hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Avner BenHanoch (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-4049) plugin for generic shuffle service
Date Mon, 28 May 2012 08:48:25 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-4049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13284326#comment-13284326
] 

Avner BenHanoch commented on MAPREDUCE-4049:
--------------------------------------------

Luke, Auburn paper is related to hadoop 0.20.2.  In Mellanox, we ported it to 0.20.203 and
to various 1.0.x versions.  I will very soon port it to 1.1.  We never saw any measurable
difference between the default plugin and the corresponding vanilla hadoop.  Please let me
know if this answers your question.

Alex, in Auburn's merge algorithm, reducers start the merge after all mappers completed. 
Hence, when using the RDMA plugin (that uses that algorithm), it makes sense to put high value
for mapred.reduce.slowstart.completed.maps.  However, Auburn has variant of their merge algorithm
called "hybrid merge" in which reducers start merge in parallel to mappers.  Anyhow, all this
is not part of the default plugin.
                
> plugin for generic shuffle service
> ----------------------------------
>
>                 Key: MAPREDUCE-4049
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4049
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: performance, task, tasktracker
>    Affects Versions: 1.1.0, 1.0.3, 2.0.0-alpha, 3.0.0
>            Reporter: Avner BenHanoch
>              Labels: merge, plugin, rdma, shuffle
>         Attachments: HADOOP-1.0.2.patch, HADOOP-1.0.x.patch, HADOOP-1.0.x.patch, Hadoop
Shuffle Consumer Plugin TLD.rtf, Hadoop Shuffle Provider Plugin TLD.rtf, MAPREDUCE-4049-branch-1.0.2.patch,
mapred-site.xml, mapred.diff, src.tgz, test.diff
>
>
> Support generic shuffle service as set of two plugins: ShuffleProvider & ShuffleConsumer.
> This will satisfy the following needs:
> # Better shuffle and merge performance. For example: we are working on shuffle plugin
that performs shuffle over RDMA in fast networks (10gE, 40gE, or Infiniband) instead of using
the current HTTP shuffle. Based on the fast RDMA shuffle, the plugin can also utilize a suitable
merge approach during the intermediate merges. Hence, getting much better performance.
> # Satisfy MAPREDUCE-3060 - generic shuffle service for avoiding hidden dependency of
NodeManager with a specific version of mapreduce shuffle (currently targeted to 0.24.0).
> References:
> # Hadoop Acceleration through Network Levitated Merging, by Prof. Weikuan Yu from Auburn
University with others, [http://pasl.eng.auburn.edu/pubs/sc11-netlev.pdf]
> # I am attaching 2 documents with suggested Top Level Design for both plugins (currently,
based on 1.0 branch)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message