VPS Setup Guide¶

Deploy the full DataEngineX platform on a single VPS using K3s.

Prerequisites¶

VPS Provider: Hetzner Cloud (recommended), DigitalOcean, or any Linux VPS
Specs: Minimum 4 vCPU, 8 GB RAM, 80 GB SSD
OS: Ubuntu 22.04 or Debian 12
Domain: Registered domain with Cloudflare DNS (optional)
SSH Key: Ed25519 key pair (ssh-keygen -t ed25519)

Quick Start¶

# 1. Clone the infradex repo
git clone https://github.com/TheDataEngineX/infradex.git
cd infradex

# 2. Copy and edit configuration
cp ansible/inventory/hosts.example.yml ansible/inventory/hosts.yml
# Edit hosts.yml with your VPS IP and SSH key path

# 3. Run the one-click deploy
./scripts/one-click-deploy.sh

Manual Setup¶

1. Provision VPS (Hetzner)¶

# Install hcloud CLI
brew install hcloud  # macOS
# or: snap install hcloud

# Create server
hcloud server create \
  --name dex-server \
  --type cx31 \
  --image ubuntu-22.04 \
  --ssh-key your-key-name \
  --location nbg1

2. Install K3s¶

ansible-playbook -i ansible/inventory/hosts.yml ansible/playbooks/install-k3s.yml

3. Deploy Platform¶

# Set kubeconfig
export KUBECONFIG=./kubeconfig

# Install core services
helm install dataenginex helm/charts/dataenginex -f helm/values/values-vps.yaml
helm install dex-studio helm/charts/dex-studio -f helm/values/values-vps.yaml

4. Setup Monitoring¶

ansible-playbook ansible/playbooks/setup-monitoring.yml

5. Configure DNS (Optional)¶

cd terraform/modules/dns-cloudflare
terraform init
terraform apply \
  -var="zone_id=YOUR_ZONE_ID" \
  -var="domain=dataenginex.dev" \
  -var="subdomain=api" \
  -var="ip_address=YOUR_VPS_IP" \
  -var="api_token=YOUR_CF_TOKEN"

Verification¶

# Check all pods are running
kubectl get pods -A

# Check service health
curl http://YOUR_VPS_IP:17000/health   # dataenginex

Resource Budget (8 GB VPS)¶

Component	CPU	Memory
K3s system	250m	512Mi
dataenginex	500m	512Mi
dex-studio	100m	128Mi
PostgreSQL	250m	256Mi
Redis	100m	128Mi
Prometheus	100m	256Mi
Grafana	50m	64Mi
Total	1.35	~1.8Gi

Leaves ~6 GB headroom for spikes and OS.