[
https://issues.apache.org/jira/browse/MAPREDUCE-4049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13542107#comment-13542107
]
Avner BenHanoch commented on MAPREDUCE-4049:
--------------------------------------------
Hi Alejandro,
thanks for your comments!
*my 2nd point above was for the trunk.* True, in MRv2 AuxiliaryServices are
plugable, still they don't get APPLICATION_INIT events; hence, they don't know
to map jobId->userId. We need to send this event to all AuxiliaryServices, or
to define 2 groups of AuxiliaryServices: 1 that get this event, and 1 that
doesn't get this event. *In the current code, this event is only sent to
"mapreduce.shuffle" using hard-coded string rather than relying on any conf
settings*.
For the branch-1 comments:
Please notice that ShuffleProvider has different semantics than
ShuffleConsumer, because a 3rd party shuffle-provider always runs in addition
to the default shuffle-provider. multiple jobs can run in parallel, resulting
in various shuffleConsumers in parallel (in different Jobs/ReduceTasks). Hence,
all possible providers should exists in parallel. Saying that, the semantic of
ShuffleProvider plugin is in addition to the default shuffle-provider. Hence
TT should not fail for that.
(for the rest of your branch-1 comments: yes, you are right on all!)
> plugin for generic shuffle service
> ----------------------------------
>
> Key: MAPREDUCE-4049
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4049
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: performance, task, tasktracker
> Affects Versions: 1.0.3, 1.1.0, 2.0.0-alpha, 3.0.0
> Reporter: Avner BenHanoch
> Assignee: Avner BenHanoch
> Labels: merge, plugin, rdma, shuffle
> Fix For: 3.0.0
>
> Attachments: HADOOP-1.x.y.patch, Hadoop Shuffle Plugin Design.rtf,
> mapreduce-4049.patch
>
>
> 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)
> # I am providing link for downloading UDA - Mellanox's open source plugin
> that implements generic shuffle service using RDMA and levitated merge.
> Note: At this phase, the code is in C++ through JNI and you should consider
> it as beta only. Still, it can serve anyone that wants to implement or
> contribute to levitated merge. (Please be advised that levitated merge is
> mostly suit in very fast networks) -
> [http://www.mellanox.com/content/pages.php?pg=products_dyn&product_family=144&menu_section=69]
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
