It does support clustered RMI servers with load balancing, but I don't think it
supports clustered singletons. Interestingly, the RMI singleton failover that
you are looking for is very close to how the HTTP session data is clustered. It
has a primary and a secondary copy of the HTTP session state, and when the
server with the primary goes down, the secondary copy is automatically made the
primary, and a new secondary copy is created. I wonder if something similar
could be written to handle RMI servers? Certainly BEA could build it in to WLS,
and you could probably build some sort of clustered RMI proxy service to handle
connecting you to the primary singleton by looking it up via JNDI, and handling
the failover by looking for a secondary via JNDI and giving you a reference to
it. What would be better is something like rmid to handling activating a new
copy of the singleton when one can't be found in the cluster.
-----Original Message-----
From: A mailing list for Enterprise JavaBeans development
[mailto:[EMAIL PROTECTED]]On Behalf Of Smith, Curt H.
Sent: Thursday, July 13, 2000 12:17 PM
To: [EMAIL PROTECTED]
Subject: Re: Resiliency: method level failover, who provides this
> There is a similar setting for use with RMI servers that has
> the exact same
> effect, and I have used this as well:
> java weblogic.rmic -clusterable -methodsAreIdempotent -O -d
> %SERVER_CLASSES%
> asep2.services.PointEngine.PointEngineImpl
>
> We use this feature quite a bit in our application, and it
> works as advertised
> <g>.
>
The above about RMI services plus the part I cut re:EJB interfaces
is very good news for WLS.
I've totaly missed this in my experiments and reading about WLS.
I probably
stopped reading after I found that WLS didn't support Corba services,
just support for being a client to external Corba servers.
Does WLS support clustered RMI servers with load balancing? I'll
guess it does. I have a need though for a resilient service in
a clustered box architecture for a Singleton service.
Can WLS manage a Singleton RMI service in a cluster of boxes, where
it'll be started somewhere if that box or server goes down?
Thanks alot!
curt
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
