[
https://issues.apache.org/jira/browse/MAPREDUCE-279?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13039993#comment-13039993
]
Tom White commented on MAPREDUCE-279:
-------------------------------------
I'm wondering what the maven modules might look like for this when integrated
into trunk. Something like:
* api - containing the user-facing public API of MapReduce (from
org.apache.hadoop.mapred(uce)). When MAPREDUCE-1638 is done it will be possible
to split the API into a self-contained tree (no dependencies on other parts of
MapReduce).
* lib - containing the user-facing public MapReduce libraries (from
org.apache.hadoop.mapred and org.apache.hadoop.mapred(uce).lib). There's a
patch in MAPREDUCE-1478 to perform this separation.
* classic-impl - containing the implementation classes for MapReduce. This is
what's left over after doing MAPREDUCE-1638 and MAPREDUCE-1478.
* nextgen-impl - this is mr-client in the MR-279 branch, which I think should
be renamed, since it's not immediately clear what it's a client of in the
context of the whole MapReduce project. It has submodules app, common, hs,
jobclient, shuffle.
* yarn - the yarn framework from the MR-279 branch. Yarn is broken into
submodules too.
Given the progress on mavenizing common (HADOOP-6671), is it worth integrating
MAPREDUCE-279 at the same time as doing the full Mavenization of MapReduce?
That would seem ideal, but perhaps there's an alternative I haven't considered.
> Map-Reduce 2.0
> --------------
>
> Key: MAPREDUCE-279
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-279
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: jobtracker, tasktracker
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Fix For: 0.23.0
>
> Attachments: MR-279.patch, MR-279.patch, MR-279.sh,
> MR-279_MR_files_to_move.txt, capacity-scheduler-dark-theme.png,
> multi-column-stable-sort-default-theme.png, yarn-state-machine.job.dot,
> yarn-state-machine.job.png, yarn-state-machine.task-attempt.dot,
> yarn-state-machine.task-attempt.png, yarn-state-machine.task.dot,
> yarn-state-machine.task.png
>
>
> Re-factor MapReduce into a generic resource scheduler and a per-job,
> user-defined component that manages the application execution.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
