Distributed Computing: Challenges And Solutions In Edge IOT Landscape
summary
A traffic‑generation platform is built to validate NSX firewall rules at scale, running as a Python daemon over 3,000–5,000 VMs via SSH, storing local state and pushing results to a central Webfront observability DB. Tests cover TCP UDP HTTP across roughly 500 million endpoint pairs and 100–150k rules. The team avoids Kubernetes, uses systemd, adds an NTP server for clock drift, implements retries acks and templates, and releases the framework as an open source python package.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:04Distributed systems overview
- 7:02Traffic generation case study
- 15:42Learnings from distributed systems
- 22:04Edge computing considerations
- 24:07Pitfalls in distributed design
related talks
32mFrom Edge To Insight: Cybersecurity Intelligence With Kubeflow...
49mA Real-World Example of Using Network Behaviour to Detect Cyber-Compromise
36mCatch Me If You Can – Leveraging Cloud Foundry, SDN, BOSH and Predictive Analytics
08mMaximizing ROI Through Distributed Computing and Multi-Cloud
33mEdge-Native: The new paradigm for operating and developing Edge Apps
29mCloud and Edge: Price, Performance and Privacy Considerations in IOT
57mWhy should you care about edge computing?
37mDevelop modern connected applications on 5G MEC/Edge Space and Azure