A working mental model.
Kubernetes becomes easier to reason about when you can see the relationships, the transitions, and the gap between what you asked for and what exists.
Learn the mechanism
This book connects practical operation with the internals that explain it. Each lesson starts with a problem, introduces the model, follows it through an interactive scene, and gives you a way to test your understanding locally.
The learning path spans foundations, intermediate operations, and advanced control-plane and platform patterns. You can move in sequence or use the concept atlas to investigate a specific question.
Use the models deliberately
Scenes are bounded educational models. Their timing, node counts, and implementation details are simplified where that makes a mechanism easier to inspect. The “About this model” note identifies material boundaries. Animation time is not a measured Kubernetes latency.
Real systems are asynchronous, eventually consistent, and implementation-dependent. The local labs help bridge the explanation and real behavior. Production HA, managed-cloud load balancers, storage drivers, and failure domains need environments beyond a small local cluster.
Follow the primary sources
Lessons link to upstream Kubernetes documentation and official project documentation. Each lesson records a review date and Kubernetes version. Stable core APIs are the baseline; ecosystem add-ons and experimental behavior are called out explicitly.
Accuracy and review
The full curriculum—60 lessons and 121 narrated scenes, including alternate paths—was checked against primary documentation on 27–28 October 2026. The reviewed baseline is Kubernetes 1.37; the reproducible local labs pin kind 0.33.0 and its Kubernetes 1.37.0 node image. Lesson review dates identify the content check, not an upstream certification.
Review includes explanations, diagram states, examples, and the distinction between baseline labs and optional extensions. Add-ons have independent support requirements: the Argo CD extension requires a separately verified compatible version pair. A documentation review cannot guarantee behavior on every runtime, network or storage provider. Follow the linked primary sources and report a concrete mismatch when you find one.
Your progress belongs to you
Completion, bookmarks, and your last lesson stay in this browser. There is no account, learning-data backend, or tracking script. Export your progress from the learning path to carry it to another browser. Clearing browser storage removes local history.
Built to be improved
Content, simulation scenarios, shared renderers, and labs have separate responsibilities. The repository includes authoring templates, a coverage validator, and a contribution guide. Improvements should preserve precise explanations, primary references, accessible visuals, and runnable exercises.
Public source publication and open-source licensing will be decided by the project owner. This site is an independent learning resource and is not affiliated with or endorsed by the Kubernetes project or CNCF.
Practice, then explain it back
Before playing a scene, predict the next state. Change one input. Explain which component acts and why. In a lab, inspect the object’s spec, status, events, and ownership before changing it. Understanding the boundary is as useful as memorizing the command.
Open the learning path →