[
https://issues.apache.org/jira/browse/YARN-6059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15862172#comment-15862172
]
Konstantinos Karanasos commented on YARN-6059:
----------------------------------------------
Thank you for the patch, [~hrsharma].
I gave a first look, here are some comments:
# The paused events are not stored in the {{NMStateStore}}. You need to add
that in the {{ContainerScheduler}}, as we do for the QUEUED containers, e.g.,
with {{this.context.getNMStateStore().storeContainerQueued}}.
# You need to make sure that, when a PAUSED container is relaunched, we add a
new entry to the {{NMStateStore}} to mark it as RUNNING again.
# In the {{ContainersLauncher}}, I think you should add to the {{running}}
containers, and then:
# In the {{RecoverPausedContainerLaunch}}, you should raise a ContainerEvent to
indicate that the container finished its execution, like we do with the other
{{*ContainerLaunch}} classes, with something like the following:
{noformat}
dispatcher.getEventHandler().handle(
new ContainerEvent(containerId,
ContainerEventType.CONTAINER_EXITED_WITH_SUCCESS));
{noformat}
Some first nits:
- In {{ContainerImpl}}:
-- Use same comment style as above ("try to recover a container that was
previously paused");
-- Avoid the extra line.
- In {{RecoveredContainerLaunch}}, indentation needs to be fixed.
> Update paused container state in the state store
> ------------------------------------------------
>
> Key: YARN-6059
> URL: https://issues.apache.org/jira/browse/YARN-6059
> Project: Hadoop YARN
> Issue Type: Sub-task
> Reporter: Hitesh Sharma
> Assignee: Hitesh Sharma
> Attachments: YARN-5216-YARN-6059.001.patch,
> YARN-6059-YARN-5972.001.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
