You are viewing a plain text version of this content. The canonical link for it is here.
Posted to mapreduce-issues@hadoop.apache.org by "Arun C Murthy (Created) (JIRA)" <ji...@apache.org> on 2012/01/08 07:14:39 UTC
[jira] [Created] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
CapacityScheduler should be more conservative assigning off-switch requests
---------------------------------------------------------------------------
Key: MAPREDUCE-3641
URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
Project: Hadoop Map/Reduce
Issue Type: Bug
Components: mrv2, scheduler
Affects Versions: 0.23.0
Reporter: Arun C Murthy
Assignee: Arun C Murthy
Priority: Blocker
In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Status: Open (was: Patch Available)
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187650#comment-13187650 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Mapreduce-0.23-Build #163 (See [https://builds.apache.org/job/Hadoop-Mapreduce-0.23-Build/163/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
svn merge --ignore-ancestry -c 1232182 ../../trunk/
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232183
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187236#comment-13187236 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Hdfs-trunk-Commit #1622 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk-Commit/1622/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232182
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Attachment: MAPREDUCE-3641.patch
Ok, tested patch on secure cluster.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Attachment: MAPREDUCE-3641.patch
Straight-fwd patch, testing...
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187257#comment-13187257 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Mapreduce-0.23-Commit #396 (See [https://builds.apache.org/job/Hadoop-Mapreduce-0.23-Commit/396/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
svn merge --ignore-ancestry -c 1232182 ../../trunk/
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232183
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Vinod Kumar Vavilapalli (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vinod Kumar Vavilapalli updated MAPREDUCE-3641:
-----------------------------------------------
Resolution: Fixed
Fix Version/s: 0.23.1
Release Note: Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration.
Hadoop Flags: Reviewed
Status: Resolved (was: Patch Available)
I just committed this to trunk and branch-023. Thanks Arun!
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Status: Patch Available (was: Open)
Writing unit tests now...
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187618#comment-13187618 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Hdfs-trunk #928 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/928/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232182
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Attachment: MAPREDUCE-3641.patch
Added unit tests, ready for review/commit.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187238#comment-13187238 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Common-trunk-Commit #1549 (See [https://builds.apache.org/job/Hadoop-Common-trunk-Commit/1549/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232182
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Vinod Kumar Vavilapalli (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187096#comment-13187096 ]
Vinod Kumar Vavilapalli commented on MAPREDUCE-3641:
----------------------------------------------------
Looking through the patch.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187267#comment-13187267 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Mapreduce-trunk-Commit #1567 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk-Commit/1567/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232182
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Status: Open (was: Patch Available)
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Vinod Kumar Vavilapalli (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187226#comment-13187226 ]
Vinod Kumar Vavilapalli commented on MAPREDUCE-3641:
----------------------------------------------------
+1. This looks good.
I also have an offline confirmation from [~sseth] that this runs fine on a 350 node cluster and improves performance for a few benchmarks.
May be good to fix FifoScheduler also to have this, but we can keep that separate.
I am going to push this in.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Attachment: MAPREDUCE-3641.patch
Update patch with some bug-fixes and some extra debug info (not for commit).
I'm still hunting a NPE down:
{noformat}
12/01/12 23:47:38 ERROR resourcemanager.ResourceManager: Error in handling event type NODE_UPDATE to the schedulerjava.lang.NullPointerException at org.apache.hadoop.yarn.server.resourcemanager.scheduler.SchedulerApp.unreserve(SchedulerApp.java:392)
at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.unreserve(LeafQueue.java:1217) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainer(LeafQueue.java:1158) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignRackLocalContainers(LeafQueue.java:1027) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainersOnNode(LeafQueue.java:989) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignReservedContainer(LeafQueue.java:821) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.assignContainers(LeafQueue.java:721) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.nodeUpdate(CapacityScheduler.java:568) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.handle(CapacityScheduler.java:615) at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.handle(CapacityScheduler.java:77) at org.apache.hadoop.yarn.server.resourcemanager.ResourceManager$SchedulerEventDispatcher$EventProcessor.run(ResourceManager.java:295) at java.lang.Thread.run(Thread.java:662)
{noformat}
I'm also seeing negative container counts for the queue etc.
Need to dig some more.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hadoop QA (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13186479#comment-13186479 ]
Hadoop QA commented on MAPREDUCE-3641:
--------------------------------------
+1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12510620/MAPREDUCE-3641.patch
against trunk revision .
+1 @author. The patch does not contain any @author tags.
+1 tests included. The patch appears to include 6 new or modified tests.
+1 javadoc. The javadoc tool did not generate any warning messages.
+1 javac. The applied patch does not increase the total number of javac compiler warnings.
+1 eclipse:eclipse. The patch built with eclipse:eclipse.
+1 findbugs. The patch does not introduce any new Findbugs (version 1.3.9) warnings.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
+1 core tests. The patch passed unit tests in .
