Kubernetes Notes for Practical Operators
Start with workloads
Most Kubernetes debugging starts by asking what workload owns the failing pod.
Check service boundaries
A service problem is often a selector, port, or readiness problem. Check those before blaming the network.
Keep manifests boring
Small manifests are easier to review, diff, and recover at 3am.