[
https://issues.apache.org/jira/browse/IMPALA-8369?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16856816#comment-16856816
]
ASF subversion and git services commented on IMPALA-8369:
---------------------------------------------------------
Commit 74a763c20d7e5b706350924ce8d71b517c7f5ca5 in impala's branch
refs/heads/master from Zoltan Borok-Nagy
[ https://gitbox.apache.org/repos/asf?p=impala.git;h=74a763c ]
IMPALA-8369: Fix HMS integration tests for Hive 3
Hive 3 changed the output format of the 'describe formatted
<table> <column>' command. Added a new function that can parse
the new output.
Hive 3 also got more permissive during scanning, e.g. it can scan
a Parquet column of INT type while the HMS metadata for that column
is STRING. It does automatic conversion during scanning. However,
it doesn't do the opposite, i.e. when the Parquet column type is
STRING and the HMS column type is INT it raises an exception. Extended
the test with this case.
Change-Id: I51c91c7fc706175295b78abaacf47a86156714ce
(cherry picked from commit 5fb83e73e83d5a6496553d88daa9a2510e5b97bd)
Reviewed-on: http://gerrit.cloudera.org:8080/13475
Reviewed-by: Impala Public Jenkins <[email protected]>
Tested-by: Impala Public Jenkins <[email protected]>
> Impala should be able to interoperate with Hive 3.1.0
> -----------------------------------------------------
>
> Key: IMPALA-8369
> URL: https://issues.apache.org/jira/browse/IMPALA-8369
> Project: IMPALA
> Issue Type: Improvement
> Reporter: Vihang Karajgaonkar
> Assignee: Vihang Karajgaonkar
> Priority: Major
> Labels: impala-acid
>
> Currently, Impala only works with Hive 2.1.1. Since Hive 3.1.0 has been
> released for a while it would be good to add support for Hive 3.1.0 (HMS
> 3.1.0). This patch will focus on ability to connect to HMS 3.1.0 and run
> existing tests. It will not focus on adding support for newer features like
> ACID in Hive 3.1.0 which can be taken up as separate JIRA.
> It would be good to make changes to Impala source code such that it can work
> with both Hive 2.1.0 and Hive 3.1.0 without the need to create a separate
> branch. However, this should be a aspirational goal. If we hit a blocker we
> should investigate alternative approaches.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
