[
https://issues.apache.org/jira/browse/YARN-6620?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Wangda Tan updated YARN-6620:
-----------------------------
Attachment: YARN-6620.011.patch
Thanks [~zhengchenyu] for your review, I replied your comments here:
bq. When storeAssignedResources failed, there will be a leak. Because
postComplete will never be called. It means usedDevices won't be removed. So
should add cleanupAssignGpus in catch section.
Fixed and added UT.
bq. I thinks NM recovery for gpu doesn't take effect. Becuase
RecoveredContainerState's ResourceMappings didn't add to the recovered
Container.
It added to ContainerImpl in constructor:
{code}
// constructor for a recovered container
public ContainerImpl(Configuration conf, Dispatcher dispatcher,
ContainerLaunchContext launchContext, Credentials creds,
NodeManagerMetrics metrics,
ContainerTokenIdentifier containerTokenIdentifier, Context context,
RecoveredContainerState rcs) {
this(conf, dispatcher, launchContext, creds, metrics,
containerTokenIdentifier, context, rcs.getStartTime());
// ....
this.resourceMappings = rcs.getResourceMappings();
}
{code}
Please let me know if I missed anything.
Attached ver.11 patch.
> [YARN-6223] NM Java side code changes to support isolate GPU devices by using
> CGroups
> -------------------------------------------------------------------------------------
>
> Key: YARN-6620
> URL: https://issues.apache.org/jira/browse/YARN-6620
> Project: Hadoop YARN
> Issue Type: Sub-task
> Reporter: Wangda Tan
> Assignee: Wangda Tan
> Attachments: YARN-6620.001.patch, YARN-6620.002.patch,
> YARN-6620.003.patch, YARN-6620.004.patch, YARN-6620.005.patch,
> YARN-6620.006-WIP.patch, YARN-6620.007.patch, YARN-6620.008.patch,
> YARN-6620.009.patch, YARN-6620.010.patch, YARN-6620.011.patch
>
>
> This JIRA plan to add support of:
> 1) GPU configuration for NodeManagers
> 2) Isolation in CGroups. (Java side).
> 3) NM restart and recovery allocated GPU devices
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
