You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hive.apache.org by "Sergey Shelukhin (JIRA)" <ji...@apache.org> on 2014/04/08 00:42:15 UTC
[jira] [Commented] (HIVE-6825) custom jars for Hive query should be
uploaded to scratch dir per query; and/or versioned
[ https://issues.apache.org/jira/browse/HIVE-6825?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13962346#comment-13962346 ]
Sergey Shelukhin commented on HIVE-6825:
----------------------------------------
[~vikram.dixit] this is the jira
> custom jars for Hive query should be uploaded to scratch dir per query; and/or versioned
> ----------------------------------------------------------------------------------------
>
> Key: HIVE-6825
> URL: https://issues.apache.org/jira/browse/HIVE-6825
> Project: Hive
> Issue Type: Improvement
> Reporter: Sergey Shelukhin
> Assignee: Sergey Shelukhin
> Fix For: 0.14.0
>
> Attachments: HIVE-6825.01.patch, HIVE-6825.patch
>
>
> Currently the jars are uploaded to either user directory or global, whatever is configured, which is a mess and can cause collisions. We can upload to scratch directory, and/or version.
> There's a tradeoff between having to upload files every time (for example, for commonly used things like HBase input format) (which is what is done now, into global/user path), and having a mess of one-off custom jars and files, versioned, sitting in .hiveJars.
--
This message was sent by Atlassian JIRA
(v6.2#6252)