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 "Robert Joseph Evans (JIRA)" <ji...@apache.org> on 2012/08/30 16:40:07 UTC
[jira] [Created] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Robert Joseph Evans created MAPREDUCE-4611:
----------------------------------------------
Summary: MR AM dies badly when Node is decomissioned
Key: MAPREDUCE-4611
URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
Project: Hadoop Map/Reduce
Issue Type: Bug
Affects Versions: 2.0.0-alpha, 0.23.3, 3.0.0
Reporter: Robert Joseph Evans
Assignee: Robert Joseph Evans
The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
I propose that in the short term the AM will only clean up if
# The process has been asked by the client to exit (kill)
# The process job has finished cleanly and is exiting already
# This is that last retry of the AM retries.
The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Updated] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Thomas Graves (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Thomas Graves updated MAPREDUCE-4611:
-------------------------------------
Resolution: Fixed
Fix Version/s: 2.2.0-alpha
3.0.0
0.23.3
Status: Resolved (was: Patch Available)
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Closed] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Arun C Murthy (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Arun C Murthy closed MAPREDUCE-4611.
------------------------------------
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 2.0.2-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446376#comment-13446376 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Mapreduce-trunk-Commit #2691 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk-Commit/2691/])
MAPREDUCE-4611. MR AM dies badly when Node is decommissioned (Robert Evans via tgraves) (Revision 1379599)
Result = FAILURE
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379599
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Thomas Graves (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446347#comment-13446347 ]
Thomas Graves commented on MAPREDUCE-4611:
------------------------------------------
+1 lgtm. Thanks Bobby!
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Updated] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Robert Joseph Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Robert Joseph Evans updated MAPREDUCE-4611:
-------------------------------------------
Priority: Critical (was: Major)
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hadoop QA (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13445229#comment-13445229 ]
Hadoop QA commented on MAPREDUCE-4611:
--------------------------------------
+1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12543142/MR-4611.txt
against trunk revision .
+1 @author. The patch does not contain any @author tags.
+1 tests included. The patch appears to include 2 new or modified test files.
+1 javac. The applied patch does not increase the total number of javac compiler warnings.
+1 javadoc. The javadoc tool did not generate any warning messages.
+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 hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app.
+1 contrib tests. The patch passed contrib unit tests.
Test results: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/2792//testReport/
Console output: https://builds.apache.org/job/PreCommit-MAPREDUCE-Build/2792//console
This message is automatically generated.
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Updated] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Robert Joseph Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Robert Joseph Evans updated MAPREDUCE-4611:
-------------------------------------------
Attachment: MR-4611.txt
This patch makes the changes to only cleanup when the job has finished, or when it is the last retry for the AM.
I have manually tested this in addition to adding in the unit tests.
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446683#comment-13446683 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Hdfs-trunk #1152 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/1152/])
MAPREDUCE-4611. MR AM dies badly when Node is decommissioned (Robert Evans via tgraves) (Revision 1379599)
Result = SUCCESS
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379599
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446668#comment-13446668 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Hdfs-0.23-Build #361 (See [https://builds.apache.org/job/Hadoop-Hdfs-0.23-Build/361/])
merge -r 1379601:1379602 from branch-2. FIXES: MAPREDUCE-4611 (Revision 1379604)
Result = UNSTABLE
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379604
Files :
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/branches/branch-0.23/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Updated] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Robert Joseph Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Robert Joseph Evans updated MAPREDUCE-4611:
-------------------------------------------
Status: Patch Available (was: Open)
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 2.0.0-alpha, 0.23.3, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446368#comment-13446368 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Common-trunk-Commit #2666 (See [https://builds.apache.org/job/Hadoop-Common-trunk-Commit/2666/])
MAPREDUCE-4611. MR AM dies badly when Node is decommissioned (Robert Evans via tgraves) (Revision 1379599)
Result = SUCCESS
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379599
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446711#comment-13446711 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Mapreduce-trunk #1183 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/1183/])
MAPREDUCE-4611. MR AM dies badly when Node is decommissioned (Robert Evans via tgraves) (Revision 1379599)
Result = SUCCESS
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379599
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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
[jira] [Commented] (MAPREDUCE-4611) MR AM dies badly when Node is
decomissioned
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MAPREDUCE-4611?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13446363#comment-13446363 ]
Hudson commented on MAPREDUCE-4611:
-----------------------------------
Integrated in Hadoop-Hdfs-trunk-Commit #2729 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk-Commit/2729/])
MAPREDUCE-4611. MR AM dies badly when Node is decommissioned (Robert Evans via tgraves) (Revision 1379599)
Result = SUCCESS
tgraves : http://svn.apache.org/viewcvs.cgi/?root=Apache-SVN&view=rev&rev=1379599
Files :
* /hadoop/common/trunk/hadoop-mapreduce-project/CHANGES.txt
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/jobhistory/JobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/MRAppMaster.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/main/java/org/apache/hadoop/mapreduce/v2/app/rm/RMCommunicator.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/jobhistory/TestJobHistoryEventHandler.java
* /hadoop/common/trunk/hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-app/src/test/java/org/apache/hadoop/mapreduce/v2/app/TestStagingCleanup.java
> MR AM dies badly when Node is decomissioned
> -------------------------------------------
>
> Key: MAPREDUCE-4611
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4611
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Affects Versions: 0.23.3, 2.0.0-alpha, 3.0.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Priority: Critical
> Fix For: 0.23.3, 3.0.0, 2.2.0-alpha
>
> Attachments: MR-4611.txt
>
>
> The MR AM always thinks that it is being killed by the RM when it gets a kill signal and it has not finished processing yet. In reality the RM kill signal is only sent when the client cannot communicate directly with the AM, which probably means that the AM is in a bad state already. The much more common case is that the node is marked as unhealthy or decomissioned.
> I propose that in the short term the AM will only clean up if
> # The process has been asked by the client to exit (kill)
> # The process job has finished cleanly and is exiting already
> # This is that last retry of the AM retries.
> The downside here is that the .staging directory will be leaked and the job will not show up in the history server on an kill from the RM in some cases.
> At least until the full set of AM cleanup issues can be addressed, probably as part of MAPREDUCE-4428
--
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