Master production-grade DevOps through hands-on labs. You take a single web service and, week by week, containerize it, wire CI/CD, provision its infrastructure, make it observable, orchestrate it on Kubernetes, and ship it through GitOps and progressive delivery — ending with a multi-service, self-healing, fully-observable system and a portfolio that proves you can run software in production.

"You build it, you run it." — Werner Vogels

The arc is build → integrate → provision → observe → orchestrate → deliver → operate at scale. Each lab builds on the last; your service grows from one container to a three-service cluster.

📅 16-week schedule. If your semester runs shorter, lectures pair up (two per week) so the lab cadence stays one-per-week. Lectures 1-16 map 1:1 to Labs 1-16; bonus labs 17-18 are covered by Lecture 16.

You start with one Python service (Lab 1) and never throw it away — every lab adds a production capability to the same project. Two course-provided plumbing services join later to make orchestration concepts real:

graph LR
    U[User] -->|HTTP| WEB[web :8080<br/>your Python service]
    WEB -->|Lab 9+| ECHO[echo :8081<br/>plumbing]
    WEB -.->|Lab 13+| HEALTH[health :8082<br/>plumbing]

    style WEB fill:#2196F3,color:#fff
    style ECHO fill:#FF9800,color:#fff
    style HEALTH fill:#607D8B,color:#fff

You never write the plumbing services — you deploy and wire them. They expose Prometheus metrics so your Lab 7-8-16 observability stack picks them up automatically.

Student-produced directories are gitignored in this repo so the upstream stays clean. You commit them to your fork.

Each main lab (1-16) is worth 10 points of main tasks + up to 2 bonus points:

• Main tasks sum to 10 pts. Split varies by lab (e.g. 6+4, or 2+3+3+2) — Task 1 is always standalone so later labs never depend on a task you skipped.
• Bonus tasks sum to 2 pts, flat. Genuinely challenging extensions, not busywork.
• Bonus labs (17, 18) follow the same 10 + 2 shape and are the exam-alternative track.

Acceptance criteria and a rubric table close every lab. Minimum passing per lab: 6/10.

Five components. Maximum contributions sum to 139%, capped at 100% — so there are multiple paths to an A and no single component is mandatory.

• Practice path: all main labs (70%) + bonuses (14%) + both bonus labs (20%) → 104% → capped A, no exam.
• Exam path: all main labs (70%) + bonuses (14%) + a solid exam (30%) → A, no bonus labs.
• Mixed: main labs + some bonuses + one bonus lab + a decent exam.

Each lecture has a 15-question post-quiz. Quizzes feed 5 rolling leaderboard windows across the semester:

The top-10 students in each window split that window's small point pool (≈1% each, ~5% total). Late-joining students can still win later windows.

graph LR
    A["Fork repo"] --> B["Branch lab<N>"]
    B --> C["Complete tasks"]
    C --> D["Push & open PR"]
    D --> E["Submit PR URL via Moodle"]
    E --> F["Receive feedback"]
    style A fill:#4CAF50,color:#fff
    style D fill:#F44336,color:#fff
    style E fill:#00BCD4,color:#fff

git checkout -b lab1
# ... complete the lab ...
git add app_python/
git commit -m "Complete lab1"
git push -u origin lab1
# Open a PR from your-fork:lab1 → your-fork:main, submit the PR URL on Moodle

DevOps foundations

• The Phoenix Project — Gene Kim et al. (2013)
• The DevOps Handbook (2e) — Kim, Humble, Debois, Willis (2021)
• Accelerate — Forsgren, Humble, Kim (2018)

Tooling

• Docker Deep Dive — Nigel Poulton
• Terraform: Up & Running (4e) — Yevgeniy Brikman
• Ansible for DevOps — Jeff Geerling
• Kubernetes Up & Running (3e) — Burns, Beda, Hightower
• Learning Helm (2e) — Butcher, Farina, Dolitsky

Observability & reliability

• Observability Engineering — Majors, Fong-Jones, Miranda
• Site Reliability Engineering — Beyer et al. (free at sre.google/books) — pairs with the SRE-Intro elective

Online

• DORA State of DevOps
• CNCF Cloud Native Landscape
• 12 Factor App

Ready? Start with Lab 1. Questions → course Moodle or office hours.