Do We Really Need Kubernetes at the Edge
summary
Argues that Kubernetes is useful sometimes but not universal: k3s and KubeEdge reduce footprint compared with vanilla Kubernetes, while edge platforms like ioFog, FogOS and Zeno handle disconnected, constrained, long-lived devices. Presents survey numbers: AI gets 30% of workloads, 48% use cloud-managed OTA updates versus 42% local, 38% still update by cable. Recommends deciding by latency predictability, statefulness, node constraints and control-plane proximity rather than assuming Kubernetes everywhere.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 2:55Do we need communities
- 3:31Agenda for today
- 4:29Cloud definition
- 6:34Edge definition
- 9:50Formal definition
- 12:46What doesnt work
- 17:36What makes Edge Native different
- 20:08Cloud Native vs Edge Native
- 20:38The Edge Continuum
- 22:47Eclipse Developer Survey
- 27:13Kubernetes at the Edge
- 32:21Other Options
- 35:36Should we be running Kubernetes
- 38:16Modern Factory
- 38:37Eclipse Basics
- 40:28Should I use Kubernetes
- 44:34Edge Computing Ecosystem
- 46:08Edge Ups
- 52:05Questions
- 58:34Closing
related talks
39mShould You Bring Kubernetes on Your Edge Roadtrip?
20mKubernetes at the Edge
36mEdge Computing in Kubernetes: Best Practices for the Future
34mKubernetes IoT and Edge Working Group
01h 16mIntro + Deep Dive: Kubernetes IoT Edge WG
34mKubernetes IoT and Edge Working Group
20mKubernetes at the Edge for Telcos
11mThe Journey to Kubernetes Everywhere