My Homelab Grew Up: From Hobby to Production Infrastructure
Three years ago my homelab was a Raspberry Pi running Pi-hole and a Plex server. Today it's a multi-node cluster handling production workloads for paying clients. Here's how that happened, and the mistakes I made getting there.
The Turning Point
The shift happened when I realized my homelab skills were directly billable. I was already running Traefik, Authentik for SSO, Docker for everything. When a client needed a containerized app deployed with proper TLS and auth, I didn't need to learn anything new. Just did what I'd been doing at home, but on their cloud account.
That changed how I invested. Every tech choice became a dual-purpose decision: does this solve my home need AND teach me something I can sell?
The Stack
Proxmox for virtualization, which lets me spin up and tear down environments fast. On top of that:
- Traefik for ingress routing with automatic Let's Encrypt certs
- Authentik for SSO across everything
- Docker Compose for deployment (tried K8s, overkill for my scale)
- Tailscale for secure remote access without open ports
- Ansible so I can rebuild from scratch
Mistakes I Made
Over-engineering early. Two weeks setting up Kubernetes for 20 containers. K8s is incredible technology that I absolutely did not need. Sometimes boring is better.
Ignoring backups. Lost a database because I thought RAID was enough. RAID is not a backup. Everything critical now gets copied off-site automatically.
Skipping monitoring. A container failed silently for three days before I noticed. Now I have Grafana dashboards and alerts on everything. If a service goes down, I know in minutes.
What Made It Professional-Grade
Three things:
- Reproducibility. Every config is in code. If my server dies tomorrow, I can rebuild the whole stack from Ansible playbooks and Compose files. No snowflakes.
- Monitoring. Problems at 3 AM come as alerts, not 9 AM client emails.
- Documentation. Not for current-me. For six-months-from-now-me who won't remember why Traefik is configured that way.
The ROI
Total cost: about $50/month in electricity, $30/month in cloud hosting for the stuff that needs public uptime. For that I get a dev environment that mirrors production, hands-on experience with everything I recommend to clients, self-hosted alternatives to expensive SaaS, and a portfolio piece that shows I actually know infrastructure.
If you're in DevOps or infrastructure and you don't have a homelab, you're leaving skill development on the table. Best professional investment I've made outside of formal education.