+1 contrib tests. The patch passed contrib unit tests.
Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1612//testReport/
Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1612//console
This message is automatically generated.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187663#comment-13187663 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Mapreduce-trunk #961 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/961/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232182
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Vinod Kumar Vavilapalli (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vinod Kumar Vavilapalli updated MAPREDUCE-3641:
-----------------------------------------------
Issue Type: Sub-task (was: Bug)
Parent: MAPREDUCE-3561
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187243#comment-13187243 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Hdfs-0.23-Commit #373 (See [https://builds.apache.org/job/Hadoop-Hdfs-0.23-Commit/373/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
svn merge --ignore-ancestry -c 1232182 ../../trunk/
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232183
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Status: Patch Available (was: Open)
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187624#comment-13187624 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Hdfs-0.23-Build #141 (See [https://builds.apache.org/job/Hadoop-Hdfs-0.23-Build/141/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
svn merge --ignore-ancestry -c 1232182 ../../trunk/
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232183
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hadoop QA (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13182212#comment-13182212 ]
Hadoop QA commented on MAPREDUCE-3641:
--------------------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12509836/MAPREDUCE-3641.patch
against trunk revision .
+1 @author. The patch does not contain any @author tags.
-1 tests included. The patch doesn't appear to include any new or modified tests.
Please justify why no new tests are needed for this patch.
Also please list what manual steps were performed to verify this patch.
+1 javadoc. The javadoc tool did not generate any warning messages.
+1 javac. The applied patch does not increase the total number of javac compiler warnings.
+1 eclipse:eclipse. The patch built with eclipse:eclipse.
-1 findbugs. The patch appears to introduce 1 new Findbugs (version 1.3.9) warnings.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
-1 core tests. The patch failed these unit tests:
org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.TestParentQueue
+1 contrib tests. The patch passed contrib unit tests.
Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1567//testReport/
Findbugs warnings: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1567//artifact/trunk/hadoop-mapreduce-project/patchprocess/newPatchFindbugsWarningshadoop-mapreduce-client-app.html
Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1567//console
This message is automatically generated.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hudson (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13187237#comment-13187237 ]
Hudson commented on MAPREDUCE-3641:
-----------------------------------
Integrated in Hadoop-Common-0.23-Commit #383 (See [https://builds.apache.org/job/Hadoop-Common-0.23-Commit/383/])
MAPREDUCE-3641. Making CapacityScheduler more conservative so as to assign only one off-switch container in a single scheduling iteration. Contributed by Arun C Murthy.
svn merge --ignore-ancestry -c 1232182 ../../trunk/
vinodkv : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1232183
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/resource/Resources.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSAssignment.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/CSQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/LeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/main/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/ParentQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestLeafQueue.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/src/test/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/capacity/TestParentQueue.java
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Fix For: 0.23.1
>
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Commented] (MAPREDUCE-3641) CapacityScheduler should be
more conservative assigning off-switch requests
Posted by "Hadoop QA (Commented) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13186400#comment-13186400 ]
Hadoop QA commented on MAPREDUCE-3641:
--------------------------------------
-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12510612/MAPREDUCE-3641.patch
against trunk revision .
+1 @author. The patch does not contain any @author tags.
-1 tests included. The patch doesn't appear to include any new or modified tests.
Please justify why no new tests are needed for this patch.
Also please list what manual steps were performed to verify this patch.
+1 javadoc. The javadoc tool did not generate any warning messages.
+1 javac. The applied patch does not increase the total number of javac compiler warnings.
+1 eclipse:eclipse. The patch built with eclipse:eclipse.
+1 findbugs. The patch does not introduce any new Findbugs (version 1.3.9) warnings.
+1 release audit. The applied patch does not increase the total number of release audit warnings.
-1 core tests. The patch failed these unit tests:
org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.TestParentQueue
+1 contrib tests. The patch passed contrib unit tests.
Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1611//testReport/
Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/1611//console
This message is automatically generated.
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Sub-task
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch, MAPREDUCE-3641.patch, MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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
[jira] [Updated] (MAPREDUCE-3641) CapacityScheduler should be more
conservative assigning off-switch requests
Posted by "Arun C Murthy (Updated) (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-3641?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy updated MAPREDUCE-3641:
-------------------------------------
Status: Patch Available (was: Open)
> CapacityScheduler should be more conservative assigning off-switch requests
> ---------------------------------------------------------------------------
>
> Key: MAPREDUCE-3641
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3641
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, scheduler
> Affects Versions: 0.23.0
> Reporter: Arun C Murthy
> Assignee: Arun C Murthy
> Priority: Blocker
> Attachments: MAPREDUCE-3641.patch
>
>
> In hadoop-1, the CS is very conservative handing out off-switch assignments, we need to do the same in YARN.
> We noticed performance regressions due to this, particularly for reduces.
--
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