Having introduced the first SRE book for O’Reilly, I am honored to be invited back for the sequel. In this book, the writing team is leaving the history of the first book to speak for itself and reaching out to a broader audience, offering direct experiences, case studies, and informal guidance. The broad themes will be familiar to anyone in IT, perhaps relabeled and reprioritized, and with a modern sense of business awareness. In place of technical descriptions, here we have user-facing services and their promises or objectives. We see human-computer systems originate from within the evolving business, intrinsic to its purpose, rather than as foreign meteorites impacting an unsuspecting and pristine infrastructure. Cooperation of all human-computer parts is the focus. Indeed, the book might be summarized as follows:
- Commit to clear promises that set service objectives, expectations, and levels.
- Assess those promises continuously, with metrics and budgetary limits.
- React quickly to keep and repair promises, be on-call, and guard autonomy to avoid new gatekeepers.
Keeping promises reliably (to all stakeholders) depends on the stability of all their dependencies, of intent, and of the lives of the people involved (e.g., see Thinking in Promises). Remarkably, the human aspects of human-computer systems only grow alongside the perceived menace of scale: it turns out that automation doesn’t eliminate humans, after all; rather, it challenges us to reassert human needs across all scales, from the genesis of an individual idea to the massive deployments on behalf of a global user base.
Teaching these lessons is a service challenge in its own right—and, like any service, hard-won knowledge is an iterative process. We make these lessons our own by questioning, trying, failing, rehearsing, and perfecting them. There’s a wealth of material to ponder and adapt in the book, so let’s go.