Start with the problem
- Cloud wasn’t built as a curriculum. It grew out of failures, constraints, scaling problems, and operational pain.
- Instead of memorising what every service does, I try to ask: why would someone need this in the first place?
- That is how I structured the sessions. Introduce the problem, then show the tool as a way to solve it.
What we worked through
- Docker images and containers, starting with the problem of an application behaving differently across environments.
- Running containers and building an image, then moving into Kubernetes concepts and deployments.
- Metrics Server and Horizontal Pod Autoscaler, connecting resource metrics to changes in replica count.
A starting point, with room to dig deeper
- One hour is nowhere near enough to understand Docker or Kubernetes properly. I wanted people to leave with a useful mental model and somewhere to start.
- Teaching also helped me see where cloud concepts become confusing, especially when tools are introduced without the problems behind them.
- If you use the slides, check commands, installation links, and tooling against the current docs. There are usually several ways to solve the same problem.
Docker & Kubernetes session slides
1.2 MB · 37 pages