[YARN-7371] Added allocateRequestId in NMContainerStatus for recovery

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: yarn-native-services
• Target Version/s: yarn-native-services
• Hadoop Flags: Reviewed
• Assignee: Chandni Singh
• Reporter: Chandni Singh
• Created: 19/Oct/17 22:13
• Updated: 09/Nov/17 23:51
• Resolved: 05/Nov/17 06:15

Description

java.lang.NullPointerException
at org.apache.hadoop.yarn.service.ServiceScheduler.recoverComponents(ServiceScheduler.java:313)
at org.apache.hadoop.yarn.service.ServiceScheduler.serviceStart(ServiceScheduler.java:265)
at org.apache.hadoop.service.AbstractService.start(AbstractService.java:194)
at org.apache.hadoop.service.CompositeService.serviceStart(CompositeService.java:121)
at org.apache.hadoop.service.AbstractService.start(AbstractService.java:194)
at org.apache.hadoop.yarn.service.ServiceMaster.main(ServiceMaster.java:150)

Steps:
1. Stopped RM and then started it
2. Application was still running
3. Killed the ServiceMaster to check if it recovers
4. Next attempt failed with the above exception

Issue Links

• is related to: YARN-6744 Recover component information on YARN native services AM restart
• relates to: YARN-7470 TestYarnNativeServices testRecoverComponentsAfterRMRestart fails due to BindingException
• relates to: YARN-5447 Consider including allocationRequestId in NMContainerStatus to allow recovery in case of RM failover