[ 
https://issues.apache.org/jira/browse/HDDS-4008?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17163035#comment-17163035
 ] 

Bharat Viswanadham edited comment on HDDS-4008 at 7/22/20, 7:58 PM:
--------------------------------------------------------------------

[~avijayan]
ozone.om.internal.serviceid is added to let OM what are its internal peers, if 
the configuration has more than one service ids defined in 
ozone.om.service.ids. Added as part of HDDS-2536. 

S3G is also having a similar issue, we have failed this with a TODO which is 
not addressed. I think all services to behave in same manner, i think we should 
check if multiple service id's are configured, then we should set mandatorily 
ozone.om.internal.service.id. In this way, S3G and Recon will behave in the 
similar manner.

Any thoughts??






was (Author: bharatviswa):
[~avijayan]
ozone.om.internal.serviceid is added to let OM what are its internal peers, if 
the configuration has more than one service ids defined in 
ozone.om.service.ids. Added as part of HDDS-2536. 

Because in S3G we have failed this. I think all services to behave in same 
manner, i think we should check if multiple service id's are configured, then 
we should set mandatorily ozone.om.internal.service.id. In this way, S3G and 
Recon will behave in the similar manner.

Any thoughts??





> Recon should fallback to ozone.om.service.ids when the internal service id is 
> not defined.
> ------------------------------------------------------------------------------------------
>
>                 Key: HDDS-4008
>                 URL: https://issues.apache.org/jira/browse/HDDS-4008
>             Project: Hadoop Distributed Data Store
>          Issue Type: Bug
>          Components: Ozone Recon
>            Reporter: Aravindan Vijayan
>            Assignee: Aravindan Vijayan
>            Priority: Critical
>
> Recon connects to OM via RPC using the "ozone.om.internal.service.id" to get 
> updates. If the above config is not defined, but the ozone.om.service.ids is 
> defined, Recon should use the latter as a fallback. Currently, a single Recon 
> instance supports only 1 OM HA cluster at a time. Hence, if multiple ids are 
> defined, Recon will pick the first.
> Thanks to [~vivekratnavel] for reporting the issue.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

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

Reply via email to