From “why build this?” to the design review: which questions to ask when designing a system and what risk each one guards against, walked through with one real example, a call-summary system for a customer support team.
From virtual machines on vSphere to cloud auto scaling, containers, and Lambda — at every stage I worked through, what monitoring cared about drifted up another layer.
A Kubernetes persistent-disk billing incident, the tradeoff between centralized and decentralized infrastructure governance, and a fix that makes monitoring appear automatically with every new resource.