[MAPREDUCE-2875] NM does not communicate Container crash to RM

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Duplicate
• Fix Version/s: 0.23.0
• Assignee: Siddharth Seth
• Reporter: Sharad Agarwal
• Created: 24/Aug/11 08:45
• Updated: 15/Nov/11 00:49
• Resolved: 19/Oct/11 12:02

Description

Faulty container crash detection code path in NodeManager.

Steps:
Run a job.
Kill the AM container in NM.

NM logs has:
org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: CONTAINER_KILLED_ON_REQUEST at RUNNING
at org.apache.hadoop.yarn.state.StateMachineFactory.doTransition(StateMachineFactory.java:297)
at org.apache.hadoop.yarn.state.StateMachineFactory.access$300(StateMachineFactory.java:39)
at org.apache.hadoop.yarn.state.StateMachineFactory$InternalStateMachine.doTransition(StateMachineFactory.java:439)
at org.apache.hadoop.yarn.server.nodemanager.containermanager.container.ContainerImpl.handle(ContainerImpl.java:685)

Issue Links

• is duplicated by: MAPREDUCE-3031 Job Client goes into infinite loop when we kill AM