Homelab on Homelab Documentation

Storage Architecture

Mon, 01 Jan 0001 00:00:00 +0000

Storage Architecture#

Philosophy#

Start simple. No NAS/RAID until you have 3+ drives. Syncthing-first sync, thunderbolt cold backups, snapraid for phase 2.

Phase 1: Syncthing-First (What You Have Now)#

Setup#

hydrogen (2TB NVMe) ←→ helium (2TB NVMe + extra) ←→ lithium (2TB NVMe)

Install syncthing on all three boxes:

# Ubuntu/Debian
sudo apt install syncthing

# Enable auto-start
systemctl enable --now syncthing@$(whoami)

Configure sync folders:

/code — active development projects
/configs — dotfiles, service configs
/models — LLM models (large, sync selectively)

How it works:

Network Topology

Mon, 01 Jan 0001 00:00:00 +0000

Network Topology#

VLAN Layout#

VLAN ID	Name	Subnet	Purpose
10	management	10.0.10.0/24	Servers, NAS, infrastructure
20	workstations	10.0.20.0/24	Developer machines
30	iot	10.0.30.0/24	Smart home devices
40	guests	10.0.40.0/24	Guest Wi-Fi (isolated)

DNS Configuration#

Pi-hole on hydrogen (10.0.10.10):

# /etc/dnsmasq.d/custom.conf
server=1.1.1.1
server=8.8.8.8
local=/mrzk.io/
address=/hydrogen.mrzk.io/10.0.10.10
address=/helium.mrzk.io/10.0.10.20
address=/lithium.mrzk.io/10.0.10.30

Traefik Reverse Proxy#

Docker Compose (helium):

version: '3.8'
services:
 traefik:
 image: traefik:v2.10
 ports:
 - "80:80"
 - "443:443"
 volumes:
 - /var/run/docker.sock:/var/run/docker.sock
 - ./traefik.yml:/etc/traefik/traefik.yml
 - ./acme.json:/acme.json
 labels:
 - "traefik.enable=true"
 - "traefik.http.routers.traefik.rule=Host(`traefik.mrzk.io`)"
 - "traefik.http.routers.traefik.tls.certresolver=letsencrypt"

traefik.yml:

Container Deployment

Mon, 01 Jan 0001 00:00:00 +0000

Container Deployment#

Docker Compose Stacks#

Core Services (helium)#

docker-compose.yml:

version: '3.8'
services:
 # Authentication
 authentik:
 image: ghcr.io/goauthentik/server:latest
 ports:
 - "9000:9000"
 - "9443:9443"
 environment:
 - AUTHENTIK_REDIS__HOST=redis
 - AUTHENTIK_POSTGRESQL__HOST=postgres
 - AUTHENTIK_POSTGRESQL__USER=authentik
 - AUTHENTIK_POSTGRESQL__PASSWORD=changeme
 - AUTHENTIK_POSTGRESQL__NAME=authentik
 volumes:
 - ./authentik/media:/media
 - ./authentik/custom-templates:/templates
 depends_on:
 - redis
 - postgres

 redis:
 image: redis:alpine
 volumes:
 - redis_data:/data

 postgres:
 image: postgres:15-alpine
 environment:
 - POSTGRES_USER=authentik
 - POSTGRES_PASSWORD=changeme
 - POSTGRES_DB=authentik
 volumes:
 - postgres_data:/var/lib/postgresql/data

 # File sharing
 nextcloud:
 image: nextcloud:latest
 ports:
 - "8081:80"
 environment:
 - MYSQL_HOST=mysql
 - MYSQL_DATABASE=nextcloud
 - MYSQL_USER=nextcloud
 - MYSQL_PASSWORD=changeme
 volumes:
 - nextcloud_data:/var/www/html
 depends_on:
 - mysql

 mysql:
 image: mysql:8.0
 environment:
 - MYSQL_ROOT_PASSWORD=changeme
 - MYSQL_DATABASE=nextcloud
 - MYSQL_USER=nextcloud
 - MYSQL_PASSWORD=changeme
 volumes:
 - mysql_data:/var/lib/mysql

 # Code hosting
 gitea:
 image: gitea/gitea:latest
 ports:
 - "3000:3000"
 - "222:22"
 volumes:
 - gitea_data:/data
 - /etc/timezone:/etc/timezone:ro
 - /etc/localtime:/etc/localtime:ro

 # Monitoring
 prometheus:
 image: prom/prometheus:latest
 ports:
 - "9090:9090"
 volumes:
 - ./prometheus.yml:/etc/prometheus/prometheus.yml
 - prometheus_data:/prometheus

 grafana:
 image: grafana/grafana:latest
 ports:
 - "3000:3000"
 volumes:
 - grafana_data:/var/lib/grafana
 environment:
 - GF_SECURITY_ADMIN_PASSWORD=changeme

 # UniFi Controller
 unifi:
 image: lscr.io/linuxserver/unifi-network-application:latest
 ports:
 - "8443:8443"
 - "3478:3478/udp"
 - "8080:8080"
 volumes:
 - unifi_data:/config

volumes:
 redis_data:
 postgres_data:
 nextcloud_data:
 mysql_data:
 gitea_data:
 prometheus_data:
 grafana_data:
 unifi_data:

NVIDIA Container Toolkit#

Install on helium:

Inference Stack

Mon, 01 Jan 0001 00:00:00 +0000

Inference Stack#

llama.cpp Server (lithium)#

Installation#

# Clone and build
git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
make -j8

# Or use prebuilt
cargo install llama-cpp-server

Server Configuration#

server.yml:

host: 0.0.0.0
port: 8080
model: /models/qwen3.5-27b-q5_k_m.gguf
n_ctx: 8192
n_batch: 512
n_ubatch: 512
n_threads: 12
n_gpu_layers: 50
flash_attn: true
cache_type_k: q8_0
cache_type_v: f16

Run:

./server -c server.yml

API Endpoints#

Chat completion:

curl http://lithium.mrzk.io:8080/completion \
 -H "Content-Type: application/json" \
 -d '{
 "prompt": "user: Hello\nassistant:",
 "n_predict": 512,
 "temperature": 0.7,
 "stop": ["user:", "</s>"]
 }'

Embeddings: