[YARN-3223] Resource update during NM graceful decommission

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.7.1
• Fix Version/s: 2.8.0, 3.0.0-alpha1
• Component/s: graceful, nodemanager, resourcemanager
• Target Version/s: 2.8.0
• Hadoop Flags: Reviewed
• Assignee: Brook Zhou
• Reporter: Junping Du
• Created: 19/Feb/15 15:00
• Updated: 30/Aug/16 01:27
• Resolved: 23/Feb/16 11:33

Description

During NM graceful decommission, we should handle resource update properly, include: make RMNode keep track of old resource for possible rollback, keep available resource to 0 and used resource get updated when
container finished.

Issue Links

• blocks: YARN-4677 RMNodeResourceUpdateEvent update from scheduler can lead to race condition
• is blocked by: YARN-3212 RMNode State Transition Update with DECOMMISSIONING state