Network Observability in Kubernetes with Cilium and Hubble
summary
Describes using Cilium and Hubble with eBPF to provide kernel-level network observability so Kubernetes pods emit labeled, structured metrics and flow metadata without sidecars or service discovery. Covers practical constraints like AWS ENI/IP pool limits, compares DaemonSet vs sidecar patterns for logs and metrics, and walks through installation and the Hubble UI (live demo failed due to expired credentials); includes a Star Wars example and mentions production validation via vendor Groundcover.
AI-written summary · about
chapters (9)
- 0:00Kubernetes observability intro
- 1:50Kubernetes networking basics
- 10:20Dynamic observability challenges
- 13:06Traditional observability patterns
- 17:24Conventional methods limits
- 20:18Cilium eBPF observability
- 24:17Hubble benefits
- 26:08Hubble demo
- 32:22Q&A
related talks
29m“Are These Things Talking to Each Other?” - Observing Kubernetes Networking...
25mHubble Beyond Cilium
36mWhat's going on in my cluster?
01h 13mCilium Workshop
20mTackling Container Networking at Scale
35mNext-level Kubernetes Networking with Cilium - Nico Meisenzahl and Philip Welz
22mOne Project, Different Angles: How to Secure and Observe with Cilium
31mBridging Dev and Ops with eBPF: Extending Observability Upwards and Downwards