Building my home lab
Linux services, local virtual machines and a place to understand how the pieces fit together.
My home lab gives me somewhere to learn by running services myself. I use Linux, Docker Compose and Traefik across local Proxmox virtual machines and Oracle Cloud.
This article began as a January 2026 overview. The tools below reflect my setup as of October 2026; the earlier list of plans is no longer a description of what is installed.
What runs where
Proxmox provides local virtual machines for experiments. Docker Compose describes the services I run, while Traefik routes requests to applications. Oracle Cloud provides another Linux environment outside my home network.
A Raspberry Pi runs Pi-hole for DNS filtering. I use Tailscale to reach home devices remotely. These tools solve separate problems: filtering DNS queries does not control every network connection, and remote access still needs an access policy.
Looking at the system
I use Wazuh for lab log analysis, and Prometheus with Grafana for system metrics and dashboards. Logs help explain what happened; metrics help show how a system is behaving over time.
Keeping those views separate is useful. An application can be reachable while still returning errors, and a healthy host does not establish that every service on it works.
DNS is one part of the path
Pi-hole answers local records and filters requests. Permitted public queries still go to an upstream resolver; in the configuration described in my DNS article, that is Cloudflare.
When looking at a connection problem, the questions are different at each step: did the name resolve, was the host reachable, and did the application return the expected response?
What I’m documenting next
The public repository currently contains a short README. More useful evidence would be selected configuration, a recovery procedure and a dated account of a specific failure. I’m keeping those separate from claims about what the lab already proves.