Welcome to SRE.
Site Reliability Engineering treats operations as a software problem: instead of protecting uptime through heroics and tribal knowledge, you define reliability as a number, budget for a controlled amount of failure, and automate everything else. This course walks the discipline end to end — starting with SLIs, SLOs, and error budgets, moving into the day-to-day work of running production (observability, on-call, postmortems, capacity planning), then the engineering patterns that make systems fail gracefully (redundancy, progressive delivery, chaos engineering), and finishing with the reporting, tooling, and drills a real on-call rotation runs on. It's written for engineers who already ship code and now own what happens after it deploys — backend, platform, and infrastructure engineers moving into or working alongside an SRE function. Expect concrete formulas, real tool names, and a full worked incident — not slogans about "reliability culture."
Imagine a school bus service that promises to get kids to school on time on 99 out of every 100 school days. That missing 1% isn't laziness — it's a budget: some days there's traffic, a flat tire, a snowstorm. The dispatcher tracks how many "late days" have been used up this month. If the budget is nearly gone, the company stops adding new routes and fixes whatever keeps making buses late instead. If there's budget to spare, it takes the risk of trying a faster route. SRE runs production software the same way: decide up front how much unreliability you can tolerate, spend that budget deliberately on the risk of moving fast, and once it's gone, reliability work jumps the queue.
What you'll learn
Reliability by the Numbers
What SRE is, SLIs/SLOs/error budgets, and toil.
2 · Running Reliable SystemsKeeping Systems Running
Observability, on-call, postmortems, and capacity planning.
3 · Engineering for ReliabilityEngineering Around Failure
Resilience patterns, progressive delivery, and chaos engineering.
4 · Practice & ReferenceProof, Tools, and Drills
Reporting, tools, a real outage story, and drills.
★ · PlatformThe Academy
Back to the hub for every course on the Academy platform, including this one.
How this course works
The course is self-paced: read pages in the order the sidebar lists them, or jump straight to what you need — nothing is gated behind a quiz. The sidebar and the prev/next pager at the bottom of every page are both generated from the same source, manifest.js, so the two never disagree about what comes next.
Everything is static HTML that runs entirely in your browser, with no account required to read it. Practice & Reference closes the course out with a searchable glossary, a deck of flip flashcards for the vocabulary and formulas, and a printable on-call readiness checklist; progress is tracked locally through the sidebar's visited-page state, not a server-side record.