Brandon Savage continues his look at using RabbitMQ and PHP together to queue up requests today in this latest post. In this new part of the series, he focuses on the last piece of the puzzle - consuming the requests in the queue. (Parts one and two)
Once you've created a RabbitMQ producer, it's fairly easy to create a consumer. In fact, the only difference is in exactly what commands you're using. The connection, envelope, channel and queue declarations are the same. While in RabbitMQ you publish to the exchange, you actually do consume a specific queue. As a result, the commands for consuming are part of the AMQPQueue class.
He shows you how to set up the code to sit in the background and wait for a queue request and how to fetch them in a non-blocking way. He finishes off the post with a look at handling success and error conditions (based on the status of message consumption, not the result of the processing).