Document for JMS Failover Implementation On WebLogic

Question

Hi,
I am looking some good links and techniques to implement JMS failover using WebLogic 10.3.
FailOver* [As we do with our Databases (Concept of Clustring)]
System will consist of two app servers and each will have its own application deployments but if one failed for some reason the application messages should redirected to the other servere and vice versa.
Above efinition is very brief but if anyone can help provide some good documents and info how to implement it it will be appriciated.
Thanks :-)

Answer 1

Thanks alot guys for your help. We successfully implemented it at our servers here by creating distributed queues targetting all servers in a cluster.
One point which I think is worth mentioning and I want to share with all us here is that; when App Server [where MDB will post the message finally after retrieving from queue] if that goes down what will happen, what MDB will do with that message?.
We impleneted the DLQ (error destination) and deploy one more MDB_DLQ_SERVER2 (Let say App SERVER 1 is down) which gets triggered when any message comes to DLQ and post that message to some other App Server, Let say message has been read by MDB_SERVER1 on SERVER1 but offcourse actaull server is down so message should get Re-directed to its Error Destination after it expiration peiod or whatever the settings are. DLQ (Error Destination) which is also a distributed destinatrion again targetting all servers in cluster same as actaull Request or Reply queues BUT MDB_DLQ_SERVER2 which is deployed on Server2 is NOT able to read this message. It get triggered but can not access the message.
After debugging for almost a day we found out its because message has been transafed to DLQ but actaully its resides in a FILESTORE_SERVER1 and MDB_DLQ_SERVER2 is not able to access it.
To work with that we have to define MDB_DLQ_SERVER1 to cater the SERVER1 failure and MDB_DLQ_SERVER2 to cater SERVER2 failure.
Reason I am mentioning this because as I said DLQ is also a normal Distributed Queue but at the same time its NOT as Distributed as its says.
Hope you all understand what I just wrote above.
Now I need to implement exactly the same scenario using four seperate physicall machine containing my four servers. I tried this scenario by creating four machines where node manager for each server is running and listning but when I am trying to start the server it gives me Certificate Exception with bad user name and password. Anyway I have seen some posts here regarding this; So i think i'll be fine.
Thanks Again,
Sheeraz