[ 
https://issues.apache.org/jira/browse/YARN-5620?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15483158#comment-15483158
 ] 

Arun Suresh commented on YARN-5620:
-----------------------------------

Thanks [~jianhe]..

Will update patch shortly with your suggestions. But, with regard to this :
bq. Looks like we don’t need the killedForReInitialization flag in 
ContainerLaunch, because container_killed event can already be distinguished 
based on whether container is at Reinit or running state.
This we actually do need the flag.. Since even while the container is 
re-initing (while resource is localizing but before the container_cleanup 
signal is sent), the container should be killable explicitly via an external 
signal. The {{ContainerLaunch}} should be able to know the difference and after 
killing, it should either {{ContainerEventType.CONTAINER_KILLED_ON_REQUEST}} or 
{{ContainerEventType.CONTAINER_KILLED_FOR_REINIT}}. That reminds me... I have 
to add the following transition as well:
{code}
.addTransition(ContainerState.REINITIALIZING,
        ContainerState.EXITED_WITH_FAILURE,
        ContainerEventType.CONTAINER_KILLED_ON_REQUEST,
        new KilledExternallyTransition())
{code}

> Core changes in NodeManager to support for upgrade and rollback of Containers
> -----------------------------------------------------------------------------
>
>                 Key: YARN-5620
>                 URL: https://issues.apache.org/jira/browse/YARN-5620
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>            Reporter: Arun Suresh
>            Assignee: Arun Suresh
>         Attachments: YARN-5620.001.patch, YARN-5620.002.patch, 
> YARN-5620.003.patch, YARN-5620.004.patch, YARN-5620.005.patch, 
> YARN-5620.006.patch, YARN-5620.007.patch, YARN-5620.008.patch
>
>
> JIRA proposes to modify the ContainerManager (and other core classes) to 
> support upgrade of a running container with a new {{ContainerLaunchContext}} 
> as well as the ability to rollback the upgrade if the container is not able 
> to restart using the new launch Context. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org

Reply via email to