Prerequisites and Environment Setup

System Requirements

• Recommended OS: Ubuntu 24.04 LTS (LTS recommended for stability)
• Hardware: Minimum 4-core CPU with 8GB RAM (2-core/4GB for testing)
• User Access: Root privileges for setup (create dedicated user with sudo access for production)

Required Toolchain

# Node.js 22 LTS (recommended via nvm)
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.7/install.sh | bash
source ~/.bashrc
nvm install 22
nvm alias default 22
# Verify installations
node -v  # Should show v22.x.x
npm -v   # Should show 9.x.x+
# Additional tools
sudo apt install git curl wget -y

LLM Service Credentials

• Obtain API access key from your cloud provider’s LLM service
• Store credentials in environment variables (e.g., OPENCLAW_MODEL_KEY)
• Avoid hardcoding credentials in configuration files

OpenClaw Framework Installation

Install Node.js Modules

# Global installation with permission handling
npm install -g openclaw@latest --unsafe-perm
# Verify installation
openclaw --version  # Should show 2026.x.x+

Troubleshooting Installation Issues

• ENOTEMPTY Errors: Clean previous installations with:

  rm -rf $(npm root -g)/openclaw
  npm cache clean --force

• Permission Problems: Configure alternative global install path:

  npm config set prefix '~/.npm-global'

LLM Service Configuration

Interactive Setup Wizard

openclaw config

Configuration Parameters

1. Provider Setup:

   • Select “Add custom provider”
   • Enter provider name (e.g., zhipu-provider)

2. API Configuration:

   • Endpoint URL: https://api.example.com/v4 (replace with actual endpoint)
   • Authentication: API Key mode
   • Key Management: Use environment variable injection

3. Advanced Settings (optional):

   • Timeout: Increase from default 30s to 60s for unstable networks
   • Retry Policy: Configure 3 retry attempts
   • Concurrency: Set max concurrent requests based on server capacity

Configuration File Location:

~/.openclaw/config.json

Web Service Deployment

Service Startup

openclaw start --port 3000 --host 0.0.0.0

Secure Local Access (SSH Tunneling)

# Forward server port 3000 to local port 8080
ssh -L 8080:localhost:3000 username@server_ip -N

• Access via browser: http://localhost:8080

Production Security Recommendations

• Configure Nginx reverse proxy
• Enable HTTPS with Let’s Encrypt certificates
• Restrict access via firewall rules (e.g., ufw or cloud provider security groups)

Operations and Maintenance

Log Management

• Service logs: /var/log/openclaw/service.log
• Access logs: /var/log/openclaw/access.log

Performance Monitoring

# Monitor resource usage
top -p $(pgrep -f openclaw)

Common Issues Resolution

• Model API failures: Verify key validity and network connectivity
• Unresponsive service: Check for OOM errors in logs
• Port conflicts: Modify startup port or terminate conflicting processes

Advanced Features

Multi-Model Support

Add multiple provider entries in config.json and implement model routing through API endpoints.

Custom Routing Configuration

{
  "/api/v1/chat": {
    "provider": "zhipu-provider",
    "model": "glm-4.7-flash"
  }
}

Rate Limiting

Implement QPS restrictions through middleware to prevent service abuse.

Production Deployment Best Practices

Containerization

FROM node:22-alpine
WORKDIR /app
COPY . .
RUN npm install --production
CMD ["openclaw", "start"]

High Availability Architecture

• Deploy multiple instances using Kubernetes
• Configure health checks and auto-restart policies
• Implement load balancing for request distribution

Backup Strategy

• Regularly backup configuration files and model data
• Store important logs in object storage services

This guide provides comprehensive coverage from initial environment setup to production deployment. Always validate configurations in testing environments before production migration, adjusting parameters according to specific business requirements. Refer to the official documentation for the latest framework updates and compatibility information.