From the clustering docs:
"When a remote client does a lookup through HA-JNDI, HA-JNDI will delegate to the local JNDI Context when it cannot find the object within the global cluster-wide Context. So, a EJB home lookup through HA-JNDI, will always be delegated to the local JNDI instance."
You should be able to tell whether HA-JNDI is correctly deployed from either the server.log or by looking in the JNDI console.
When you say that you're using the same JNDI for all instances do you mean that you're setting the JNDI and/or HA-JNDI ports to be the same for seperate instances of JBoss? I'm not aware that you can do this - each JNDI and HA-JNDI should run on a seperate port. When you perform a remote lookup you can list multiple server instances like this:
java.naming.provier.url=localhost:1100,localhost:1200,localhost:1300
Hope that helps.
-Andrew
Stefan Groschupf wrote:
Hi Bax,
nice to hear you! ;) I use such an pattern and it is generated by Xdoclet. I use the BeanUtil classes.
They make an normal remote lookup. As i understand the documentation i does not need an other lookup. Since the request will be forwarded to the HA JNDI. Isn't it?
Stefan
Holger Baxmann wrote:
your setup reminds me on the Service Activator Pattern http://java.sun.com/blueprints/corej2eepatterns/Patterns/ServiceActivator.html in conjunction with Session Facade and Service Locator for decoupling the lookup and the do-something stuff.
It could be more easy to solve the clustering jndi problem if you have a facade between for 'hiding' the cluster to the 'client' --- but i could be totaly wrong here ...
regards
bax
On Mon, 01 Sep 2003 10:12:23 +0200, Stefan Groschupf <[EMAIL PROTECTED]> wrote:
Dear all,
I have massive trouble to get an simple cluster to run. It sounds in all publications so easy, but something went wrong here.
I can run multiple instances on my machine via binding service. The nodes are discovered and the farming deployment running well.
As you may be remember i have a set of node with my project A: It is manually deployed on all nodes. I have my main application that has an mdb that run my "job" queue.
My problem is that it looks like i have no HA JNDI. To get something running i setup my instances on my machine to use all the same jndi of localhost. ;/
But when i setup my nodes in the network to use on jndi on my main machine it does of curse say that my project is already deployed. ;(
When i does not setup an jndi host as described in the clustering documentation i get an Object no bound exception.
My MDB:
TRExtractorRemote extractor = TRExtractorUtil.getHome().create();
ProcessedDocument document = extractor.process(oneValue.getValueExternal());
(The Remote Object is generated by xdoclet)
Since my nodes are in a win network and the farming works i think the multi casting working.
Really guys i had try everything i had found in the net or in the documentation and had tried it now 2 days and 2 nights, since i only can play with the nodes over night. Someone has any idea where my problem could be located?
I use jboss-3.2.2RC3_jetty-4.2.11.
Thanks for giving me light! Stefan
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user
