The Life of On-Call, how SRE is changing DevOps
summary
Claim: SRE shifts on‑call from NOC spreadsheets to an engineering‑led, observability‑driven practice. Concrete moves: adopt OpenTelemetry for instrumentation and multi‑destination export, run a service maturity checklist before deploy, pick 3–5 global metrics (service tiers keep 5–15), avoid >6–7 per service, make developers on‑rotation and SREs own rotation/escalation strategy. Observability = metrics + logs + traces/spans; incidents without a response strategy see acknowledgments of up to 45 minutes, so build runbooks, service maps and incident tooling now.
AI-written summary · about
chapters (4)
related talks
25mUnleashing the potential of SRE for modern operations
25mSRE Guiding Principles & Transformational Framework
42mDemystifying SRE and DevOps
39mSRE or Devops - I am an Architect - Why Do I Care?
41mAn Introduction to SRE at Google (with DevOps Focus)
23mSite Reliability Evangelism: Practice Start-up within an Established...
25mSRE in the Third Age
58mEnterprise Roadmap for SRE