Kubernetes

This is the course material for Kubernetes for Operations — five days of running Kubernetes rather than just deploying to it. The focus is on the things that page you at three in the morning: storage that will not bind, networking that drops traffic silently, clusters that will not scale, and backups that turn out not to restore.

How a Day Runs

The mornings are for understanding, the afternoons are for doing. Ask questions during the presentations — that is cheaper for everyone than discovering a gap three labs later.

The Week

Work through the modules in the sidebar in order — each one assumes the one before it, and the labs are numbered to match.

Two Things, Not One

You have a textbook and a workbook, and they are deliberately separate:

• This site is the textbook. Concepts, terminology, worked examples. Read it at your own pace, and come back to it after the course.
• The workbook is a repository cloned into your workspace at /home/coder/course-material. It contains the labs and a single command, labctl, that drives them.

labctl env      # is my environment working?
labctl list     # what labs are there, and how far have I got?
labctl docs 04  # open the module for a lab, in this site

When You Are Stuck

There are four levels of help, and they exist so that nobody sits and stews:

1. labctl explain 04 — the idea and what “done” should look like.
2. labctl steps 04 — a guided recipe, if you want one instead of working it out.
3. labctl hint 04.2 — a nudge for one specific checkpoint. Ask twice and it gets more specific.
4. labctl solution 04 — the reference answer. Available to anyone.

Using the hints and the solution is recorded, and that is on purpose: it tells the instructor who to come and sit with. It is not a mark against you. Being stuck in silence for an hour is the only thing that actually wastes the week.

Still being written. The structure and the links are in place. Some pages are complete and some are marked Scaffold, which means the prose is still to come.