Service Mesh: Service-to-Service Traffic
Sidecars, control vs data plane, mTLS identity, retries and circuit breaking in the mesh, traffic splitting, ambient/sidecarless — and when not to adopt one.
About 75 minutes · 15 exam questions · 9 interview questions · 2 code labs
What you will learn
- What a service mesh is (and why it exists)
- Control plane vs data plane (how a request actually flows)
- mTLS and workload identity
- Timeouts, retries and circuit breaking in the mesh
- Traffic splitting, header routing and mirroring
- Observability: what the mesh gives you and what it can’t
- Sidecar vs sidecarless: ambient, eBPF and proxyless gRPC
- Mesh failure modes you will meet in production
- When NOT to adopt a mesh, anti-patterns, and choosing one
Previous: Reverse Proxy, Forward Proxy, API Gateway & BFF · Next: Load Balancing: L4/L7, Algorithms, Consistent Hashing