[
https://issues.apache.org/jira/browse/YARN-816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13706653#comment-13706653
]
Abhishek Kapoor commented on YARN-816:
--------------------------------------
Preemption is one of the case where container can be killed while application
is still running.
We can take inspiration from CPU scheduling algorithms done in OS.
Also if application is preempted we can provide a way to let app know that if
it is going to get preempted and during recovery we aware app then it was bring
preempted.
Probably a event fired to app letting it know what is going(preempt) to happen
and what has happened(preempted).
Sorry if it sounds confusing....
I am open for discussion
> Implement AM recovery for distributed shell
> -------------------------------------------
>
> Key: YARN-816
> URL: https://issues.apache.org/jira/browse/YARN-816
> Project: Hadoop YARN
> Issue Type: Improvement
> Components: applications/distributed-shell
> Reporter: Vinod Kumar Vavilapalli
>
> Simple recovery to just continue from where it left off is a good start.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
