Seamless List-Watch Improvements Across Data Center and Massive Unstable Network Edge
Talkat Kubernetes on Edge Day Europe 2025 · on Tuesday, 2025-04-01· at 08:45· Level 3 | ICC Capital Suite 10-12·
summary
KubeEdge replaces direct API server list-watch bursts with a KubeEdge cloud Hub that enqueues objects, sends only the latest state on reconnect, and requires an ACK from edge nodes; edges run a light kubelet and persist metadata in a local DB so apps auto‑recover after node restart. No changes are made to the Kubernetes master. A sync controller detects drift and re-applies objects. Field uses in commercial vehicles, offshore oil and CDN claim fewer breakdowns, lower bandwidth use and stable operation.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:07H Computing and Kubernetes background
- 2:39Kubernetes List-Watch challenges
- 4:26Cube Edge architecture overview
- 6:11Cloud to edge deployment workflow
- 7:38Cube Cloud and Hub module details
- 8:46Improved List-Watch to edge
- 10:35Sck controller for consistency
- 10:56Cube Edge use cases
- 12:45Cube Edge in oil fields
- 14:12Cube Edge for CDNs
- 15:45Summary
related talks
33mKubeEdge: From Fixed Location to Movable Edge, Latest Updates and...
34mManaging Edge Nodes With “KubeEdge”
27mManaging Large-Scale Edge Cluster Over Unstable Network with KubeEdge
22mKubeEdge: From Fixed Location To Movable Edge, Latest Updates And Future
35mKubeEdge – Kubernetes Native Edge Computing Framework
41mKubeEdge
20mKubernetes at the Edge
26mSeparation of Orchestration and Computation in KubeEdge, Understand the Why and How