[
https://issues.apache.org/jira/browse/MAPREDUCE-1333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12796570#action_12796570
]
Zhaoning Zhang commented on MAPREDUCE-1333:
-------------------------------------------
I think it's a general fall back though I run sort benchmark on my cluster
with 20 nodes.
I set the mapred.tasktracker.{map|reduce}.tasks.maximum = 1 individually. In
a single job, map task running simultaneously with a reduce task will be
more slower than the solo one. And for the inter-dependency of the tasks,
shuffles in the reduce tasks will waiting for the maps and the response time
of the job will increase.
2010/1/5 Hong Tang (JIRA) <[email protected]>
--
Thank you!
谢谢!
张钊宁
zzningxp
> Parallel running tasks on one single node may slow down the performance
> -----------------------------------------------------------------------
>
> Key: MAPREDUCE-1333
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-1333
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: jobtracker, task, tasktracker
> Affects Versions: 0.20.1
> Reporter: Zhaoning Zhang
>
> When I analysis running tasks performance, I found that parallel running
> tasks on one single node will not be better performance than the serialized
> ones.
> We can set mapred.tasktracker.{map|reduce}.tasks.maximum = 1 individually,
> but there will be parallel map AND reduce tasks.
> And I wonder it's true in the real commercial clusters?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
