I am trying to establish the following setup:
(a) 2 servers running aretmis brokers: 1 live servers + 1 backup servers
(b) server discovery via JGroups with a jdbc_ping
(c) clients (producers & consumers) are running on both servers in a process
separate from the process running the brokers
(d) everything runs inside Spring containers, so the brokers are embedded
and the clients (consumers) extend DefaultMessageListenerContainer
(e) fail-back is enabled
So far so good, but now things work not as expected.
All consumers are configured to connect to the localhost broker -- in the
documentation it says:
/Server discovery is a mechanism by which servers can propagate their
connection details to:
Messaging clients. A messaging client wants to be able to
connect to the
servers of the cluster without having specific knowledge of which servers in
the cluster are up at any one time./
On the live server everything works fine.
On the backup server, however, I would have expected the client to prompt
the local backup broker, (somehow?) discover the connection details of the
currently live server and then connect to it.
Instead, the clients are only able to connect to the backup server after a
failover occurs (when live is stopped), and are never automatically
redirected to the live server. And even after the live server is up again
and fail-back occurs, the clients on the backup server lose connection.
Does my setup makes sense? Am I trying to achieve something which is not
supported?
NOTE:
* I can not use udp multicast in my environment (that's why I am trying to
use jgroups with jdbc_ping)
--
Sent from: http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html
