[
https://issues.apache.org/jira/browse/YARN-3464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14503588#comment-14503588
]
Karthik Kambatla commented on YARN-3464:
----------------------------------------
bq. So I'd look into changing the logic from "kill me when there's no more work
in my queue" to "kill me when my container is ready to be launched."
Thanks [~jlowe]. Yes, that would simplify the localizer-NM communication
significantly.
We could add the kill-container-localizer logic to either
{{LocalizedTransition}} or {{LaunchTransition}} in ContainerImpl. I prefer we
do it in the {{LocalizedTransition}}.
> Race condition in LocalizerRunner causes container localization timeout.
> ------------------------------------------------------------------------
>
> Key: YARN-3464
> URL: https://issues.apache.org/jira/browse/YARN-3464
> Project: Hadoop YARN
> Issue Type: Bug
> Components: nodemanager
> Reporter: zhihai xu
> Assignee: zhihai xu
> Priority: Critical
>
> Race condition in LocalizerRunner causes container localization timeout.
> Currently LocalizerRunner will kill the ContainerLocalizer when pending list
> for LocalizerResourceRequestEvent is empty.
> {code}
> } else if (pending.isEmpty()) {
> action = LocalizerAction.DIE;
> }
> {code}
> If a LocalizerResourceRequestEvent is added after LocalizerRunner kill the
> ContainerLocalizer due to empty pending list, this
> LocalizerResourceRequestEvent will never be handled.
> Without ContainerLocalizer, LocalizerRunner#update will never be called.
> The container will stay at LOCALIZING state, until the container is killed by
> AM due to TASK_TIMEOUT.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
