[jira] [Assigned] (HIVE-19804) msck repair should hold locks

["Vihang Karajgaonkar (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HIVE-19804?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Vihang Karajgaonkar reassigned HIVE-19804:
------------------------------------------

    Assignee: Vihang Karajgaonkar

> msck repair should hold locks
> -----------------------------
>
>                 Key: HIVE-19804
>                 URL: https://issues.apache.org/jira/browse/HIVE-19804
>             Project: Hive
>          Issue Type: Improvement
>            Reporter: Vihang Karajgaonkar
>            Assignee: Vihang Karajgaonkar
>            Priority: Major
>
> {msck repair table <tablename>}} does not hold locks. This can lead to weird race conditions when concurrent sessions are running on the same table.  For example if two sessions run msck on the same table at the same time, they both try to add partitions and they might both end up with failures due to AlreadyExistsException. Another example would be if a query is running on a partitioned table while some other session issues msck repair which add/drops the partitions, it could trigger errors during query execution.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)