Quartz 5

Homelab

3 min read

Fastigiata Homelab

Purpose

Fastigiata Homelab is a small multi-machine Linux and Apple lab used for local AI experimentation, cybersecurity practice, containerisation, networking, web development and technical learning. The design prioritises reusable older hardware, local-first services, secure LAN access and practical experimentation. As of August 2026 the official laboratory inventory sat at 9 devices however for core operational function it has been standardised to 5 devices.

Homelab Architecture

SystemPrimary RoleKey Technology
MintBookProMain Linux production workstationLinux Mint, Docker, Open WebUI, cyber tools
M1 MacBook ProLocal AI + developmentLM Studio, Ollama, MLX, VS Code
MintBookAirCybersecurity / demonstration machineLinux Mint, Kali Docker, LM Studio, native cyber tools
MX Linux MacBookLightweight AI/container experimentationOllama, Podman, Open WebUI
Fedora ServerLAN services / server experimentationFedora, Docker/Podman, Apache, SSH
Raspberry Pi 4Network monitoring / infrastructureUbuntu/Linux, Docker, Uptime Kuma, Pi-hole experiments

🧠 Local AI

The lab supports running AI models locally rather than relying exclusively on cloud services.

Core platforms

  • LM Studio— LM Studio is used for model testing, GGUF/MLX models and LAN model serving.

  • Ollama — Ollama runs as lightweight local inference and API access.

  • Open WebUI — browser-based interface for locally hosted models.

  • Obsidian + Local AI — knowledge-base querying, note analysis and future RAG experimentation.

Typical model range: 1B–14B, depending on machine capability.

Current model families include:

  • Llama (Llama3.2:3B)

  • Qwen (Qwen3:9b, Qwen2.5 Coder:3B and 7B)

  • Gemma (Gemma 3 and 4, 1B -12B)

  • Phi (Phi 3.5 Vision)

  • Ministral 3B

  • Nomic Embed and Embedding Gemma

Local AI experiments focus on RAG, coding assistance, knowledge retrieval, agentic workflows and private AI services.

🛡️ Cybersecurity Lab

The MintBookAir is the primary cybersecurity experimentation machine. I am currently studying a Certificate IV in Cybersecurity at CIT. The homelab is my best effort at simulation of a multi device networked environment for learning.

Kali Environment

Kali Linux is primarily operated as a Docker-based CLI toolbox, providing an isolated environment without requiring Kali as the host operating system.

Native / Lab Tools

  • Nmap

  • Wireshark

  • Burp Suite

  • Metasploit

  • Hydra

  • Nikto

  • Netcat

  • DNS utilities

  • Whois

  • Traceroute

Typical activities include:

Network Discovery → Enumeration → Traffic Analysis → Web Testing → Vulnerability Research

Warning

Cybersecurity testing is restricted to systems, networks and services owned by or explicitly authorised for the Fastigiata Lab.

🐳 Docker & Containers

Docker provides isolated services and disposable development environments across the lab.

Typical containers include:

  • Open WebUI

  • Kali Linux

  • Uptime Kuma

  • AI supporting services

  • Web and development experiments

Useful commands:

docker ps
docker ps -a
docker images
docker start <container>
docker stop <container>
docker exec -it <container> bash
docker logs <container>

Podman is also used on selected Linux systems for daemonless/rootless container experimentation.

🌐 Network & Services

The lab primarily operates as a private LAN environment.

Services include:

  • SSH administration

  • Local AI APIs

  • Web interfaces

  • Apache/web hosting experiments

  • Network monitoring

  • DNS/Pi-hole experimentation

  • Docker services

Remote services are deliberately reviewed to avoid unnecessary internet exposure.

flowchart LR
    LAN[Fastigiata LAN]

    LAN --> Mint[MintBookPro]
    LAN --> Air[MintBookAir]
    LAN --> M1[M1 MacBook Pro]
    LAN --> MX[MX Linux]
    LAN --> Fedora[Fedora Server]
    LAN --> Pi[Raspberry Pi]

    Mint --> Docker[Docker / Open WebUI]
    Air --> Kali[Kali Cyber Lab]
    M1 --> AI[LM Studio / Ollama]
    MX --> Podman[Podman / Ollama]
    Fedora --> Services[LAN Services]
    Pi --> Monitor[Monitoring / DNS]

Design Principles

Local First · Open Source · Small Experiments · Secure LAN · Reusable Hardware · Documentation

The lab is designed as an evolving learning environment rather than a fixed production architecture. New technologies are tested on a small scale, documented in Obsidian, and retained only when they provide a clear improvement to the overall system.

homelab local-ai linux docker cybersecurity ollama lmstudio openwebui fastigiata

Graph View

Backlinks