[ 
https://issues.apache.org/jira/browse/HDFS-14162?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16763971#comment-16763971
 ] 

Erik Krogen commented on HDFS-14162:
------------------------------------

I had some discussion offline with [~shv] and he pointed me towards 
{{ProtocolProxy}}, which appears to maintain some mapping of the methods 
implemented by the proxy.

After looking into this further, I realized that it is actually keeping track 
of which methods are implemented by the _server_, rather than the proxy on the 
client. Actually, its methods won't even work on HDFS proxies, because HDFS 
protocol interfaces do not implement {{VersionedProtocol}}, and 
{{ProtocolProxy}} does a cast to {{VersionedProtocol}} to fetch the server-side 
information.

However this pointer inspired me to make some modifications, relying more on 
reflection, to greatly simplify the {{CombinedProxyBuilder}} (renamed to 
{{ProxyCombiner}}). [~shv], let me know your thoughts on v001.

> Balancer should work with ObserverNode
> --------------------------------------
>
>                 Key: HDFS-14162
>                 URL: https://issues.apache.org/jira/browse/HDFS-14162
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Konstantin Shvachko
>            Assignee: Erik Krogen
>            Priority: Major
>         Attachments: HDFS-14162-HDFS-12943.wip0.patch, HDFS-14162.000.patch, 
> HDFS-14162.001.patch, testBalancerWithObserver-3.patch, 
> testBalancerWithObserver.patch
>
>
> Balancer provides a substantial RPC load on NameNode. It would be good to 
> divert Balancer RPCs {{getBlocks()}}, etc. to ObserverNode. The main problem 
> is that Balancer uses {{NamenodeProtocol}}, while ORPP currently supports 
> only {{ClientProtocol}}.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to