liveBook · Manning

Knative Serving

As explained in Google Anthos in Action

Knative Serving is the Knative component for deploying, serving, and routing serverless workloads packaged as containers.

Knative Serving deploys container-based serverless workloads on Kubernetes and exposes them through HTTP(S) requests. It abstracts deployment and networking details while scaling workloads from zero to N instances and routing traffic among application revisions. section 9.0 section 9.3.2

A Knative Service coordinates the workload lifecycle through related custom resources:

  • Service defines the container version and traffic rules, then creates the resources required for the workload lifecycle. section 9.3.2
  • Revision records an immutable snapshot of a container version and its configuration. section 9.3.2
  • Configuration separates application code from desired-state settings; changing it creates a new Revision. section 9.3.2
  • Route maps endpoints to one or more revisions, enabling controlled traffic distribution without directly managing low-level networking resources. section 9.2.2 section 9.3.2

Deploy and split traffic

A minimal Service manifest names the workload, selects a container image, and supplies environment variables such as TARGET. Traffic rules can then send 95% of requests to helloworld-v1 and 5% to helloworld-v2, supporting a canary rollout.

Knative can run on managed or self-managed Kubernetes environments, including Anthos GKE, EKS, IBM IKS, OpenShift, and Minikube. Installing Serving requires CRDs, core components, networking, DNS, and optional extensions; YAML files or an operator can perform the installation. section 9.3.5 section 9.0

How Google Anthos in Action covers it

How deep it goes
Explains Serving resources, autoscaling, traffic splitting, and installation stages, but defers detailed operational procedures.
Examples
Uses a Hello World Service manifest, resource listings, revision traffic percentages, and an e-commerce frontend scenario.
What it assumes
Assumes Kubernetes familiarity and teaches Serving’s resource model and installation requirements rather than Kubernetes fundamentals.
What you can do after
Operate container deployments, scale workloads on demand, and manage revision traffic through Knative resources.
How it connects
Connects Kubernetes, custom resource definitions, revisions, Knative Eventing, service mesh networking, and canary deployments.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Google Anthos in Action. The section links point to the parts of the book it draws on.