[jboss-dev-forums] [Design of Clustering on JBoss] - API modifications for better extensibility ?

[deruelle_jean]

Hi all,

I'm starting to work on the task of providing clustering and failover capabilities to our Mobicents Sip Servlets container working on top of JBoss AS 5.1.CR1. See http://code.google.com/p/mobicents/issues/detail?id=521

I took the AS 5.1.CR1 as a basis for my work and the corresponding 2.0.0.GA tags for the different HA sub projects.

the initial import of my code is located here : 
http://code.google.com/p/mobicents/source/detail?r=6013

This is not working yet but as I extended the JBoss cluster classes I noticed that few API changes would make my life a lot easier and would provide better integration, so here are my requests, please comment or tell me if I need to open one or multiple JIRA issues if you do agree with the changes :

JBossCacheManager

*  initSnapshotManager to be made protected so that it can be overriden (so that we can set our own converged snapshotManager)
* getSnapshotManager() and setSnapshotManager() public methods or snapShotManager_ field to be made protected
* installValves and installValve methods to be made protected so that they can be overriden (we need that to install our custom JVMRouteValve)
* Ideally, methods such as start/stopEmbedded and UnEmbedded and the methods they use internally should be made protected too

AbstractJBossCacheService

* make the storeSessionOwners protected so that we can reuse it

IntervalSnapshotManager

* make the fields protected

View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231452#4231452

Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231452