Super Charge Your Service Mesh With eBPF
summary
Claim: moving service‑mesh work into the kernel with eBPF cuts overhead and adds new observability and security hooks. Shows Cilium as the eBPF CNI replacing iptables’ linear rule evaluation with near constant‑time service IP→pod mapping, demonstrates sockmap short‑circuiting to avoid full TCP traversals and eBPF-assisted QUIC steering, and describes shipping Glue Mesh integrates Istio + Cilium to discover, manage, and combine telemetry and policy — reported as implemented and used.
AI-written summary · about
chapters (8)
- 0:00Intro
- 0:59Kernel
- 1:50What is eBPF
- 2:50Extended eBPF
- 3:36Kubernetes Networking
- 6:50Psilium
- 7:53Envoy
- 13:35What can we do
related talks
31mThe Future of Service Meshes: Sidecarless or eBPF?
20mEnvoy Mesh Acceleration: From Iptables to Fully BPF
36mCloud Native Superpowers with eBPF
33mThe Future of eBPF in Cloud Native
32mBuzzing across the eBPF Landscape and into the Hive
40mPanel Discussion: eBPF: A New Era in Cloud Infrastructure Tools
47mUsing eBPF to Bring Kubernetes-Aware Security to the Linux Kernel
35mAn Introduction to eBPF - Superpowers for Networking, Observability & Security