An Introduction to SRE at Google (with DevOps Focus)
summary
SRE is presented as DevOps implemented with measurable rules: pick SLOs, track SLI metrics on dashboards, and enforce an error budget that balances reliability with feature velocity. Operations claim 40–70% of post‑launch cost and an internal study finds 70% of outages are due to recent change, so use progressive rollouts, automation, capacity planning, and change controls. Run blameless post-mortems, staff roughly 50/50 engineers and sysadmins who code, and start small with one service and the Google SRE books as reference.
AI-written summary · about
chapters (7)
- 0:00Introduction to SRE
- 1:48What is SRE
- 8:56Error budget principle
- 16:33Five key SRE practices
- 25:25Emergency response
- 31:42Blameless culture
- 36:28Summary and next steps
related talks
40mHow Google SRE and developers work together
32mHow Google SRE and Developers Work Together
22mSRE for Enterprises
05mHow do Devops and SRE relate?
06mIGNITE: Want to start with SRE? Start with this talk.
08mSRE vs DevOps: The Difference & Why It Matters
27mHow Google SRE and developers work together
07mIGNITE: Metrics, Margins, Mutiny – How to make your SREs (not) run away