You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-dev@hadoop.apache.org by "Zhijie Shen (JIRA)" <ji...@apache.org> on 2013/08/14 23:05:50 UTC

[jira] [Created] (MAPREDUCE-5459) Update the doc of running MRv1 examples jar on YARN

Zhijie Shen created MAPREDUCE-5459:
--------------------------------------

             Summary: Update the doc of running MRv1 examples jar on YARN
                 Key: MAPREDUCE-5459
                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-5459
             Project: Hadoop Map/Reduce
          Issue Type: Bug
            Reporter: Zhijie Shen
            Assignee: Zhijie Shen


In addition to adding two env vars: HADOOP_USER_CLASSPATH_FIRST and HADOOP_CLASSPATH, we still need to add
{code}
    <property>
        <name>mapreduce.job.user.classpath.first</name>
        <value>true</value>
    </property>
{code}
in mapred-site.xml to make sure that the MRv1 examples jar runs correctly on YARN. Some examples will use Java reflection to find the classes in the examples jar dynamically when they are running. With this configuration, the MRv1 examples jar will appear before the MRv2 examples jar in CLASSPATH of the processes in YARN containers. Therefore, the classes found via reflection will be picked from MRv1 examples jar instead of MRv2 examples jar as well.

MapReduce_Compatibility_Hadoop1_Hadoop2.apt.vm needs to be updated to document this.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira