[MAPREDUCE-6659] Mapreduce App master waits long to kill containers on lost nodes.

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 2.6.0
• Component/s: mr-am
• Assignee: Nicolas Fraison
• Reporter: Laxman
• Created: 24/Mar/16 12:01
• Updated: 08/Nov/17 15:05

Description

MR Application master waits for very long time to cleanup and relaunch the tasks on lost nodes. Wait time is actually 2.5 hours (ipc.client.connect.max.retries * ipc.client.connect.max.retries.on.timeouts * ipc.client.connect.timeout = 10 * 45 * 20 = 9000 seconds = 2.5 hours)

Some similar issue related in RM-AM rpc protocol is fixed in YARN-3809.
As fixed in YARN-3809, we may need to introduce new configurations to control this RPC retry behavior.

Also, I feel this total retry time should honor and capped maximum to global task time out (mapreduce.task.timeout = 600000 default)

Issue Links

• is duplicated by: MAPREDUCE-6982 Containers on lost nodes are considered failed after a too long time.
• relates to: YARN-3809 Failed to launch new attempts because ApplicationMasterLauncher's threads all hang