Each *cluster* should ideally use the same UDP address and port. If that's not possible then I see two options:
1) Specify a unique cluster user and password for each cluster. The nodes from the different clusters will still attempt to cluster with each other since they will be able to "see" each other, but security will prevent the nodes from one cluster from clustering with the nodes in the other cluster. 2) Use static clustering. In other words, manually specify the nodes in the cluster rather than relying on UDP discovery. Justin ----- Original Message ----- From: "anton.mithun" <[email protected]> To: [email protected] Sent: Wednesday, August 3, 2016 10:21:34 PM Subject: Re: Artemis - Two clusters and one udp group ip:port Hi, Thanks for the reply. As per my understanding, specifying the user/password for the cluster doesn't allow a client to differentiate between clusters on the same udp address. It gives cluster the power to permit other nodes whether to get connection to that cluster or not. May be I am wrong. Referring the document at the moment. And also, when u said "I would keep each server on its own UDP address. " , did you mean "each cluster" instead of "each server"? Cheers -- View this message in context: http://activemq.2283324.n4.nabble.com/Artemis-Two-clusters-and-one-udp-group-ip-port-tp4714902p4715049.html Sent from the ActiveMQ - Dev mailing list archive at Nabble.com.
