Knative 0.18: Serverless Workloads Near GA

K8s Guru
2 min read
Knative 0.18: Serverless Workloads Near GA

Introduction

Knative 0.18, released on November 23, 2020, marks a turning point toward general availability. Serving APIs move to v1, Eventing matures broker/trigger contracts, and autoscaling handles bursty traffic with fewer cold starts.


Serving API Stabilization

  • Service, Configuration, and Route resources adopt the v1 API version with minor spec cleanups, locking down fields before GA.
  • ksvc revisions now expose rollout metadata for better observability and debugging.
  • Domain mapping and custom domain support improve multi-tenant routing without addon ingress controllers.

Eventing Enhancements

  1. Broker v1beta1: Standardizes delivery options, dead-letter queues, and tracing attributes.
  2. Trigger Filtering: Attribute-based filters allow fine-grained event targeting and fan-out.
  3. Sources: Amazon SQS, Apache Kafka, and GitHub sources gain production readiness docs and sample manifests.

Autoscaling & Performance

  • Activator optimizations reduce request queueing and bring pods online sooner, cutting cold start times.
  • Concurrency and burst capacity settings expose more precise controls for latency-sensitive workloads.
  • HTTP Probe improvements ensure that revisions switch traffic only when ready.

Getting Started

kubectl apply -f https://github.com/knative/serving/releases/download/v0.18.0/serving-crds.yaml
kubectl apply -f https://github.com/knative/serving/releases/download/v0.18.0/serving-core.yaml
kubectl apply -f https://github.com/knative/net-istio/releases/download/v0.18.0/release.yaml

Deploy a sample service:

kn service apply hello --image gcr.io/knative-samples/helloworld-go:v1 --revision-name hello-v1

Summary

AspectDetails
Release DateNovember 23, 2020
Headline FeaturesServing v1 APIs, Broker v1beta1, faster autoscaling
Why it MattersPositions Knative for GA with predictable APIs and operational guidance

Knative 0.18 signals that serverless on Kubernetes is ready for prime time, with APIs and tooling that production teams can trust.