"Knative Serving: Serverless Workloads on Kubernetes Without the Chaos"
Running "serverless" on Kubernetes is easy to start and notoriously hard to operate well. This book is written for experienced Kubernetes engineers, platform operators, and senior application developers who need Knative Serving to behave predictably under real production constraints—multi-tenancy, strict change control, demanding latency SLOs, and incident-driven debugging—without hand-wavy abstractions.
You'll build an operator-grade mental model of Knative Serving from the ground up: the Kubernetes primitives it relies on, the workload packaging contract (OCI images, digests, private registries), and the installation/upgrade discipline required to keep a fast-moving control plane compatible. From there, you'll master the Serving API—Services, Routes, Configurations, and immutable Revisions—learning to reason from status conditions and reconciliation behavior instead of guesswork. Deep coverage of the request path (ingress/router, activator, queue-proxy, user container) turns cold starts, buffering, and timeout behavior into tunable systems.
The book then focuses on production outcomes: autoscaling and concurrency tuning for bursty traffic, revision-based progressive delivery with fast rollbacks, and the operational essentials—golden-signal observability, OpenTelemetry-forward tracing, TLS and identity patterns, and systematic troubleshooting playbooks. Expect trade-off-driven guidance, failure-mode catalogs, and workflows you can apply immediately in day-2 operations.
