[YARN-8065] Application is failing when AM node is stopped

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Fix Version/s: 3.0.0
• Assignee: Bilwa S T
• Reporter: Bilwa S T
• Created: 23/Mar/18 08:41
• Updated: 29/Mar/18 12:42

Description

Configure yarn.scheduler.capacity.schedule-asynchronously.enable as true and 

yarn.resourcemanager.nodemanagers.heartbeat-interval-ms as 60000 .Run application and make AM node down. Application will fail.

If same node is picked up to launch AM attempt again, application can fail.More likely to occur with lesser number of nodes.