ArgoBox Infrastructure
Enterprise-grade homelab environment
Two sites, 40 miles apart, one unified network
The Story
It started with a Raspberry Pi running Pi-hole. Then a used Dell server from eBay. Then another. Then "I should probably set up proper backups." Then "what if I had a remote site for disaster recovery?"
Now there are two sites 40 miles apart, connected by Tailscale mesh VPN. The primary site runs Proxmox hypervisors with VMs and LXC containers. The remote site has an Unraid array for media and backups. Both work together as one logical infrastructure.
This is the homelab that runs everything else.
Multi-Site Architecture
Milky Way (Primary)
Local Network- Build Swarm Gateway & Orchestrator
- K3s Kubernetes Cluster
- Plex Media Server
- DNS & DHCP Services
- Development VMs
Andromeda (Remote)
40 Miles Away- Backup Orchestrator (Failover)
- Build Drones (24 cores)
- Media Archive
- Offsite Backups
- Remote Access Point
Sites connected via Tailscale mesh VPN with subnet routing. ~38ms latency, feels like a LAN.
What's Running
🖥️ Virtualization
- Proxmox VE clusters
- LXC containers
- KVM virtual machines
- GPU passthrough (RTX 4070 Ti)
💾 Storage
- TrueNAS SCALE (ZFS)
- Unraid (200TB+ media)
- Synology NAS units
- Btrfs with snapshots
🔨 Build Infrastructure
- Distributed Build Swarm
- 5 build drones (66 cores)
- Binary package hosting
- Automated CI/CD
🎬 Media
- Plex Media Server
- Sonarr / Radarr / Prowlarr
- Tautulli analytics
- Overseerr requests
🌐 Networking
- OPNsense firewall
- Tailscale mesh VPN
- Pi-hole DNS
- VLAN segmentation
📊 Monitoring
- Uptime Kuma
- Prometheus + Grafana
- Loki log aggregation
- Custom dashboards
Hardware Inventory
| System | Role | CPU | RAM | Storage |
|---|---|---|---|---|
| Izar-Orchestrator | Primary Proxmox | Xeon E5-2680 v4 | 128GB ECC | 2TB NVMe + 8TB HDD |
| Arcturus-Prime | Secondary Proxmox | i7-4790 | 32GB | 1TB SSD |
| Capella-Outpost | Desktop Workstation | i7-4790K + RTX 4070 Ti | 32GB | 2TB NVMe |
| Tarn-Host | Remote Proxmox | Xeon E5-2670 | 64GB ECC | 1TB SSD |
| Meridian-Mako-Silo | Unraid NAS | i7-4770 | 32GB | 200TB+ (14 drives) |
Design Philosophy
Redundancy Over Simplicity
Two orchestrators, two sites, automated failover. Single points of failure get eliminated.
Document Everything
If it's not documented, it doesn't exist. Every system has runbooks for recovery.
Automate the Boring
Backups, updates, health checks - anything done more than twice gets scripted.
Learn by Building
This isn't just infrastructure. It's a learning lab for enterprise concepts at home scale.
Dive Deeper
Explore the infrastructure in detail or see what's running right now.