Keptn for Delivery & SLOs

"Keptn for Delivery & SLOs: Event‑Driven Automation for Cloud‑Native Ops"
Modern Kubernetes delivery fails most often not because teams can’t deploy, but because they can’t decide—quickly, repeatably, and defensibly—whether a change is safe. This book is written for experienced platform engineers, SREs, and senior developers who want to replace subjective approvals and brittle pipeline logic with event-driven, evidence-based automation. If you’re operating multi-team clusters and need progressive delivery that survives real-world noise, drift, and async telemetry, this is your playbook.
You’ll learn how to treat SLIs, SLOs, and error budgets as control signals that drive quality gates and promotion policy, not as after-the-fact dashboards. The book builds a precise SRE vocabulary, then shows how Keptn (Kubernetes-native lineage) implements lifecycle checks, tasks, and evaluations through operators, CRDs, and reconciliation. You’ll design robust evaluation windows, baseline and multi-window gating to reduce false positives, and safe blocking vs non-blocking strategies—including failure handling that avoids deadlocks. Along the way, you’ll connect telemetry to release context, bind automation to app/workload/version identity, and integrate promotion across stages with GitOps.
Expect deep Kubernetes-native implementation detail: Helm install and upgrades, component responsibilities, RBAC and secrets/TLS hardening, Job/Pod execution semantics for tasks, and pre-deployment verification using Kubernetes Scheduling Gates. Familiarity with Kubernetes, GitOps/pipelines, and basic