Build a data platform, one verified step at a time¶
A hands-on lab for learning Linux, Kubernetes, data engineering and AI engineering. Follow the build from a Windows laptop to a working platform, with commands, explanations and screenshots from the actual setup.
Start with the foundation See the latest progress
The lab at a glance¶
| Layer | Technology |
|---|---|
| Virtual machines | Hyper-V on Windows |
| Operating system | Red Hat Enterprise Linux 9.8 |
| Kubernetes | kubeadm, CRI-O and Calico |
| Planned CI | GitHub Actions runner and a local registry |
| Planned workloads | Data pipelines, notebooks and AI experiments |
Current milestone¶
The cluster now has one control plane and two workers, all verified Ready. Calico is available, and worker-to-worker Pod HTTP, DNS and Service tests passed. The temporary test namespace was deleted (user-confirmed). The supporting GitHub runner/local registry VM remains planned.
Follow the build¶
- Set up Git and GitHub.
- Create the Hyper-V network.
- Install RHEL.
- Verify networking, SSH and package access.
- Prepare the Kubernetes node.
- Bootstrap the control plane and configure Calico.
- Set up and join the first worker.
- Join the second worker and test networking.
- Explore the CI and registry plan.
Evidence, not just instructions¶
The progress register separates verified outcomes from planned work. The screenshot index contains 51 reviewed captures, including intermediate failures and their resolution. Commands describe a guided build; read their context before rerunning them on an existing system.
This is an independent personal lab using synthetic data, not a production deployment.