The path to a Serverless native era with Kubernetes
summary
Shows a working migration path: a CSV→JSON translator built on Google Cloud Functions that fails when retargeted to AWS because proprietary event formats differ, motivating a move to Kubernetes with Knative Serving, Eventing and Tekton Pipelines. Demonstrates scale-to-zero cost savings, cold-starts ranging from milliseconds to several seconds (10–15s reported on some JVM setups), and reuse across clouds when using CloudEvents. Compares Cloud Run, Cloud Run on GKE, Pivotal riff and OpenShift preview and flags governance and vendor-lock issues.
AI-written summary · about
chapters (14)
- 0:00Introduction and speaker
- 0:44Understanding Gartner hype cycles
- 3:59Hype cycle phases explained
- 10:28What is serverless computing
- 14:49Function as a service basics
- 24:40Serverless challenges today
- 28:40Why Kubernetes for serverless
- 31:58Knative building blocks
- 39:10Tekton pipelines overview
- 41:16Knative native platforms
- 44:27Knative governance news
- 45:51Simpler serverless on Kubernetes
- 47:17CNCF serverless landscape
- 49:24Future of Kubernetes and serverless
related talks
32mThe Path to a Serverless-Native Era with Kubernetes
01h 01mServerless containers with Knative and Cloud Run
48mEvent-driven serverless architectures using Knative and Cloud Run.
55mServerless On Your Own Terms using Knative
01h 09mTutorial: Cloud-Agnostic Serverless
52mServerless with Knative
55mEvent-driven serverless architectures using Knative and Cloud Run
23mServerless with Knative
