Traffic Splitting and Distributed Tracing with Linkerd

with Charles Pretzer (Buoyant)

Kubernetes Colorado, January 2020

Service Mesh Overview

• Observability: traffic metrics, request rate, seeing headers
• Security: mTLS between servics, on by default, transparent
• Reliability: latency aware load balancing, retries, timeouts
• Traffic Management: route traffic in a way you want, either canary or blue/green deploys

Distributed Tracing

• I dunno.

Deploying Linkerd

Splitting Traffic