Foreword

In the spring of 2014, the Delivery Engineering team at Netflix set out to achieve a lofty goal: enable end-to-end global continuous delivery via a software platform that facilitates both extensibility and resiliency. My team had previously built two different applications attempting to address Netflix’s delivery and deployment needs, but both were beginning to show the telltale signs of monolith-ness and neither met the goals of flexibility and resiliency. What’s more, the most stymieing effect of these monolithic applications was ultimately that we were unable to keep pace with our partner’s innovation. Users had begun to move around our tools rather than with them.

It became apparent that if we wanted to provide real value to the company and rapidly innovate, we needed to break up the monoliths into small, independent services that could be released at will. Embracing a microservice architecture gave us hope that we could also address the twin goals of flexibility and resiliency. But we needed to do it on a credible foundation where we could count on real concurrency, legitimate monitoring, reliable and easy service discovery, and great runtime performance.

With the JVM as our bedrock, we looked for a framework that would give us rapid velocity and steadfast operationalization out of the box. We zeroed in on Spring Boot.