Hi, I am new in ActiveMQ, and try to get the following scenario working: I have external system, lets call it SysA. I have my own systems, lets call it SysB, SysC and so on. SysA is already fault tolerant, active/active etc, and is capable to communicate with multiple peers. My systems (SysB, SysC, SysN) are mutually independent and all have to communicate with SysA. I want to put SysA interfacing logic in separate module, let's call it SysA adapter. It will be stateless "gateway" module. It also has to be fault tolerant. I have experience with buses where multiple instances of the program can publish the same data, and subscribers will automatically get only one copy of the message (according to priorities between publishers or some other logic), allowing active/active publishers setup without any coding, just by configuration. So, if I were using such bus, I'd put two instances of SysA adapter on different physical hosts, they would communicate with SysA simultaneously, and publish incoming data on the bus. SysB, SysC etc would get each message only once. Is it possible to get the same net effect on ActiveMQ, even if with some coding ? I don't want to invent, test and certify "home brew" high availability solution from scratch. P.S. Obviously ActiveMQ itself has to be deployed using one of the high availability modes, this question is about how I protect myself against failure of the SysA gateway, not the failure of ActiveMQ engine. Thanks, Alex.
-- View this message in context: http://activemq.2283324.n4.nabble.com/High-availability-multiple-publishers-scenario-tp4720767.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.
