[
https://issues.apache.org/jira/browse/MAPREDUCE-2905?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13092995#comment-13092995
]
Jeff Bean commented on MAPREDUCE-2905:
--------------------------------------
Example:
5 node cluster, 5 slots per node.
Set assign multiple to true.
Submit a sleep job with 7 mappers.
All 7 mappers are assigned to nodes A and B.
Set assign multiple to false. Bounce job tracker.
Submit a sleep job with 7 mappers.
7 mappers now spread evenly across cluster.
I should be able to submit 2 sleep jobs, maybe job 1 has 7 mappers, job 2 has
700 mappers. I want to set assignmultiple on job 2 so that evey node saturates
quickly, while job 1 has assignmultiple set to false, which means it gets one
task assigned at a time and therefore spreads out more evently.
> Allow mapred.fairscheduler.assignmultple to be set per job
> ----------------------------------------------------------
>
> Key: MAPREDUCE-2905
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2905
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: contrib/fair-share
> Affects Versions: 0.20.2
> Reporter: Jeff Bean
>
> We encountered a situation where in the same cluster, large jobs benefit from
> mapred.fairscheduler.assignmultiple, but small jobs with small numbers of
> mappers do not: the mappers all clump to fully occupy just a few nodes, which
> causes those nodes to saturate and bottleneck. The desired behavior is to
> spread the job across more nodes so that a relatively small job doesn't
> saturate any node in the cluster.
> Testing has shown that setting mapred.fairscheduler.assignmultiple to false
> gives the desired behavior for small jobs, but is unnecessary for large jobs.
> However, since this is a cluster-wide setting, we can't properly tune.
> It'd be nice if jobs can set a param similar to
> mapred.fairscheduler.assignmultiple on submission to better control the task
> distribution of a particular job.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
