[
https://issues.apache.org/jira/browse/MAPREDUCE-3087?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13121297#comment-13121297
]
Ravi Prakash commented on MAPREDUCE-3087:
-----------------------------------------
No No Hadoop QA. Please don't be cross with me. I like you. I really do.
The trouble you see is, that MRApps:setMRFrameworkClasspath() queries the
classloader for mrapp-generated-classpath. If the CLASSPATH doesn't have the
mr-app.jar, reader (MRApps.java:182) throws an NPE (because classpathFileStream
is null). So the classpath MUST have a valid jar containing
mrapp-generated-classpath. When I modified pom.xml to include
hadoop-mapreduce-client-app, a circular dependency was introduced.
So should I really go through the trouble of having a test-jar containing
mrapp-generated-classpath or will a nice brave committer approve my one line
change? That would be the first patch.
> CLASSPATH not the same after MAPREDUCE-2880
> -------------------------------------------
>
> Key: MAPREDUCE-3087
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3087
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 0.23.0
> Reporter: Ravi Prakash
> Assignee: Ravi Prakash
> Fix For: 0.23.0
>
> Attachments: MAPREDUCE-3087.patch, MAPREDUCE-3087.patch,
> notworking.classpath.uniq, working.classpath.uniq
>
>
> After MAPREDUCE-2880, my classpath was missing key jar files.
--
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
