Skip to content
Hands-on IT IT Drills

Foreword

Read little, type a lot. A development team wants to run a small API with a database. That takes a Deployment, a Service, a Secret, a StatefulSet, a volume, a NetworkPolicy and, ideally, checks that report when something is stuck. That is seven objects with more than 200 lines of YAML between them. The team wants nothing to do with any of it and would rather just say: “I’d like a Notes app with two replicas.”

That is exactly what you build in this book. kro (Kube Resource Orchestrator) is a controller for Kubernetes. In it you describe your own API, such as NotesApp, and define which objects are created for each instance and how their values relate to each other. From that, kro generates a CRD, checks every expression as soon as you create it, and then keeps all objects alive in the right order. You need no Go code, no operator project of your own and no templates for this.

The common thread is notes-api, the small notes API from the sister books. On Day 4 it gets its own API, NotesApp. After that it grows day by day: status and readiness, Postgres with a protected volume, an optional NetworkPolicy, checks per path, central defaults from the platform, a fleet with several environments, permissions for teams and, at the end, GitOps.

This book is for platform people, admins and developers who know Kubernetes and kubectl and want to offer their teams their own simple APIs.

The Training Rules

  1. Type, don’t copy. Run every command yourself once.
  2. Compare with Expected. After every drill you’ll find what you should see. If it differs, that is where the real lesson is.
  3. First the RGD, then the instance. You check every change to a ResourceGraphDefinition with kubectl get rgd before you create instances. kro shows most errors right there.
  4. Pass the checkpoint. Only when the check commands show what is written there do you move on.
  5. Solve the challenge on your own. First the hint, then the solution.
  6. Lab values stay in the lab. lab-api-key-123 and the other lab-… values are practice data. Real secrets have no place in this book.

Warning: All output comes from real test runs. Times, IDs, hashes and pod names will differ on your machine. What matters is what happens.

Structure of Each Training Day

Section Purpose
Briefing The essentials in a few bullet points
Drills Commands and files with expected output
Checkpoint Individual check commands: did everything work?
Challenge A task without instructions
Enterprise Note What is different in production

The Training Plan

Part Day Unit
1 – Basics 0 The training ground
1 Installing kro, first RGD
2 SimpleSchema
3 CEL expressions
2 – The NotesApp 4 Several resources, one graph
5 Status and readiness
6 Postgres and the deletion policy
7 Conditions with includeWhen
8 Collections with forEach
9 External references
10 RGD chaining: the fleet
3 – Operations 11 Lifecycle and revisions
12 Permissions, feature gates, metrics
13 kro, Helm, Kustomize or operator?
14 GitOps, upgrade and outlook
Competition 15 Five failures
Appendix A Cheat sheet
B Final exam

The Lab

You work over SSH on an Ubuntu 24.04 VM kro with 2 CPUs and 4 GB RAM. You start it with Multipass, but any other Linux VM with sudo and internet access will do. k3s, kro and all instances run on it. Your laptop only needs SSH and curl.

The tested versions: k3s 1.37.2, Helm 4.3.1, kro 0.10.0, notes-api 0.5.1 and Postgres 18.6.

You’ll find the companion code at https://github.com/thomaszachmann/kro-zum-anfassen. Day 0 fetches it with git clone. For each day it contains the finished files in loesung/ and the broken files for the competition in wettkampf/. You are meant to write the files yourself. The solutions are there for comparison.

Where this book only touches on an area, the sister books go into depth: “Kubernetes hands-on” for the cluster, “Kubernetes operators hands-on” for your own controllers in Go, “Kyverno & NetworkPolicies hands-on” for policies.

Let’s go.