How to Provide GPU Resources

This guide explains how to become a GPU provider on WorldLand Cloud and earn WL tokens.

Prerequisites

Hardware Requirements

Component	Minimum	Recommended
GPU	NVIDIA GTX 1080+	RTX 3090/4090 or A100
VRAM	8GB	24GB+
CPU	4 cores	8+ cores
RAM	16GB	32GB+
Storage	100GB SSD	500GB NVMe
Network	100 Mbps	1 Gbps
Public IP	Required	Static IP recommended

Software Requirements

• Ubuntu 20.04/22.04 LTS
• NVIDIA Driver 525+
• Docker with NVIDIA Container Toolkit
• Kubernetes (kubeadm)

Step 1: Install NVIDIA Drivers

# Add NVIDIA repository
sudo apt-get update
sudo apt-get install -y nvidia-driver-535

# Reboot
sudo reboot

# Verify installation
nvidia-smi

Step 2: Install Docker & NVIDIA Container Toolkit

# Install Docker
curl -fsSL https://get.docker.com | sh
sudo usermod -aG docker $USER

# Install NVIDIA Container Toolkit
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/libnvidia-container/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/libnvidia-container/$distribution/libnvidia-container.list | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list

sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker

# Verify
docker run --rm --gpus all nvidia/cuda:12.0.0-base-ubuntu22.04 nvidia-smi

Step 3: Install WorldLand Provider SDK

# Download Provider SDK
wget https://github.com/worldland/releases/download/v1.0.0/worldland-provider-sdk-linux

# Make executable
chmod +x worldland-provider-sdk-linux

# Run installation
sudo ./worldland-provider-sdk-linux install

Step 4: Configure Provider

Create configuration file:

sudo nano /etc/worldland/provider.yaml

provider:
  name: "My GPU Server"
  public_ip: "123.45.67.89"

resources:
  gpu_price_per_hour: 0.5 # WL per hour
  cpu_cores: 8
  memory_gb: 32
  storage_gb: 200

wallet:
  address: "0x1234...abcd"

network:
  master_endpoint: "https://api.worldland.cloud"

Step 5: Register Provider

# Start provider agent
sudo systemctl start worldland-provider

# Check status
sudo systemctl status worldland-provider

The agent will:

1. Register with the WorldLand network
2. Receive a kubeadm join token
3. Join the Kubernetes cluster as a worker node

Step 6: Join Kubernetes Cluster

# The agent will automatically execute the join command
# Or manually run:
sudo kubeadm join <master-ip>:6443 \
  --token <token> \
  --discovery-token-ca-cert-hash sha256:<hash>

Provider Status

Status Progression

Pending → Approved → Joined → Available

Status	Description
Pending	Registration submitted, awaiting approval
Approved	Approved, join token issued
Joined	Successfully joined K8s cluster
Available	Ready to accept GPU jobs

Check Status

# View provider status
curl https://api.worldland.cloud/api/v1/providers/my-provider-id

# View node status in cluster
kubectl get nodes

Resource Management

How Resources Are Allocated

When a customer creates a job on your node:

1. GPU - Dedicated allocation via NVIDIA device plugin
2. CPU - Guaranteed QoS (Request = Limit)
3. Memory - Guaranteed QoS (Request = Limit)
4. Storage - Ephemeral storage limit

Capacity Tracking

Your Node Capacity:
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
GPU:     [████████░░░░] 2/4 used
CPU:     [██████░░░░░░] 12/24 cores used
Memory:  [████████░░░░] 48/64 GB used
Storage: [████░░░░░░░░] 80/200 GB used
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Earnings

How You Earn

• Per-hour billing - Customers pay WL per hour of GPU usage
• 90% to provider - You receive 90% of the service fee
• 10% to protocol - Network maintenance fee

Example Earnings

GPU	Price/Hour	Daily (100% util)	Monthly
RTX 4090	0.50 WL	12 WL	360 WL
RTX 3090	0.35 WL	8.4 WL	252 WL
A100	1.00 WL	24 WL	720 WL

Monitoring

Provider Dashboard

Access your provider dashboard at:

https://cloud.worldland.io/provider/dashboard

Features:

• Real-time GPU utilization
• Active jobs
• Earnings history
• Resource availability

Command Line

# View running GPU pods
kubectl get pods -l worldland.io/gpu-rental=true

# View resource usage
kubectl top nodes

# View GPU allocation
kubectl describe node <your-node-name> | grep nvidia.com/gpu

Maintenance

Graceful Maintenance Mode

Before maintenance:

# Cordon node (prevent new jobs)
kubectl cordon <your-node-name>

# Wait for existing jobs to complete
kubectl get pods -l worldland.io/gpu-rental=true --field-selector spec.nodeName=<your-node>

# Perform maintenance
# ...

# Uncordon node
kubectl uncordon <your-node-name>

Updating Drivers

# Cordon node first
kubectl cordon <your-node-name>

# Update NVIDIA driver
sudo apt-get update
sudo apt-get install -y nvidia-driver-545

# Reboot
sudo reboot

# Verify and uncordon
nvidia-smi
kubectl uncordon <your-node-name>

Troubleshooting

Node Not Joining

# Check agent logs
sudo journalctl -u worldland-provider -f

# Check kubelet status
sudo systemctl status kubelet

# Check network connectivity
curl -v https://api.worldland.cloud/health

GPU Not Detected in Cluster

# Check NVIDIA device plugin
kubectl get pods -n kube-system | grep nvidia

# Check node labels
kubectl get node <your-node> -o yaml | grep nvidia

Jobs Not Scheduling

# Check node taints
kubectl describe node <your-node> | grep Taints

# Check node conditions
kubectl describe node <your-node> | grep Conditions -A 10

Best Practices

Maximizing Earnings

1. High availability - Keep uptime > 99%
2. Competitive pricing - Research market rates
3. Quality hardware - Better GPUs attract more customers
4. Fast network - Low latency improves experience

Security

1. Keep system updated
2. Use firewall rules
3. Monitor for unusual activity
4. Regular security audits

Next Steps

• Provider Policy - Terms and requirements
• Provider Rewards - Detailed reward structure