Docker and Kubernetes for Developers in 14 Days

Fourteen days of drills on Docker and Kubernetes, built around one real application and one real cluster that stays up the whole way through.

This is a drill manual, not a narrative course. Every listing runs against a real Docker Engine and a real local Kubernetes cluster, and every printed output block is a paste from that run — including a real crash loop caused by a liveness probe set too eager, a real registry migration that broke on the very first try, and the real VM restarts that interrupted the work more than once. Each chapter states a rule in as few words as it can, shows the rule working in a short numbered example with its captured output, and hands it straight back as problems. Every example builds one application, hiretrack, a small job-application tracker that grows a real feature every day: a SQLite file, then Postgres and Redis behind Compose, then a full Kubernetes Deployment with probes, an init container, a CI/CD pipeline, and a real container registry.

The first week covers Docker: images and layers, a fast Dockerfile, multi-stage builds, volumes and bind mounts, container networking, a real Compose stack, and a systematic method for debugging a container that will not start. The second week moves the same application onto Kubernetes: Deployments, Services, ConfigMaps and Secrets, health probes and rolling updates, init containers and paused rollouts, kubectl-based debugging of planted failures, a real CI/CD pipeline against a real registry, and a closing chapter that runs the whole book back as a timed interview drill.

The measurements are reported as they came out. A one-line source edit that should touch nothing costs a full nine-second dependency reinstall, because of Dockerfile order, not bad luck. A liveness probe with too short a delay kills a perfectly healthy container, repeatedly, until the delay is fixed. A registry-based deploy fails on the very first attempt with a pull-policy conflict this book's own earlier chapters caused. A book that only printed the clean successes would be easier to write and worth less to work through.

Your 14-Day Path

• 1. IMAGES, LAYERS AND THE CONTAINER



• 2. WRITING A DOCKERFILE THAT BUILDS FAST



• 3. MULTI-STAGE BUILDS AND SMALL IMAGES



• 4. VOLUMES, BIND MOUNTS AND DATA



• 5. CONTAINER NETWORKING



• 6. COMPOSE FOR A REAL STACK



• 7. DEBUGGING A CONTAINER THAT WON'T START



• 8. IMAGE SECURITY AND SUPPLY CHAIN



• 9. FROM COMPOSE TO KUBERNETES



• 10. DEPLOYMENTS, SERVICES AND CONFIG



• 11. HEALTH CHECKS, PROBES AND ROLLOUTS



• 12. LOGS, METRICS AND kubectl DEBUGGING



• 13. CI/CD: BUILD, PUSH, DEPLOY



• 14. THE INTERVIEW

Who It's For

Developers preparing for a DevOps or platform-engineering interview. Backend engineers whose team just adopted Kubernetes. Anyone who has run docker run a hundred times and still cannot explain what a Service actually routes to. No prior Kubernetes is assumed; comfort with a shell and a text editor is.

302 worked examples. 302 figures. 2,053 numbered problems. 350 pages. Measured with Docker Engine 29.6.1, kubectl v1.36.4, and a kind v0.33.0 cluster running Kubernetes v1.35.0.

You need Docker Desktop or an equivalent Docker Engine, and about an hour a day. Start Chapter 1.