You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-issues@hadoop.apache.org by "Jack Dintruff (JIRA)" <ji...@apache.org> on 2012/06/28 18:50:45 UTC
[jira] [Created] (HADOOP-8539) RMContainerImpl State Machine
Doesn't Handle RELEASED State from AM Properly
Jack Dintruff created HADOOP-8539:
-------------------------------------
Summary: RMContainerImpl State Machine Doesn't Handle RELEASED State from AM Properly
Key: HADOOP-8539
URL: https://issues.apache.org/jira/browse/HADOOP-8539
Project: Hadoop Common
Issue Type: Bug
Affects Versions: 0.23.3
Reporter: Jack Dintruff
Encountered this error with an RMContainerImpl unable to go from the RUNNING state to the RELEASED state. Upon further inspection, the state machine has no means of dealing with this and since the RELEASED state can be sent from the AM at any time, it should be able to handle this in both the RUNNING state and the ALLOCATED state.
--
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] (HADOOP-8539) RMContainerImpl State Machine
Doesn't Handle RELEASED State from AM Properly
Posted by "Jack Dintruff (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-8539?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jack Dintruff updated HADOOP-8539:
----------------------------------
Attachment: NewStateMachine.jpg
Here's the new state machine which can properly handle the RELEASED state if it's sent from the AM.
> RMContainerImpl State Machine Doesn't Handle RELEASED State from AM Properly
> ----------------------------------------------------------------------------
>
> Key: HADOOP-8539
> URL: https://issues.apache.org/jira/browse/HADOOP-8539
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 0.23.3
> Reporter: Jack Dintruff
> Attachments: NewStateMachine.jpg, OldStateMachine.jpg
>
>
> Encountered this error with an RMContainerImpl unable to go from the RUNNING state to the RELEASED state. Upon further inspection, the state machine has no means of dealing with this and since the RELEASED state can be sent from the AM at any time, it should be able to handle this in both the RUNNING state and the ALLOCATED state.
--
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] (HADOOP-8539) RMContainerImpl State Machine
Doesn't Handle RELEASED State from AM Properly
Posted by "Jack Dintruff (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-8539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13403214#comment-13403214 ]
Jack Dintruff commented on HADOOP-8539:
---------------------------------------
Here's the raw log message:
[ResourceManager Event Processor]2012-06-18 19:58:45,112 ERROR org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: Can't handle this event at current state
[IPC Server handler 25 on 8030]org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: RELEASED at RUNNING
at org.apache.hadoop.yarn.state.StateMachineFactory.doTransition(StateMachineFactory.java:301)
at org.apache.hadoop.yarn.state.StateMachineFactory.access$300(StateMachineFactory.java:43)
at org.apache.hadoop.yarn.state.StateMachineFactory$InternalStateMachine.doTransition(StateMachineFactory.java:443)
at org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl.handle(RMContainerImpl.java:200)
at org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl.handle(RMContainerImpl.java:44)
at org.apache.hadoop.yarn.server.resourcemanager.scheduler.SchedulerApp.containerCompleted(SchedulerApp.java:203)
at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.LeafQueue.completedContainer(LeafQueue.java:1312)
at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.completedContainer(CapacityScheduler.java:726)
at org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler.allocate(CapacityScheduler.java:463)
at org.apache.hadoop.yarn.server.resourcemanager.ApplicationMasterService.allocate(ApplicationMasterService.java:277)
at org.apache.hadoop.yarn.api.impl.pb.service.AMRMProtocolPBServiceImpl.allocate(AMRMProtocolPBServiceImpl.java:56)
at org.apache.hadoop.yarn.proto.AMRMProtocol$AMRMProtocolService$2.callBlockingMethod(AMRMProtocol.java:87)
at org.apache.hadoop.yarn.ipc.ProtoOverHadoopRpcEngine$Server.call(ProtoOverHadoopRpcEngine.java:352)
at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:1520)
at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:1516)
at java.security.AccessController.doPrivileged(Native Method)
at javax.security.auth.Subject.doAs(Subject.java:396)
at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1177)
at org.apache.hadoop.ipc.Server$Handler.run(Server.java:1514)
> RMContainerImpl State Machine Doesn't Handle RELEASED State from AM Properly
> ----------------------------------------------------------------------------
>
> Key: HADOOP-8539
> URL: https://issues.apache.org/jira/browse/HADOOP-8539
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 0.23.3
> Reporter: Jack Dintruff
>
> Encountered this error with an RMContainerImpl unable to go from the RUNNING state to the RELEASED state. Upon further inspection, the state machine has no means of dealing with this and since the RELEASED state can be sent from the AM at any time, it should be able to handle this in both the RUNNING state and the ALLOCATED state.
--
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] (HADOOP-8539) RMContainerImpl State Machine
Doesn't Handle RELEASED State from AM Properly
Posted by "Jack Dintruff (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-8539?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jack Dintruff updated HADOOP-8539:
----------------------------------
Attachment: OldStateMachine.jpg
Here's how things operate now. Notice that RUNNING and ALLOCATED are unable to handle the RELEASED state.
> RMContainerImpl State Machine Doesn't Handle RELEASED State from AM Properly
> ----------------------------------------------------------------------------
>
> Key: HADOOP-8539
> URL: https://issues.apache.org/jira/browse/HADOOP-8539
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 0.23.3
> Reporter: Jack Dintruff
> Attachments: NewStateMachine.jpg, OldStateMachine.jpg
>
>
> Encountered this error with an RMContainerImpl unable to go from the RUNNING state to the RELEASED state. Upon further inspection, the state machine has no means of dealing with this and since the RELEASED state can be sent from the AM at any time, it should be able to handle this in both the RUNNING state and the ALLOCATED state.
--
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] (HADOOP-8539) RMContainerImpl State Machine
Doesn't Handle RELEASED State from AM Properly
Posted by "Jack Dintruff (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HADOOP-8539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13403291#comment-13403291 ]
Jack Dintruff commented on HADOOP-8539:
---------------------------------------
The fix is only 4 lines long but there are no tests for RMContainerImpl so I'll write some state change tests for it to verify the state machine performs as intended.
> RMContainerImpl State Machine Doesn't Handle RELEASED State from AM Properly
> ----------------------------------------------------------------------------
>
> Key: HADOOP-8539
> URL: https://issues.apache.org/jira/browse/HADOOP-8539
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 0.23.3
> Reporter: Jack Dintruff
> Attachments: NewStateMachine.jpg, OldStateMachine.jpg
>
>
> Encountered this error with an RMContainerImpl unable to go from the RUNNING state to the RELEASED state. Upon further inspection, the state machine has no means of dealing with this and since the RELEASED state can be sent from the AM at any time, it should be able to handle this in both the RUNNING state and the ALLOCATED state.
--
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