Our solution is fully reactive, so we need to use Reactor RabbitMQ, which allows us to interact with the RabbitMQ broker using the reactive paradigm.
On this new microservice, we do not need to send messages through the message broker. Our solution will listen to the RabbitMQ queues and push the received Tweets for the connected clients.
The Reactor RabbitMQ tries to provide a reactive library to interact with the RabbitMQ rboker. It enables developers to create non-blocking applications based on the reactive stream, using RabbitMQ as a message-broker solution.
As we learned before, this kind of solution, in general, does not use a lot of memory. The project was based on the RabbitMQ Java client and has similar functionalities, if we compare it to the blocking solution.
We are not using the spring-amqp-starter, so the magic will not happen. We will need to code the beans declarations for the Spring context and we will do that in the following...