It is time to move on to our real-world solution for real-time processing. By now, we should have a great grasp of what Bunny is, and what it allows us to do in the space of RabbitMQ.

To do real-time processing, we will need to be able to use the job queue/workers semantics that we've seen just earlier; but by having this, we may want to just handle the actual logic—what can we do to avoid rewriting the boilerplate that is a RabbitMQ worker?

Furthermore, what can we do to avoid the boilerplate that is a production-grade RabbitMQ worker? Such a production-grade worker should have available configuration, logging, metrics, and a good way to abstract away all of the AMQP/RabbitMQ gritty detail so that we're left with a familiar Ruby-like programming model.

When I started doing this, Ruby didn't have anything to offer. I ended up building my own library, and I've called it Sneakers.