[
https://issues.apache.org/jira/browse/MAPREDUCE-4784?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13494803#comment-13494803
]
Jason Lowe commented on MAPREDUCE-4784:
---------------------------------------
Looking at the test output when it fails, there's an invalid state transition:
{noformat}
2012-11-10 23:19:07,665 INFO [AsyncDispatcher event handler]
impl.TaskAttemptImpl (TaskAttemptImpl.java:handle(993)) -
attempt_0_0000_m_000000_1 TaskAttempt Transitioned from NEW to UNASSIGNED
TaskAttempt State is : FAILED
TaskAttempt State is : STARTING Waiting for state : FAILED progress : 0.0
2012-11-10 23:19:07,667 ERROR [AsyncDispatcher event handler]
impl.TaskAttemptImpl (TaskAttemptImpl.java:handle(984)) - Can't handle this
event at current state for attempt_0_0000_m_000000_1
org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event:
TA_CONTAINER_LAUNCH_FAILED at UNASSIGNED
at
org.apache.hadoop.yarn.state.StateMachineFactory.doTransition(StateMachineFactory.java:302)
at
org.apache.hadoop.yarn.state.StateMachineFactory.access$300(StateMachineFactory.java:43)
at
org.apache.hadoop.yarn.state.StateMachineFactory$InternalStateMachine.doTransition(StateMachineFactory.java:445)
at
org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl.handle(TaskAttemptImpl.java:982)
at
org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl.handle(TaskAttemptImpl.java:1)
at
org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:996)
at
org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:1)
at
org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:128)
at
org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:77)
at java.lang.Thread.run(Thread.java:662)
{noformat}
I think the problem occurs because the test is trying to inject
TA_CONTAINER_LAUNCH_FAILED into the attempt state machine asynchronously.
Sometimes that event arrives at the appropriate state and the test passes,
sometimes it arrives at an inappropriate state and the test fails.
> TestRecovery occasionally fails
> -------------------------------
>
> Key: MAPREDUCE-4784
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4784
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2, test
> Affects Versions: 2.0.3-alpha
> Reporter: Jason Lowe
>
> TestRecovery is occasionally failing with this error:
> {noformat}
> testCrashed(org.apache.hadoop.mapreduce.v2.app.TestRecovery): TaskAttempt
> state is not correct (timedout) expected:<FAILED> but was:<STARTING>
> {noformat}
--
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
