[
https://issues.apache.org/jira/browse/YARN-2883?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15222196#comment-15222196
]
Arun Suresh commented on YARN-2883:
-----------------------------------
I think we have a tradeoff here :
# On one hand, I agree with [~kkaranasos] that {{ContainersMonitor}} has more
insight into the available/utilized resource (as well, in the future, possibly
the knowledge of the queue length and wait-time).. not to mention, the fact
that the {{resourceCalculaterPlugin}} etc are all present in the Monitor, and
therefore, as with the current patch, the monitor should decide when to start /
stop a container.
# But on the other hand, I see [~kasha]'s point.. Moving the queues to the
manager means, there is no back and forth, via events, between the manager and
the monitor. We would of-course maybe need to expose methods in the monitor to
return available / utilized resources which can be used by the Manager to
decide when to queue / start / kill containers.
Going thru the patch again, I feel maybe Option 2 might be easier to implement
and additionally would have minimal (on first-pass, I see only the need to
notify the Manager of a finished container) impact on the existing
ContainersMonitor.
Thoughts ?
> Queuing of container requests in the NM
> ---------------------------------------
>
> Key: YARN-2883
> URL: https://issues.apache.org/jira/browse/YARN-2883
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: nodemanager, resourcemanager
> Reporter: Konstantinos Karanasos
> Assignee: Konstantinos Karanasos
> Attachments: YARN-2883-trunk.004.patch, YARN-2883-trunk.005.patch,
> YARN-2883-trunk.006.patch, YARN-2883-trunk.007.patch,
> YARN-2883-trunk.008.patch, YARN-2883-yarn-2877.001.patch,
> YARN-2883-yarn-2877.002.patch, YARN-2883-yarn-2877.003.patch,
> YARN-2883-yarn-2877.004.patch
>
>
> We propose to add a queue in each NM, where queueable container requests can
> be held.
> Based on the available resources in the node and the containers in the queue,
> the NM will decide when to allow the execution of a queued container.
> In order to ensure the instantaneous start of a guaranteed-start container,
> the NM may decide to pre-empt/kill running queueable containers.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
