[
https://issues.apache.org/jira/browse/MAPREDUCE-4049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13413344#comment-13413344
]
Arun C Murthy commented on MAPREDUCE-4049:
------------------------------------------
Avner, apologies for taking this long.
The patch looks reasonable, and small(!) which is great.
The concern I have is that this patch introduces an interface (i.e.
ShuffleProvider/Consumer) which isn't present in hadoop-2.x. Should we do both
hadoop-2 and hadoop-1 simultaneously? Else, this 'feature' will break as soon
as we upgrade to hadoop-2.x.
Other nits:
# We should get TaskTracker.MapOutputServlet to implement ShuffleProvider
interface, else it's very easy to break an interface if no one in the core
implements it. For e.g. I have no idea about ShuffleProvider.taskDone or
ShuffleProvider.jobDone are used.
# Minor nits: ShuffleProvider is mis-spelt in a couple of places.
# We should add the new configs for provider/consumer in mapred-default.xml
Again, apologies it took me so long to get to your patch and thanks for being
super-patient! I'd like to work with you to get this committed asap!
> 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: 1.0.3, 1.1.0, 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-1.1.patch, HADOOP-1.x.y-review-oriented.patch, Hadoop Shuffle Consumer
> Plugin TLD.rtf, Hadoop Shuffle Provider Plugin TLD.rtf, mapred-site.xml
>
>
> 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
