You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Lars Hofhansl (JIRA)" <ji...@apache.org> on 2013/12/02 23:24:36 UTC
[jira] [Commented] (HBASE-8553) improve unit-test coverage of
package org.apache.hadoop.hbase.mapreduce.hadoopbackport
[ https://issues.apache.org/jira/browse/HBASE-8553?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13836992#comment-13836992 ]
Lars Hofhansl commented on HBASE-8553:
--------------------------------------
This is a good test! It does break with Hadoop 2.x. Looks like this is expected; nevertheless it is annoying.
[~iveselovsky], do you have any insight here? Is that something that we could fix easily? There are more folks these days who run 0.94 against a Hadoop 2.x version. Not a big deal, just asking.
> improve unit-test coverage of package org.apache.hadoop.hbase.mapreduce.hadoopbackport
> --------------------------------------------------------------------------------------
>
> Key: HBASE-8553
> URL: https://issues.apache.org/jira/browse/HBASE-8553
> Project: HBase
> Issue Type: Test
> Affects Versions: 0.94.9
> Reporter: Ivan A. Veselovsky
> Assignee: Ivan A. Veselovsky
> Fix For: 0.94.13
>
> Attachments: HBASE-8553-0.94--N2.patch
>
>
> The patch is for branch 0.94 only.
> The class InputSampler is modified because need to fix bug there: in method "run(String[] args)" should be "TotalOrderPartitioner.setPartitionFile(job.getConfiguration(), outf);" instead of "TotalOrderPartitioner.setPartitionFile(getConf(), outf);". Otherwise it is impossible to set output file correctly.
--
This message was sent by Atlassian JIRA
(v6.1#6144)