[MAPREDUCE-7240] Exception ' Invalid event: TA_TOO_MANY_FETCH_FAILURE at SUCCESS_FINISHING_CONTAINER' cause job error

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 2.8.2
• Fix Version/s: 3.3.0, 3.1.4, 3.2.2, 2.10.1
• Labels: Reviewed applicationmaster mrv2
• Hadoop Flags: Reviewed
• Assignee: luhuachao
• Reporter: luhuachao
• Created: 12/Sep/19 03:16
• Updated: 11/Jun/21 06:52
• Resolved: 27/Nov/19 12:55

Description

log in appmaster

2019-09-03 17:18:43,090 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Too many fetch-failures for output of task attempt: attempt_1566552310686_260041_m_000052_0 ... raising fetch failure to map
2019-09-03 17:18:43,091 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Too many fetch-failures for output of task attempt: attempt_1566552310686_260041_m_000049_0 ... raising fetch failure to map
2019-09-03 17:18:43,091 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Too many fetch-failures for output of task attempt: attempt_1566552310686_260041_m_000051_0 ... raising fetch failure to map
2019-09-03 17:18:43,091 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Too many fetch-failures for output of task attempt: attempt_1566552310686_260041_m_000050_0 ... raising fetch failure to map
2019-09-03 17:18:43,091 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: Too many fetch-failures for output of task attempt: attempt_1566552310686_260041_m_000053_0 ... raising fetch failure to map
2019-09-03 17:18:43,092 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1566552310686_260041_m_000052_0 transitioned from state SUCCEEDED to FAILED, event type is TA_TOO_MANY_FETCH_FAILURE and nodeId=yarn095:45454
2019-09-03 17:18:43,092 ERROR [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Can't handle this event at current state for attempt_1566552310686_260041_m_000049_0
org.apache.hadoop.yarn.state.InvalidStateTransitionException: Invalid event: TA_TOO_MANY_FETCH_FAILURE at SUCCESS_FINISHING_CONTAINER
	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.TaskAttemptImpl.handle(TaskAttemptImpl.java:1206)
	at org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl.handle(TaskAttemptImpl.java:146)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:1458)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:1450)
	at org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:184)
	at org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:110)
	at java.lang.Thread.run(Thread.java:745)
2019-09-03 17:18:43,093 ERROR [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: Can't handle this event at current state for attempt_1566552310686_260041_m_000051_0
org.apache.hadoop.yarn.state.InvalidStateTransitionException: Invalid event: TA_TOO_MANY_FETCH_FAILURE at SUCCESS_FINISHING_CONTAINER
	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.TaskAttemptImpl.handle(TaskAttemptImpl.java:1206)
	at org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl.handle(TaskAttemptImpl.java:146)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:1458)
	at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$TaskAttemptEventDispatcher.handle(MRAppMaster.java:1450)
	at org.apache.hadoop.yarn.event.AsyncDispatcher.dispatch(AsyncDispatcher.java:184)
	at org.apache.hadoop.yarn.event.AsyncDispatcher$1.run(AsyncDispatcher.java:110)
	at java.lang.Thread.run(Thread.java:745)
2019-09-03 17:18:43,093 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1566552310686_260041_m_000050_0 transitioned from state SUCCEEDED to FAILED, event type is TA_TOO_MANY_FETCH_FAILURE and nodeId=yarn095:45454
2019-09-03 17:18:43,093 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskAttemptImpl: attempt_1566552310686_260041_m_000053_0 transitioned from state SUCCEEDED to FAILED, event type is TA_TOO_MANY_FETCH_FAILURE and nodeId=yarn095:45454
2019-09-03 17:18:43,094 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.TaskImpl: task_1566552310686_260041_m_000052 Task Transitioned from SUCCEEDED to SCHEDULED
2019-09-03 17:18:43,096 FATAL [IPC Server handler 27 on 35972] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Task: attempt_1566552310686_260041_r_000005_0 - exited : org.apache.hadoop.mapreduce.task.reduce.Shuffle$ShuffleError: error in shuffle in fetcher#22
	at org.apache.hadoop.mapreduce.task.reduce.Shuffle.run(Shuffle.java:134)
	at org.apache.hadoop.mapred.ReduceTask.run(ReduceTask.java:376)
	at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:175)
	at java.security.AccessController.doPrivileged(Native Method)
	at javax.security.auth.Subject.doAs(Subject.java:422)
	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1961)
	at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:169)
Caused by: java.io.IOException: Exceeded MAX_FAILED_UNIQUE_FETCHES; bailing-out.
	at org.apache.hadoop.mapreduce.task.reduce.ShuffleSchedulerImpl.checkReducerHealth(ShuffleSchedulerImpl.java:367)
	at org.apache.hadoop.mapreduce.task.reduce.ShuffleSchedulerImpl.copyFailed(ShuffleSchedulerImpl.java:289)
	at org.apache.hadoop.mapreduce.task.reduce.Fetcher.copyFromHost(Fetcher.java:355)
	at org.apache.hadoop.mapreduce.task.reduce.Fetcher.run(Fetcher.java:193)

