[
https://issues.apache.org/jira/browse/HAMA-848?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14291649#comment-14291649
]
Edward J. Yoon commented on HAMA-848:
-------------------------------------
Here's more explanation:
When the job is submitted, classic BSP system calculates the number of
available task slots. In YARN case, cluster will be launched on demand and
terminated when job is finished. So, getClusterStatus() returns null.
Please check how mesos module solved this problem.
> Refactor YARN module for hadoop 2.x stable version
> --------------------------------------------------
>
> Key: HAMA-848
> URL: https://issues.apache.org/jira/browse/HAMA-848
> Project: Hama
> Issue Type: Improvement
> Affects Versions: 0.7.0
> Reporter: Edward J. Yoon
> Assignee: Minho Kim
> Labels: patch
> Fix For: 0.7.0
>
> Attachments: HAMA-HDFS2XBuild.Patch, HAMA-YARNDraft.patch,
> HAMA_YARN.patch
>
>
> Current yarn-module was implemented based on Hadoop 0.23.1 version. You'll
> need to use the Hama RPC instead of Hadoop RPC.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
