Full-Scale Edge Computing with Eclipse ioFog: A Case Study
summary
Cloud round trip latency is ~250 ms versus edge needs as low as ~20 µs so the project deploys Eclipse ioFog: a tiny ioFog agent (15 MB) on Linux gateways, a fog controller control plane, optional CommSet mesh, Docker microservices, REST/WebSocket local APIs, and resource management to pick which nodes can run jobs. A smart agriculture/cellular base station case enforces a no-data-leaves policy, provides data provenance, automated quarantine and guaranteed wipe of edge services, and runs image recognition locally with proven production deployments.
AI-written summary · about
chapters (20)
- 0:00Start
- 1:28Why You Need To Be at the Edge
- 2:27Autonomous Vehicles
- 4:14Key Points Compliance and Security
- 4:53Seven Reasons for Being at the Edge
- 5:04Why Do You Want Your Edge To Be Open-Source
- 6:23Blockchain
- 7:51Organizational Risk
- 8:17Licenses
- 14:15Components
- 17:11Edge Kernel
- 17:40Resource Management
- 21:17Why Would You Use Eclipse
- 22:14Data Provenance
- 24:51Eight Requirements for Secure Commercial Grade Edge Computing
- 25:09Hardware Root of Trust
- 28:44Automated Quarantine
- 30:15Architectural Principles
- 33:10Software To Overcome Hardware Issues
- 36:22Asset Tracking
related talks
55mFog Computing and IoT Solution Architecture
39mExploring the Trend Toward the Edge
41mCloud Computing: The Edge talks - long road into the Cloud
51mNo-compromise solutions for the integration between Edge and Cloud
31mWhat is Edge Native?
30mEclipse IoT and Eclipse Edge Native: Making Sense of the Puzzle
30mEnabling a Cloud-native Edge for IoT Scale
36mIntegration of Edge and Cloud IoT Platforms