Building My Ultimate Home Server: From Old Laptop to Cloud Lab
Command Palette
Search for a command to run...
Comments
Join the discussionNo comments yet. Be the first to comment.
Search for a command to run...
No comments yet. Be the first to comment.
I’ve always been fascinated by self-hosting and running my own services. Instead of relying on third-party providers, I wanted control, security, and the ability to experiment with real-world DevOps practices at home.
So, I turned an old laptop into a Proxmox-based home server and gradually expanded it into a full stack of self-hosted services. Along the way, I also built a small Kubernetes lab for learning cloud-native concepts.
Centralized storage for personal files, photos, and backups.
Self-hosting services like Nextcloud, Plex, and Home Assistant.
Hands-on DevOps learning with Docker, Kubernetes, Terraform, and Ansible.
Cost efficiency compared to cloud hosting.
Complete control over security and networking.
Base machine: Old laptop with 8-core CPU, 16 GB RAM, 2x NVMe SSDs + external drives.
Hypervisor: Proxmox VE for virtualization and VM/LXC management.
Storage: ZFS pools for redundancy + backups synced to external drives.
I split services into multiple VMs and containers for better isolation and flexibility:
Nextcloud → Personal cloud storage & file sharing.
Immich → AI-powered photo and video management.
Home Assistant → Smart home automation hub.
NGINX Proxy Manager → Reverse proxy with SSL and hostname-based routing.
Authentik → Single sign-on and access control for all services.
Media Stack → Plex/Emby + Arr apps (Sonarr, Radarr, etc.) with qBittorrent.
Storage VM → Shared storage LXC for other services.
Tailscale VPN → Secure remote access to all services.
Reverse proxy + SSL → Clean, secure access with custom hostnames.
Isolated networks in Proxmox → Separate traffic between critical and non-critical workloads.
To practice cloud-native concepts, I also built a K3s cluster:
3 control plane nodes + 3 worker nodes (all on lightweight VMs).
Deployed workloads using Helm and Terraform.
Explored concepts like scaling, service discovery, and observability.
This cluster doubles as both a learning environment and a testbed for containerized apps before production-like deployment.
Local redundancy using ZFS snapshots and RAID setups.
Cloud sync to AWS S3 for disaster recovery.
Automated scripts for scheduled backups and restores.
Virtualization with Proxmox makes it easy to isolate and manage services.
Proper networking (VPN + reverse proxy) is key for security.
Kubernetes at home is not overkill—it’s a great way to learn in a safe environment.
Always plan backup and recovery before running production-like services.
I plan to expand my cluster with mini PCs for high availability, add monitoring via Prometheus + Grafana, and experiment with advanced GitOps workflows using ArgoCD.
What started as a simple home server project on an old laptop became a complete self-hosting ecosystem and learning lab. It gives me real-world DevOps experience, full control of my data, and the satisfaction of building something reliable from scratch.
If you’ve ever thought about running your own server, you don’t need enterprise hardware—just curiosity, patience, and the willingness to experiment.