Self-Hosted Dev Platform Guide
This guide walks you from bare metal or a fresh VPS to a fully working self-hosted platform: source control, project management, CI/CD, AI-assisted development, and monitoring.
No prior Kubernetes experience is required. The guide offers two tracks — Docker-only and Kubernetes (K3s) — so you can start where your skills are and grow from there.
Who this is for
- Developers who want full control over their tools and data.
- Hobbyists building a home lab with real dev infrastructure.
- Small teams who want to run their own Gitea, YouTrack, and CI/CD stack without paying SaaS prices.
The two deployment tracks
Choose one track when you reach the Choose Your Deployment Path page:
| Track | Best for | Complexity |
|---|---|---|
| Docker-only | Single-node setups, simpler operations, getting started fast | Lower |
| Kubernetes (K3s) | Multi-node clusters, GitOps, production-grade resilience | Higher |
Sections that differ between tracks present both options side by side.
Guide sections
Work through these sections in order:
- Getting Started — prerequisites, OS setup, and choosing your deployment track
- Infrastructure Foundation — Docker, DNS, reverse proxy, TLS
- Kubernetes (K3s) — cluster setup, storage, GitOps (K3s track)
- Source Control — Gitea — install and configure Gitea
- Project Management — YouTrack — install and configure YouTrack
- CI/CD Pipeline — Gitea Actions, container registry, deployment
- AI-Assisted Development — agent pipeline, LLM setup, Claude Code
- Monitoring and Observability — Prometheus, Grafana, alerting
- Maintenance and Operations — backups, upgrades, incident response
Start with Getting Started.
What the platform includes
When you finish the guide, you will have a running self-hosted platform with:
- Gitea — Git hosting and Gitea Actions CI/CD
- YouTrack — Issue tracking and project management
- Container registry — Nexus or Distribution Registry
- Traefik — Reverse proxy with automatic TLS
- Prometheus + Grafana — Metrics and dashboards (K3s track)
- Flux CD — GitOps deployments (K3s track)
- AI agent pipeline — Automated issue-to-PR workflow with Claude Code