[
https://issues.apache.org/jira/browse/YARN-2917?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14241755#comment-14241755
]
Jian He commented on YARN-2917:
-------------------------------
agree that if RM is abruptly killed, no events will be drained. In fact, things
continue to work irrespective of whether events are drained or not either when
RM is transiting or restarting. Draining events is just one optimization to
process the remaining events before RM is shutting down. Similarly in this
jira, there's no correctness concern whether to drain or not, it's just a
optimization, unless I miss some other side effects. do you think if there's
any ?
> Potential deadlock in AsyncDispatcher when system.exit called in
> AsyncDispatcher#dispatch and AsyscDispatcher#serviceStop from shutdown hook
> --------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: YARN-2917
> URL: https://issues.apache.org/jira/browse/YARN-2917
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager
> Reporter: Rohith
> Assignee: Rohith
> Priority: Critical
> Attachments: 0001-YARN-2917.patch
>
>
> I encoutered scenario where RM hanged while shutting down and keep on logging
> {{2014-12-03 19:32:44,283 INFO org.apache.hadoop.yarn.event.AsyncDispatcher:
> Waiting for AsyncDispatcher to drain.}}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
