[
https://issues.apache.org/jira/browse/HIVE-5297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13779243#comment-13779243
]
Sergey Shelukhin commented on HIVE-5297:
----------------------------------------
Btw, as we have discussed this doesn't cover the case where column type is
changed when the data is already there, and thus data becomes invalid for a
column (imho in such cases the alter should failed).
Do you want to file a separate JIRA
> Hive does not honor type for partition columns
> ----------------------------------------------
>
> Key: HIVE-5297
> URL: https://issues.apache.org/jira/browse/HIVE-5297
> Project: Hive
> Issue Type: Bug
> Components: CLI
> Affects Versions: 0.11.0
> Reporter: Vikram Dixit K
> Assignee: Vikram Dixit K
> Attachments: HIVE-5297.1.patch, HIVE-5297.2.patch, HIVE-5297.3.patch,
> HIVE-5297.4.patch, HIVE-5297.5.patch, HIVE-5297.6.patch, HIVE-5297.7.patch,
> HIVE-5297.8.patch
>
>
> Hive does not consider the type of the partition column while writing
> partitions. Consider for example the query:
> {noformat}
> create table tab1 (id1 int, id2 string) PARTITIONED BY(month string,day int)
> row format delimited fields terminated by ',';
> alter table tab1 add partition (month='June', day='second');
> {noformat}
> Hive accepts this query. However if you try to select from this table and
> insert into another expecting schema match, it will insert nulls instead. We
> should throw an exception on such user error at the time the partition
> addition/load happens.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
