[
https://issues.apache.org/jira/browse/BIGTOP-1522?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14230542#comment-14230542
]
jay vyas commented on BIGTOP-1522:
----------------------------------
actually, can i ask that you add a comment to your patch ? so its clear why
we're adding that mapred dependency ?
i can see this regressing when someone goes to clean up the puppet code
otherwise !
> nodemanager deployment needs to include hadoop-mapreduce package
> ----------------------------------------------------------------
>
> Key: BIGTOP-1522
> URL: https://issues.apache.org/jira/browse/BIGTOP-1522
> Project: Bigtop
> Issue Type: Bug
> Components: deployment
> Affects Versions: 0.8.0
> Reporter: Konstantin Boudnik
> Assignee: Konstantin Boudnik
> Fix For: 0.9.0
>
> Attachments: BIGTOP-1522.patch
>
>
> It appears that if a cluster is deployed with only
> {code}
> components,hadoop,yarn
> {code}
> set in {{bigtop-deploy/puppet/config/site.csv}} the nodemanager refuses to
> start as it can not find {{org.apache.hadoop.mapred.ShuffleHandler}}
> Clearly, the mapreduce component needs to be installed in order for the
> nodemanager to function properly.
> YARN is totally "not a MapReduce". Yeah, right!
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
