Hello, I am currently using HZ topics to publish messages between two machines, M1 runs HZ and M2 uses a HZ-Client. M1's SW creates (and publishes over) topic1 and subscribes to topic2. M2's SW creates (and publishes over) topic2 and subscribes to topic1.
This all works well because I have just to machines here. But I plan to add more M1-like machines and also M2-like, therefore having more HZ clustered nodes.
The problem I am fearing is that my messages on topic* are to be processed by just one HZ node (whichever of them). My concrete usage cases are sending mobile push notifications and XMPP messaging. I do not want M1-like SW to publish something and be processed by *every* M2-like, but *just by one* of them.
From my (limited) HZ understanding, I could refactor and use IQueue's instead. I would then expect for any node to put a message in the queue and then *any -but just one-* of the others pop it and process.
¿Am I right assuming this will solve my scenario? ¿Is there any other approach I could take?
Thanks a lot in advance!