Step By Step Kubernetes Observability with eBPF
summary
Shows how eBPF captures every TCP connect in-cluster, compiles it with libbpf or BCC, packages the kernel-only probe as an OCI image using BumbleBee, and runs it as a DaemonSet so one lightweight binary runs on each node. Metrics are exported as Prometheus metrics, correlated with Pod and Service IPs, and rendered into a service-graph from Bookinfo traffic; the demo counts connections per source/destination and proves the pipeline works end-to-end.
AI-written summary · about
chapters (6)
- 0:00Introduction and welcome
- 3:12What is Kubernetes observability
- 7:27Why use eBPF for observability
- 14:30Demo introduction
- 18:25Deploying eBPF in Kubernetes
- 23:07Summary and Q&A
related talks
29mKubernetes Observability with eBPF
42mMonitoring Kubernetes Traffic with eBPF
31mBridging Dev and Ops with eBPF: Extending Observability Upwards and Downwards
18mFind out what your processes are capable of with ebpf
31mA High-Schooler’s Guide to Kubernetes Network Observability
36mBumbleBee: Or How I Learned To Stop Worrying And Love eBPF
27mLow-Overhead Tracing Using eBPF for Observability into Kubernetes Apps and Services
34mNetwork Observability with eBPF and OpenTelemetry