[
https://issues.apache.org/jira/browse/SQOOP-2339?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14520179#comment-14520179
]
ASF subversion and git services commented on SQOOP-2339:
--------------------------------------------------------
Commit fa4e90365024b7aa0b07591a873b07d87c75e071 in sqoop's branch
refs/heads/trunk from [~jarcec]
[ https://git-wip-us.apache.org/repos/asf?p=sqoop.git;h=fa4e903 ]
SQOOP-2339: Move sub-directory might fail in append mode
(Qian Xu via Jarek Jarcec Cecho)
> Move sub-directory might fail in append mode
> --------------------------------------------
>
> Key: SQOOP-2339
> URL: https://issues.apache.org/jira/browse/SQOOP-2339
> Project: Sqoop
> Issue Type: Bug
> Affects Versions: 1.4.5
> Reporter: Qian Xu
> Assignee: Qian Xu
> Fix For: 1.4.7
>
> Attachments: SQOOP-2339.patch
>
>
> In append mode, Sqoop will first create dataset in a temporary directory and
> then move it target directory. I noticed that if the temporary directory
> contains sub-directory and the target directory has a same directory, the
> import might fail with an IOException. This will happen in old hadoop
> distribution such as hadoop-2.0.4-alpha.
> There is no problem with hadoop-2.5.0. With hadoop-2.5.0, duplicated
> directory will be renamed by adding a suffix "-0000" (the digit might be
> increased).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
