I’m about to start a new project that is gonna be based in micro-services.
They idea is that the micro-services are gonna be running in a k8s cluster, with different micro-services written in different languages. I like the way Nameko thinks about micro-services, about organizing the entry-points and dependencies explicitly.
However, there’s already a message queue + api libraries in place, so replacing them with Nameko’s built-in is not an option (we would be writing our own entry-points/dependency providers for those).
It it possible to run Nameko services without having them connected to a rabbitmq server (hence disabling built-in RPC + amqp support). Since they are gonna be running in a k8s cluster, the services would be able to find each other using the URLs generated for them.