insert into src_null values (2, null);
insert into src_null values (2, 'A');
insert into src_null values (2, 'B');
{code}
And the results are only ordered by a column, and not by b, so as I see both
results are valid:
{code}
NULL NULL
1 A
2 B
2 A
2 NULL
3 NULL
{code}
or
{code}
NULL NULL
1 A
2 A
2 B
2 NULL
3 NULL
{code}
What do you think?
Thanks,
Peter
> Flaky test: TestCliDriver.testCliDriver[vector_order_null]
> ----------------------------------------------------------
>
> Key: HIVE-16467
> URL: https://issues.apache.org/jira/browse/HIVE-16467
> Project: Hive
> Issue Type: Sub-task
> Reporter: Peter Vary
>
> Seems to me, that we have to add a {{-- SORT_QUERY_RESULTS}} to the query
> test.
> The error message is
> {code}
> Client Execution succeeded but contained differences (error code = 1) after
> executing vector_order_null.q
> 153a154
> > 2 B
> 156d156
> < 2 B
> 247a248
> > 2 B
> 250d250
> < 2 B
> 531a532
> > 2 B
> 534d534
> < 2 B
> 626a627
> > 2 B
> 629d629
> < 2 B
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
