🧪 AWS Homestyle Lab (Proxmox LXC)

A lightweight homelab simulating AWS-style infrastructure using Proxmox LXC containers.

Built for learning, testing, breaking, and rebuilding — without burning cloud credits.

---

🎯 Objective

To replicate a simplified cloud architecture locally:

• App servers → EC2
• HAProxy → Load Balancer (ELB)
• PostgreSQL → RDS
• Ansible → Configuration management
• Netdata → Infrastructure monitoring

---

🏗️ Architecture

Client (Laptop) ↓ HAProxy (lb-haproxy:80) ↓

| | | | app1 app2 app3 (round robin) ↓ PostgreSQL (db-postgres:5432)

Monitoring:

• Netdata agents installed on all nodes
• Centralised monitoring via lb-haproxy

---

🖥️ Nodes

Node	Role	IP Address	Port
lb-haproxy	Load Balancer	192.168.x.x	80
app1	App Server	192.168.x.x	3000
app2	App Server	192.168.x.x	3000
app3	App Server	192.168.x.x	3000
db-postgres	Database (Postgres)	192.168.x.x	5432

---

⚙️ Tech Stack

• Proxmox (LXC containers)
• HAProxy (Load balancing)
• Nginx / App service (port 3000)
• PostgreSQL
• Ansible (automation & roles)
• Netdata (monitoring)

---

✅ What Has Been Verified

• Static IPs configured for all containers
• SSH connectivity working across all nodes
• HAProxy round-robin working across app1, app2, and app3
• PostgreSQL reachable on port 5432
• Ansible inventory aligned with current infrastructure
• Health check script created for quick validation
• Centralised Netdata monitoring working across all 5 nodes

---

🔁 Load Balancing

Round-robin verified:

app1 → app2 → app3 → app1 → app2 → app3

HAProxy routes traffic correctly across all backend nodes.

---

🔍 Health Check

Run:

./scripts/check_lab.sh

Checks:

• Host reachability (ping)
• SSH access
• Load balancer rotation
• Database port availability
• Optional direct app access

---

📊 Monitoring

Netdata is installed on all nodes.

Current setup:

• lb-haproxy acts as the Netdata parent
• app1, app2, app3, and db-postgres stream metrics to the parent
• A central dashboard is available from the parent node

Monitoring currently provides:

• Per-node CPU usage
• Memory usage
• Load
• Disk activity
• Network traffic

---

🧠 Notes

• Static IP is configured via Proxmox, not inside the containers
• SSH access uses root
• Backend app ports (3000) are primarily intended for internal access via the load balancer
• Netdata uses a parent-child streaming model for centralised visibility
• Lab is designed for deploy → test → destroy → repeat

---

📦 Backups

Configs stored locally under:

/backups

Includes:

• HAProxy config
• SSH configs
• PostgreSQL configs

---

📸 Proof

See /screenshots for:

• HAProxy round robin test
• Lab validation outputs
• Netdata multi-node monitoring view

---

🚀 Next Steps

• Refine Ansible roles and playbooks further
• Add alerting for resource thresholds
• Expand deployment workflows
• Simulate node failure and recovery under monitoring
• Optionally add Grafana later for aggregated visual dashboards

---

💭 Reflection

This homelab started as a simple setup, but quickly turned into a full learning playground.

“Just one more check…” → 4 hours later.