Deployments and rollbacks without drama
For developers, DevOps and platform engineers, and the operations staff who press the deploy button or get paged when a release goes wrong. You should know what a build artifact and a production environment are; the Kubernetes examples assume you can run basic kubectl commands, but the method applies to any platform. The course covers why small, reproducible, automated releases are easier to fix, how to plan a staged canary rollout with clear stop criteria, how to judge a canary against its control (including the error-budget arithmetic), how to choose between rolling back, switching off a feature flag and rolling forward, and how to run and verify a Kubernetes Deployment rollback. It's grounded in Google's SRE books and the Kubernetes documentation; your own change and release policy decides who approves what. Ends with a supervisor-graded release exercise.
- Level
- Intermediate
- Length
- About 85 minutes
- Contents
- 5 lessons · final exam
- Status
- Published · updated 10 Oct 2026
Skills you'll practise
- Explain why small, automated and reproducible releases are easier to test, troubleshoot and roll back
- Plan a staged canary rollout with population, duration, metrics and stop criteria
- Calculate a canary's error-budget impact and decide from canary-versus-control metrics whether to continue or roll back
- Choose between rolling back, disabling a feature flag and rolling forward for a failing release
- Run and verify a Kubernetes Deployment rollout and rollback with kubectl rollout status, history and undo
Course outline
- 1.Explain why small, reproducible releases are easier to fixLesson · 15 min
- 2.Plan a staged canary rollout with stop criteriaLesson · 18 min
- 3.Judge a canary against its control and do the error-budget arithmeticLesson · 16 min
- 4.Choose between rollback, feature flag and roll forwardLesson · 14 min
- 5.Run and verify a Kubernetes rollout and rollbackLesson · 12 min
- 6.Deployments and rollbacks without drama: knowledge checkKnowledge check · 14 questions
- 7.Deployments and rollbacks without drama: practical exerciseKnowledge check · 1 question
- 8.Final exam10 questions · passing it completes the course, so people who already know the material can test out
Sources it draws on
The lessons and questions are written from these references, so learners can go back to the original.
- Google SRE book, Chapter 8: Release Engineering (Dinah McNutt; self-service, high velocity, hermetic builds, staged deployment)
- Google SRE book, Chapter 27: Reliable Product Launches at Scale (gradual and staged rollouts, feature flag frameworks, client retry behaviour)
- Google SRE Workbook, Chapter 16: Canarying Releases (Alec Warner and Štěpán Davidovič; canary versus control, error budget, metric selection, blue/green)
- Kubernetes documentation: Deployments (rolling update, rollout status, history, undo, maxUnavailable/maxSurge, progress deadline, revision history)
See it with your own jobs and topics
Tell us about your team and we'll walk you through setup, from choosing jobs to your first skills check.