[ https://issues.apache.org/jira/browse/MRUNIT-157?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14739079#comment-14739079
]
Brock Noland commented on MRUNIT-157:
-------------------------------------
Yep NullWritable.get() is the correct method.
> allow null keys and values as output, expected output
> -----------------------------------------------------
>
> Key: MRUNIT-157
> URL: https://issues.apache.org/jira/browse/MRUNIT-157
> Project: MRUnit
> Issue Type: Bug
> Affects Versions: 0.9.0
> Reporter: Carlos Espinoza
> Assignee: Jim Donofrio
> Priority: Critical
> Fix For: 1.2.0
>
>
> from comments at the end of MRUNIT-66:
> Hadoop allows null keys and values, when not used as intermediate keys,values between
mappers and reducers. MRUnit should also allow allow null keys and values as output and expected
output. The Serialization class will need to be modified to not try to copy null's into the
output, expected output
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
|