Seeking SRE

Chapter 11. SRE Patterns Loved by DevOps People Everywhere

Gene Kim, IT Revolution

This is an edited excerpt from The DevOps Handbook by Gene Kim, Jez Humble, John Willis, and Patrick Debois (O’Reilly, 2016).

When David Blank-Edelman asked me to contribute a chapter on the SRE body of knowledge and the impact it has had on the DevOps community, I gave a very enthusiastic “Yes!”

Although some might argue that SRE and DevOps are mutually exclusive, I argue the opposite. In my opinion, it is difficult to overstate the impact that SRE has had on framing how the operations community can best contribute to organizational goals and improving the productivity of developers. As Ben Treynor Sloss, VP of SRE at Google, famously said in his 2014 SREcon presentation: “I define SRE as what happens when software engineers create an operations group.”

In that famous presentation, Treynor Sloss introduces the breathtaking concept of a truly self-balancing system, where an organization first decides on an acceptable error budget, which then guides the prioritization of nonfunctional requirements and gates the decision to deploy and release.

During the research and writing of The DevOps Handbook (along with my coauthors Jez Humble, John Willis, and Patrick Debois), I couldn’t help but notice how many of the DevOps patterns that we love and now can take for granted were pioneered at Google.

Here are three of my favorite patterns, excerpted from The DevOps Handbook, that can be traced to the SRE ...