[
https://issues.apache.org/jira/browse/MAPREDUCE-5813?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13950315#comment-13950315
]
Sangjin Lee commented on MAPREDUCE-5813:
----------------------------------------
The patch looks good to me. The only nit is to add the right job conf before
Limits.init() so that Limits.init() can use the right job conf.
It's nearly the same fix, but how about
{code}
final JobConf job = new JobConf(MRJobConfig.JOB_CONF_FILE);
Limits.init(job);
...
{code}
> YarnChild does not load job.xml with mapreduce.job.classloader=true
> --------------------------------------------------------------------
>
> Key: MAPREDUCE-5813
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-5813
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, task
> Affects Versions: 2.3.0
> Reporter: Gera Shegalov
> Assignee: Gera Shegalov
> Priority: Blocker
> Attachments: MAPREDUCE-5813.v01.patch
>
>
> {{YarnChild.main}} uses {{JobConf.addResource(String)}} to load {{job.xml}}
> that relies on class loading. When {{mapreduce.job.classloader=true}} the
> job-speicific part of the class path is separated from {{CLASSPATH}} into
> {{APP_CLASSPATH}}. Therefore {{job.xml}} is inaccessible for the default
> class loader. Later {{writeLocalJobFile}} overwrites the correct localized
> {{job.xml}} on disk as well.
> This problem is easily avoided by using {{JobConf.addResource(Path)}} to
> read the localized {{job.xml}} without relying on class loading.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
