In this chapter, we saw how to extend the concept of RabbitMQ clustering with mirrored queues, which allow us to establish high availability at the level of the broker. Furthermore, we discussed additional mechanisms for improving reliability (in terms of connecting to the broker and processing messages such as AMQP transactions), publisher confirms, and client reconnections.