Planning

Two meetings a week, about twenty weeks. As a rule, the first meeting of the week is a theory class (read the class pages before or after it; each has slides, a quiz and playground scenarios) and the second meeting is a hands-on lab on BoxOffice. Deliverables are due on the last class day of the week, at 23:59, unless announced otherwise.

flowchart LR
    F["Foundations<br/>weeks 1–3"] --> B["Building blocks<br/>weeks 4–9"]
    B --> O["Operating at scale<br/>weeks 11–13"]
    O --> A["AI<br/>weeks 14–15"]
    A --> P["Prove it<br/>weeks 16–19"]

The order follows the story of the course: each class starts from the problem the previous one left, and each lab builds, in BoxOffice, the answer of a class you have already seen. Some weeks pair two classes and the next week pairs their two labs, because the labs need both: caching (lab 5) and data at scale (lab 6) after classes 6 and 7; observability (lab 8) and resilience (lab 9) after classes 9 and 10; Kubernetes (lab 10) and delivery (lab 11) after classes 11 and 12. Observability comes before reliability on purpose: you cannot improve what you cannot see, the SLO is defined where it is measured, and in the labs you instrument BoxOffice (lab 8) before you break it on purpose (lab 9) — otherwise the chaos experiments have nothing to show.

How a week works

1. Before the class: skim the class overview and try the first playground scenario of the page.
2. In class: theory with the slides, a "think first" question, a live playground run.
3. After class: read the subpages, do the quiz of the class, and the matching activity of the workbook.
4. Lab day: build the BoxOffice piece and measure it; push the code to your repository.
5. Every week: one project step (see the milestones).