[ 
https://issues.apache.org/jira/browse/HADOOP-9933?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13768329#comment-13768329
 ] 

Karthik Kambatla commented on HADOOP-9933:
------------------------------------------

Thanks Steve. Sorry for the delayed response here.

Currently (yarn-1027-9.patch), we are able to get away with destroy and 
recreate. transitionToStandby() stops the current RM, creates and initializes a 
new RM instance. Looks like this works okay for now. Given the transitions are 
not expected to be so frequent, the performance implications might also be few.

That said, currently the RM doesn't run any "Always On" services. Our plan is 
to move some of the Active services to Always-On; e.g. ClientRMService - moving 
this to Always-On makes sure it is always available to service the clients. 
Destroy-and-recreate takes down the ClientRMService for the duration of 
destroy-and-recreate. Another issue in handling these Always-On services is 
that it needs to be destroy-recreate-start to start those services, which makes 
it all a little ugly. What do you think?
                
> Augment Service model to support starting stopped services
> ----------------------------------------------------------
>
>                 Key: HADOOP-9933
>                 URL: https://issues.apache.org/jira/browse/HADOOP-9933
>             Project: Hadoop Common
>          Issue Type: Improvement
>    Affects Versions: 2.1.0-beta
>            Reporter: Karthik Kambatla
>            Assignee: Karthik Kambatla
>              Labels: service
>
> For ResourceManager-HA (YARN-149 and co), we would want to start/stop/start 
> RM's active services as it transitions to Active/Standby/Active respectively. 
> In the current service model, we can't start the services that are already 
> stopped.
> Would be nice to augment this. To avoid accidental restart of stopped 
> services, we can add another API: start(boolean restartIfStopped). Thoughts?

--
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

Reply via email to