I built this pubsub module because I wanted a simple solution that uses only node.<p>Why did I want pubsub? I wanted to build a complex app but segregate it into microservices that work together through event messages. I also wanted to use this method instead of node's built-in clustering for multiple instances. I've been running it for a few weeks now with several messages per second bugfree. I plan to use this setup for future apps as it reduces complexity of the code and allows me to restart/update one part of the app without interrupting the other parts.<p>I would love to receive feedback and suggestions on the code and features, and hope it can be useful to you also!
I'd say one of the important ones for me is what are the mechanisms for retries under network loss or lack of connectivity in general. What is the behaviour/semantics of that? Can I buffer messages that then get backfilled when a recipient reconnects or do I need to do that in my code / application layer.
Not being mean but:<p><a href="https://www.google.com/search?q=postgres+pub+sub" rel="nofollow">https://www.google.com/search?q=postgres+pub+sub</a><p><a href="https://redis.io/docs/manual/pubsub/" rel="nofollow">https://redis.io/docs/manual/pubsub/</a><p><a href="https://www.rabbitmq.com/web-stomp.html" rel="nofollow">https://www.rabbitmq.com/web-stomp.html</a>