[
https://issues.apache.org/jira/browse/MAPREDUCE-3260?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13135232#comment-13135232
]
Todd Lipcon commented on MAPREDUCE-3260:
----------------------------------------
There seem to be some reducers stuck in KILLING state on some of the nodes. The
only non-daemon thread is:
{code}
"main" prio=10 tid=0x0000000046f7b800 nid=0x3774 waiting on condition
[0x000000004033e000]
java.lang.Thread.State: TIMED_WAITING (sleeping)
at java.lang.Thread.sleep(Native Method)
at java.lang.Thread.sleep(Thread.java:298)
at java.util.concurrent.TimeUnit.sleep(TimeUnit.java:328)
at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:117)
{code}
Logs in the NM show the following which looks like a race:
{code}
2011-10-25 00:44:42,842 INFO
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container:
Processing container_1319528200416_0004_01_002409 of type KILL_CONTAINER
2011-10-25 00:44:42,842 INFO
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container:
Container container_1319528200416_0004_01_002409 transitioned from LOCALIZED to
KILLING
2011-10-25 00:44:42,842 INFO
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container:
Processing container_1319528200416_0004_01_002409 of type CONTAINER_LAUNCHED
2011-10-25 00:44:42,860 WARN
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container:
Can't handle this event at current state: Current: [KILLING], eventType:
[CONTAINER_LAUNCHED]
org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
CONTAINER_LAUNCHED at KILLING
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.nodemanager.containermanager.container.ContainerImpl.handle(ContainerImpl.java:803)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.ContainerImpl.handle(ContainerImpl.java:70)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.ContainerManagerImpl$ContainerEventDispatcher.handle(ContainerManagerImpl.java:373)
at
org.apache.hadoop.yarn.server.nodemanager.containermanager.ContainerManagerImpl$ContainerEventDispatcher.handle(ContainerManagerImpl.java:366)
at
org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:116)
at
org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:75)
at java.lang.Thread.run(Thread.java:619)
2011-10-25 00:44:42,879 INFO
org.apache.hadoop.yarn.server.nodemanager.containermanager.container.Container:
Container container_1319528200416_0004_01_002409 transitioned from KILLING to
null
{code}
> Yarn app stuck in KILL_WAIT state
> ---------------------------------
>
> Key: MAPREDUCE-3260
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3260
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, resourcemanager
> Affects Versions: 0.23.0
> Reporter: Todd Lipcon
> Priority: Critical
>
> Last night I killed an MR2 app using "hadoop job -kill". This morning I
> noticed it's still running, but in "KILL_WAIT" state with no tasks running.
--
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
