DevOps Mastery Guide

Home / How to learn from this guide

How to learn from this guide

This page is the study contract. Each module assumes you follow it.

The three levels

Every module is split into three bands. Read them in order the first time. After that, skip by heading.

Band Heading prefix Purpose You should leave able to…
Beginner Beginner Mental model + one working path Explain the tool in one sentence and repeat the first success
Intermediate Intermediate Worked examples and comparisons Choose between two options and say why
Advanced / Production Production Security, scale, real-world constraints Apply a practice and name the failure it prevents

Do not start a Production section until you can do the “you should already be able to…” checklist at the top of that section.

Use every module the same way

  1. Read the 60-second overview and the mermaid until you can redraw it. Then read the beginner concept cards (What / Why / How it looks / Common confusion).
  2. Run First success from the directory the page names. Type the commands; do not treat the block as decoration.
  3. Check your output against Expected output. Small differences (IPs, versions, timestamps) are normal. If it failed, use that note, then the pitfall table.
  4. Keep evidence: the command, working directory, exit code, and a line of output that proves it worked.
  5. Continue only when you can explain what each command did. Do the Basic exercises the same day (exercises/ on Linux, Git, Docker, and Kubernetes; inline Practice on the other modules until those folders exist). Linux, Git, Docker, and Kubernetes also have a short quiz on the module page (quiz.json). Intermediate and Production wait for a second sitting.

A module’s beginner band is sized for about 15–45 minutes of focused work, not a weekend.

Machine check after clone:

./scripts/preflight.sh

It lists which first-success paths will run now. It does not install packages and it does not invent lab output.

Path files the site and this page share: from-zero, apps, machines, observe.

What to install (by module)

You do not need the whole toolchain on day one.

You are opening… Have this ready
01 Linux A Linux shell (Ubuntu 22.04+ or Debian, native / VM / WSL2) with sudo.
02 Git Linux shell plus git on the PATH. sudo apt install -y git if missing. No GitHub account for first success.
03 Docker Docker Engine or Docker Desktop. Your user in the docker group.
04 Kubernetes Docker plus kind plus kubectl. Not kubeadm, not a cloud cluster.
05 GitHub Actions A GitHub account and a repo you can push to. No self-hosted runner required.
06 Prometheus Docker. Kubernetes Operator comes later.
07 Grafana Docker. This module has its own compose stack (Grafana + Prometheus). Stop module 06’s stack first — both bind host port 9090.
08 Ansible Linux shell plus ansible on the control machine. Localhost is enough for first success.
09 Terraform terraform or tofu. First success uses the local provider — no cloud account.
10 GitLab CI A GitLab.com (or self-managed) project you can push to. Shared runners. Nothing to install on the laptop for first success.

Python 3 and a text editor are assumed after Linux.

Lab options

  • Best default: Ubuntu in a VM or WSL2, 4+ GB RAM, 20+ GB disk.
  • Kubernetes modules: kind on that same machine. Minikube is fine if you already use it.
  • Cloud free tier: useful later for Terraform Intermediate, not required to learn the core idea.
  • Do not start with a production cluster, kubeadm on bare metal, or a company AWS account.

When to move to the next module

You are ready when you can, without looking at the page:

  • Say what the tool is for in one sentence.
  • Name the two or three objects in its mental model (image vs container, playbook vs inventory, state file vs plan).
  • Repeat the first-success commands and explain what each one did.
  • Hit the Basic exercises’ success checks.

You do not need every Intermediate example or any Production section before moving on.

  • New to servers: 01 → 02.
  • Ship an application: 01 → 02 → 03 → 04 → 05.
  • Configure machines: 01 → 02 → 08 → 09.
  • See what you shipped: 06 → 07.
  • Capstones: only after the modules listed on projects/README.md.

Folder numbers follow the default (apps) study order. Other paths skip modules. JSON for the same routes: learning-paths/.

If you get stuck

  1. Compare your output to the “expected output” block. Small differences (IPs, versions, timestamps) are normal.
  2. Check the module’s pitfall table (Symptom | Cause | Fix).
  3. Look up the term in the glossary.
  4. Read the official “start here” link at the bottom of the module — not a random blog post from the integration section.

What this guide will not do

  • It will not replace official docs for API details.
  • It will not provision your company’s cloud.
  • It will not pretend a planned topic (GitOps, AWS, service mesh) exists as a module.