"Flagger Progressive Delivery: Automated Canary Releases with Metrics"
Progressive delivery is no longer a niche practice—it's a core reliability capability for teams shipping continuously on Kubernetes. This book is written for experienced platform engineers, SREs, and senior developers who already operate production clusters and want to replace "hope-and-rollback" releases with automated, evidence-driven promotions. If you've felt the pain of noisy telemetry, ambiguous rollbacks, or traffic-shifting that behaves differently than expected, you'll find a disciplined approach here.
You'll build a rigorous model of progressive delivery as a feedback control system, then learn how Flagger implements that model as a Kubernetes operator. The book goes deep on the Canary CRD as a declarative rollout contract—analysis cadence, step weights, thresholds, promotion/rollback state, and safety controls—while connecting each knob to real SLO decision boundaries. You'll also master traffic management semantics across meshes, ingress controllers, and Kubernetes-native routing, including weighted shifts, blue/green switchovers, and A/B cohorting. Finally, you'll design high-signal metrics gates with Prometheus (and alternatives), integrate webhook-based tests, wire actionable notifications, and apply repeatable troubleshooting playbooks and governance guardrails.
Expect configuration detail, operational trade-offs, and failure-mode thinking throughout. Familiarity with Kubernetes workloads, services, and observability is assumed; the differentiator is the end-to-end, production-grade tr
