[MAPREDUCE-6898] TestKill.testKillTask is flaky

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Component/s: client, test
• Hadoop Flags: Reviewed
• Assignee: Peter Bacsko
• Reporter: Peter Bacsko
• Created: 14/Jun/17 15:11
• Updated: 16/Jun/17 20:13
• Resolved: 16/Jun/17 19:52

Description

TestKill.testKillTask() can fail if the async dispatcher thread is slower than the test's thread.

2017-05-26 11:43:26,532 INFO  [AsyncDispatcher event handler] impl.JobImpl (JobImpl.java:handle(1006)) - job_0_0000Job Transitioned from INITED to SETUP
Job State is : RUNNING
Job State is : RUNNING Waiting for state : SUCCEEDED   map progress : 0.0   reduce progress : 0.0
2017-05-26 11:43:26,538 INFO  [CommitterEvent Processor #0] commit.CommitterEventHandler (CommitterEventHandler.java:run(231)) - Processing the event EventType: JOB_SETUP
2017-05-26 11:43:26,540 INFO  [AsyncDispatcher event handler] impl.TaskImpl (TaskImpl.java:handle(661)) - task_0_0000_m_000000 Task Transitioned from NEW to KILLED
2017-05-26 11:43:26,540 ERROR [AsyncDispatcher event handler] impl.JobImpl (JobImpl.java:handle(998)) - Can't handle this event at current state
org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: JOB_TASK_COMPLETED at SETUP
	at org.apache.hadoop.yarn.state.StateMachineFactory.doTransition(StateMachineFactory.java:305)
	at org.apache.hadoop.yarn.state.StateMachineFactory.access$300(StateMachineFactory.java:46)
	at org.apache.hadoop.yarn.state.StateMachineFactory$InternalStateMachine.doTransition(StateMachineFactory.java:448)
	at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl.handle(JobImpl.java:996)
	at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl.handle(JobImpl.java:138)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$JobEventDispatcher.handle(MRAppMaster.java:1366)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$JobEventDispatcher.handle(MRAppMaster.java:1362)
	at org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:182)
	at org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:109)
	at java.lang.Thread.run(Thread.java:745)
2017-05-26 11:43:26,541 INFO  [AsyncDispatcher event handler] impl.JobImpl (JobImpl.java:handle(1006)) - job_0_0000Job Transitioned from SETUP to ERROR
2017-05-26 11:43:26,542 INFO  [AsyncDispatcher event handler] app.MRAppMaster (MRAppMaster.java:serviceStop(978)) - Skipping cleaning up the staging dir. assuming AM will be retried.

We have to wait until the job's internal state is JobInternalState.RUNNING and not JobInternalState.SETUP.

Issue Links

• duplicates: MAPREDUCE-6815 Fix flaky TestKill.testKillTask()