[YARN-7565] Yarn service pre-maturely releases the container after AM restart

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.1.0
• Target Version/s: 3.1.0
• Hadoop Flags: Reviewed
• Assignee: Chandni Singh
• Reporter: Chandni Singh
• Created: 27/Nov/17 18:24
• Updated: 17/Feb/19 17:53
• Resolved: 15/Dec/17 01:11

Description

With YARN-6168, recovered containers can be reported to AM in response to the AM heartbeat.

Currently, the Service Master will release the containers, that are not reported in the AM registration response, immediately.

Instead, the master can wait for a configured amount of time for the containers to be recovered by RM. These containers are sent to AM in the heartbeat response. Once a container is not reported in the configured interval, it can be released by the master.

Issue Links

• is related to: YARN-6168 Restarted RM may not inform AM about all existing containers
• is related to: YARN-9195 RM Queue's pending container number might get decreased unexpectedly or even become negative once RM failover
• relates to: YARN-7958 ServiceMaster should only wait for recovery of containers with id that match the current application id