The Network in Kubernetes: Can We See It More Clearly?
summary
Kubernetes networking adds layers (CNI, network policies, service meshes) on top of traditional networks, making visibility and troubleshooting harder. The talk surveys network observability technologies for Kubernetes — from NetFlow to eBPF to service meshes — and presents Red Hat’s Netobserv project for OpenShift/Kubernetes, which aims to improve network understanding and troubleshooting with an emphasis on ease of use.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 0:15Speaker Introductions
- 1:42What is the Network of En-Culé?
- 4:12What is Observability?
- 5:32Basic Needs
- 6:26New Tools
- 7:48The Wick Service
- 9:25Wick Service Metrics
- 10:43Other Solutions
- 11:29CNN Example
- 12:19What is CNN?
- 12:45CNN Disadvantages
- 14:06Technical Details
- 15:32A Solution?
- 17:52IPF Advantages
- 19:58Connections
- 21:23Demo Presentation
- 22:41Cilium Hub
- 23:42Pixie
- 26:45Unit Flows
- 30:46The Local Team
- 32:48The Maradona Example
- 36:17Conclusion
related talks
46mNetworking in Kubernetes: can we make it clearer?
29m“Are These Things Talking to Each Other?” - Observing Kubernetes Networking...
31mFrom CNI to Service Mesh: Demystifying Kubernetes Networking
34mUnderstanding the Kubernetes Networking Layer
01h 28mIntro + Deep Dive: Kubernetes (Network) SIG
29mInside Kubernetes Networking
49mHow to reconcile networking in Kubernetes with legacy systems
20mKubernetes Networking 101