Monitoring Out, Observability In
summary
Argues that monitoring’s static checks fail for distributed systems and that observability requires outputs the system emits itself; shows deployment on an AKS cluster using Istio sidecar injection and the bookinfo sample with Prometheus/Grafana/Zipkin integration. Demonstrates control plane vs data plane, service mesh for uniform telemetry, and an ingress gateway exposing the app. Build mostly succeeds but several pods remain pending from resource limits and a container fails during live demo, so the pattern works but cluster sizing and config matter.
AI-written summary · about
chapters (7)
- 0:00Start
- 2:55Black Box Monitoring
- 9:00Definition of Observability
- 10:23Policies of Distributed Computing
- 10:44Eight Policies of Distributed Computing
- 19:27Domain Driven Design
- 36:25Custom Resource Definitions
related talks
28mImplementing Observability in Kubernetes
32mTesting Kubernetes apps' observability end-to-end
32mUnderstanding Observability in Kubernetes
24mObservability for Istio Service Mesh
31mCloud Native Observability
21mtalks about mastering Observability for DevOps
01h 29mTutorial: A Hitchhiker's Guide to CNCF/OSS Observability So...
46mObservability Ecosystem in Kubernetes