CKA Intensive — 4-Day Training (Trainee Materials)

A 4-day, hands-on Certified Kubernetes Administrator (CKA) training course targeting Kubernetes v1.36. This repo contains everything trainees see: labs, slides, primers, reference sheets, and the platform scripts that build the per-trainee Kubernetes cluster.

Trainer: Luis Torres

Trainers: the trainer-only material (schedule, day-by-day word-for-word scripts, lab solutions, Lab-9 break/fix scripts) lives in a separate private repo: cka-intensive-trainer.

---

What's in this repo

.
├── trainees/                    ← what you read during the course
│   ├── README.md                 ← course overview
│   ├── pre-course-setup.md       ← REQUIRED before Day 1
│   ├── docker-primer.md          ← mandatory pre-course Docker primer (2-3h)
│   ├── linux-primer.md           ← mandatory pre-course Linux primer (1-2h)
│   ├── vm-setup.md               ← Day-1-morning cluster bootstrap walkthrough
│   ├── lab-reset.md              ← inter-lab reset decision tree
│   ├── cheatsheet.md             ← 1-pager exam survival kit
│   ├── resources.md              ← further-study links
│   ├── day1/  day2/  day3/  day4/   ← 18+ hands-on labs
│   └── slides/                   ← Marp markdown + exported PDFs
└── infra/scripts/                ← what you run on your VM
    ├── template-bake.sh          ← TRAINER-ONLY (you'll never run this)
    ├── kind-bootstrap.sh         ← creates the 3-node kind cluster
    ├── kind-reset.sh             ← full rebuild ~90s
    ├── lab-clean.sh              ← soft cleanup between labs ~5s
    ├── verify-cluster.sh         ← 14-check cluster sanity
    ├── verify-template.sh        ← 30+ check machine sanity
    └── preload-images.sh         ← pre-pull lab images into kind nodes

---

Course-day overview

Day	Theme	Labs
1	Containers from scratch, architecture deep-dive, kubectl, workloads, ConfigMap/Secret, probes	lab0, lab1, lab2, lab2b
2	Services, DNS, Ingress, Gateway API, NetworkPolicy + CNI internals	lab3, lab3b, lab4
3	Scheduling, storage, RBAC, Pod Security + immutable infra, Helm, Kustomize, HPA	lab5, lab5b, lab6, lab6b, lab6c, lab6d, lab6e, lab6f
4	kubeadm, edge/lightweight K8s (k3s), TLS certs, etcd, CRDs, troubleshooting, mock	lab7, lab8, lab8b, lab9, mock-exam

CKA curriculum weights this defends against:

Domain	Weight
Cluster Architecture, Install & Config	25%
Workloads & Scheduling	15%
Services & Networking	20%
Storage	10%
Troubleshooting	30%

---

How the labs run

Each trainee gets a Debian 13 VM with everything pre-installed (Docker, kubectl, helm, kind, kubeadm tooling, VS Code, the course repo, pre-pulled images). Days 1-3 use a 3-node kind cluster running on that VM. Day 4 deletes the kind cluster and runs kubeadm init on the VM directly for the real kubeadm/etcd/troubleshooting experience.

Two-tier reset model:

Tier	Cost	Used for
./lab-clean.sh	~5s	90% of lab-to-lab transitions
./kind-reset.sh	~90s	Cluster wedged, before Lab 9, etc.

See trainees/vm-setup.md for the full setup walkthrough and trainees/lab-reset.md for the inter-lab decision tree.

---

Before Day 1 (mandatory pre-work)

Trainees must complete, in order:

1. trainees/pre-course-setup.md — connect to your assigned VM, confirm the pre-installed tooling
2. trainees/docker-primer.md — what a container actually is (2-3h)
3. trainees/linux-primer.md — systemd, journalctl, vim survival (1-2h)
4. vimtutor — 30 minutes; the exam uses vim

---

License

MIT. See LICENSE.

You're welcome to use, fork, and adapt this for your own training. If you do, a link back to the upstream is appreciated but not required.