2019-09-03 17:18:43,096 INFO [IPC Server handler 27 on 35972] org.apache.hadoop.mapred.TaskAttemptListenerImpl: Diagnostics report from attempt_1566552310686_260041_r_000005_0: Error: org.apache.hadoop.mapreduce.task.reduce.Shuffle$ShuffleError: error in shuffle in fetcher#22
	at org.apache.hadoop.mapreduce.task.reduce.Shuffle.run(Shuffle.java:134)
	at org.apache.hadoop.mapred.ReduceTask.run(ReduceTask.java:376)
	at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:175)
	at java.security.AccessController.doPrivileged(Native Method)
	at javax.security.auth.Subject.doAs(Subject.java:422)
	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1961)
	at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:169)
Caused by: java.io.IOException: Exceeded MAX_FAILED_UNIQUE_FETCHES; bailing-out.
	at org.apache.hadoop.mapreduce.task.reduce.ShuffleSchedulerImpl.checkReducerHealth(ShuffleSchedulerImpl.java:367)
	at org.apache.hadoop.mapreduce.task.reduce.ShuffleSchedulerImpl.copyFailed(ShuffleSchedulerImpl.java:289)
	at org.apache.hadoop.mapreduce.task.reduce.Fetcher.copyFromHost(Fetcher.java:355)
	at org.apache.hadoop.mapreduce.task.reduce.Fetcher.run(Fetcher.java:193)

2019-09-03 17:18:43,097 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl: job_1566552310686_260041Job Transitioned from RUNNING to ERROR
2019-09-03 17:18:43,099 INFO [AsyncDispatcher event handler] org.apache.hadoop.mapreduce.v2.app.job

  

nodemanager's  log is like same with log in  MAPREDUCE-6869.

the code in TaskAttemptImpl indicate the Invalid event: TA_TOO_MANY_FETCH_FAILURE at SUCCESS_FINISHING_CONTAINER cause the job state turn into error; what i confused is 

1. what cause the appmater  handle the TA_TOO_MANY_FETCH_FAILURE  event on SUCCESS_FINISHING_CONTAINER，illegal event on this state.  but some other can successfully transitioned from state SUCCEEDED to FAILED on TA_TOO_MANY_FETCH_FAILURE  event.
2. restart the nodemanager would solve the error in nm; the shuffle error would fix too. what cause this phenomenon.

Correct me if I am wrong.

Issue Links

• relates to: MAPREDUCE-7249 Invalid event TA_TOO_MANY_FETCH_FAILURE at SUCCESS_CONTAINER_CLEANUP causes job failure
• relates to: MAPREDUCE-7264 overall reduction of ApplicationMaster exit because of unhandled TA_TOO_MANY_FETCH_FAILURE event
• links to: GitHub Pull Request #1618
• links to: GitHub Pull Request #1674