humble_dev

I am looking for the best practises around managing RabbitMQ changes/configuration.

In my current micro-services/worker based architecture I have:

* many publisher that publish to exchanges, these exchanges are linked to one or more queues. * Each worker has it's own queue that is later bind to the exchange.

The goal is that publisher and workers do not know anything about each other or where message goes to.

I saw that many people put a configuration for Exchanges/Queues in the code, but then a worker depends on the created exchange. That means I need to deploy publisher first. Otherwise worker won't be able to bind the queue to the exchange if it won't be created. The other solution is to manually bind the exchanges with the queues after the deployment but I would rather have everything automated. I wonder what are the best practises around that subject.

As most of developers, I try to create as much reusable code as possible. But in reality, I rarely use my code again. I could easily say that less than 5% of my code will be reused in the future.

I wonder how it looks for other people, is it worth to make your code reusable in the first place?