🖥️ Core Objectives

• Test and develop hybrid infrastructure solutions
• Train and benchmark AI models with GPU servers
• Maintain large-scale, high-throughput storage for media and data-intensive workloads
• Provide zero-trust remote access to distributed systems (including Helium nodes)
• Continuously improve my skills in Linux, networking, automation, and virtualization

---

🧩 Hardware Stack

Device	Role
Supermicro (Proxmox)	pfSense firewall, internet gateway, nginx
Dell 1Gb Switch	Layer 2 distribution to WiFi + Desktop
TP-Link Deco Mesh	Wireless access for home + IoT devices
Mellanox SX6036	10GbE/40gbe backbone switch
Dell R720XD #1	JBOD host (2x 60-bay, ~2PB total)
Dell R720XD #2	Media server, backups, VM host
Supermicro (8x GPU)	AI/ML workloads, training environment
Desktop (Win11)	Orchestration + management
Mac Mini, VPS, Phone	Remote devices via Tailscale VPN

---

🛠️ Virtualization Stack

• Proxmox VE on all nodes
• Ubuntu as the primary VM OS
• LXC containers for Media, DevOps tools, sysadmin services
• Windows Server & Windows 11 test environments
• AI Workloads run inside Ubuntu VMs with Docker + CUDA

---

🧱 Storage Layout

• 2x Dell 60-bay JBODs (~2PB) directly attached via HBA to R720XD #1
• Media content and daily snapshots backed up to R720XD #2
• ZFS used for data integrity and snapshots

---

🛡️ Network & Access

• pfSense handles WAN, LAN, VLANs, DNS, DHCP
• Tailscale provides secure remote access across devices
• VLANs isolate traffic (Management, Storage, AI, Media, Remote Access)
• Management via Proxmox Web UI + SSH

---

🧪 Use Cases

• Infrastructure prototyping
• Script development + automation (Bash, Python)
• AI experimentation (text, vision models)
• Storage benchmarking
• Homelab documentation for GitHub

---

🔗 Future Plans

• Deploy Kubernetes across GPU + storage nodes
• Implement Grafana/Prometheus for advanced monitoring
• Publish Terraform modules for full infrastructure replication
• Public GitHub repo for open-source homelab playbooks