[
https://issues.apache.org/jira/browse/HADOOP-8599?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13423158#comment-13423158
]
Mariappan Asokan commented on HADOOP-8599:
------------------------------------------
Can someone take a look at MAPREDUCE-4470 and propose a proper fix or retract
HADOOP-8599 from the trunk? This is causing the trunk build to fail for more
than a week. I think returning a 0 split for an empty input is not a good
idea. The original behavior of 1 split with size 0 was good. Can others jump
in and comment on this?
Thanks.
> Non empty response from FileSystem.getFileBlockLocations when asking for data
> beyond the end of file
> -----------------------------------------------------------------------------------------------------
>
> Key: HADOOP-8599
> URL: https://issues.apache.org/jira/browse/HADOOP-8599
> Project: Hadoop Common
> Issue Type: Bug
> Components: fs
> Affects Versions: 1.0.3, 0.23.1, 2.0.0-alpha
> Reporter: Andrey Klochkov
> Assignee: Andrey Klochkov
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: HADOOP-8859-branch-0.23.patch
>
>
> When FileSystem.getFileBlockLocations(file,start,len) is called with "start"
> argument equal to the file size, the response is not empty. There is a test
> TestGetFileBlockLocations.testGetFileBlockLocations2 which uses randomly
> generated "start" and "len" arguments when calling
> FileSystem.getFileBlockLocations and the test fails randomly (when the
> generated start value equals to the file size).
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
