GitHub user shivaram opened a pull request:

    https://github.com/apache/spark/pull/6652

    [SPARK-2774] Set preferred locations for reduce tasks

    Set preferred locations for reduce tasks. 
    The basic design is that we maintain a map from reducerId to a list of 
(sizes, locations) for each shuffle. We then pick the machines with the top 5 
sizes and set that as preferred locations. 
    
    Picking the top5 needs sorting (well its a little cheaper than sorting, but 
still), so we restrict this feature to cases where we have fewer than 1000 map 
tasks and 1000 reduce tasks.
    
    Resubmission of #4576 and  #1697. Hopefully third time is a charm ? cc 
@kayousterhout for review

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/shivaram/spark-1 reduce-locations

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/spark/pull/6652.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #6652
    
----
commit 3b464b7c488372a305202d17d5efad315eb0e024
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-02-12T21:40:02Z

    Set preferred locations for reduce tasks
    This is another attempt at #1697 addressing some of the earlier concerns.
    This adds a couple of thresholds based on number map and reduce tasks
    beyond which we don't use preferred locations for reduce tasks.
    
    This patch also fixes some bugs in DAGSchedulerSuite where the MapStatus
    objects created didn't have the right number of reducers set.

commit 34d0283fb173c6bf2f4b49d57d2cc10860ed8797
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-02-12T21:47:13Z

    Fix style issues

commit 774751bb08cab26ea2eaf233d466c3509103dc98
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-02-13T00:01:39Z

    Fix bug introduced by line length adjustment

commit bc4dfd66969d3cedd11d166faf83e4bce337c397
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-02-18T02:11:19Z

    Merge branch 'master' of https://github.com/apache/spark into 
reduce-locations

commit 0171d3cf794e4b318a5f07bc299cb0d5acdaf92e
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-03-03T01:16:08Z

    Merge branch 'master' of https://github.com/apache/spark into 
reduce-locations

commit 5093aea0e1c1d2c6614a89276dfc13ee0293a7e8
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-03-08T22:33:47Z

    Merge branch 'master' of https://github.com/apache/spark into 
reduce-locations

commit df14cee75617673f710608b726106434e2a1c8e1
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-03-21T04:16:13Z

    Merge branch 'master' of https://github.com/apache/spark into 
reduce-locations

commit ad7cb5377ae32f428ecd3b57cbb56fa1ecdb4856
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-06-04T19:45:59Z

    Merge branch 'master' of https://github.com/apache/spark into 
reduce-locations
    
    Conflicts:
        core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala
        core/src/test/scala/org/apache/spark/scheduler/DAGSchedulerSuite.scala

commit e7d5449238a6f674ab4500c01ac0be0bffd53832
Author: Shivaram Venkataraman <[email protected]>
Date:   2015-06-04T19:57:02Z

    Fix merge issues

----


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to