Deploying a Self-Hosted AI Assistant on Ubuntu 24.04: Full Integration Guide for Large Language Models

Oct 161 views

This guide provides step-by-step instructions for deploying a self-hosted AI assistant platform on Ubuntu 24.04 LTS, covering system requirements, core platform installation, LLM service integration, security configurations, and maintenance best practices.

System Requirements & Setup

Server Specifications

Recommended Ubuntu 24.04 LTS configuration:

CPU: 4 cores (8+ preferred for production)
Memory: 16GB RAM (32GB for high concurrency)
Storage: 100GB+ SSD-based disk space
Network: Static public IP or VPN solution

Security Recommendation: While root access is used in examples, create dedicated users for production:

sudo useradd -m aiuser && sudo passwd aiuser
sudo usermod -aG sudo aiuser

Node.js Environment

Install using nvm for version management:

# Install dependencies
sudo apt update && sudo apt install -y curl git
# Install nvm
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.7/install.sh | bash
source ~/.bashrc
# Install LTS Node.js
nvm install 22
nvm alias default 22
node -v  # Should show v22.x.x

Core Platform Installation

Global Installation

npm install -g openclaw@latest --unsafe-perm=true --allow-root

Parameters Explained:

--unsafe-perm: Resolves permission issues
--allow-root: Permits root installation (avoid in production)

Troubleshooting:

For ENOTEMPTY errors:

rm -rf /root/.npm-global/lib/node_modules/openclaw
npm cache clean --force

Network timeouts: Use registry mirrors:

npm config set registry https://registry.npmjs.org/

Verify Installation

openclaw --version  # Should show version like 2026.3.2

LLM Service Integration

Obtain API Credentials

Register with your preferred LLM provider
Create dedicated project
Generate API key (note the sk- prefix)
Select nearest service region

Configuration Wizard

openclaw config

Interactive setup steps:

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

Set API endpoint (example format):

https://api.example.com/v1/chat/completions

Paste API key (mask sensitive characters)
Test connection (optional)

Advanced Configuration (config.json):

{
  "providers": {
    "zhipu-glm": {
      "apiBase": "https://api.example.com/v1",
      "apiKey": "sk-xxxxxxxxxxxxxxxx",
      "maxTokens": 4096,
      "temperature": 0.7
    }
  }
}

Secure Access Methods

SSH Tunneling

ssh -L 3000:localhost:3000 username@server_ip -N

Access via http://localhost:3000

Advantages:

No exposed web ports
End-to-end encryption
Network-agnostic

Nginx Reverse Proxy (Production)

Install Nginx:
```
sudo apt install -y nginx
```

Configure SSL (Let’s Encrypt example):

sudo apt install -y certbot python3-certbot-nginx
sudo certbot --nginx -d yourdomain.com

Proxy configuration:

server {
 listen 443 ssl;
 server_name yourdomain.com;
 location / {
     proxy_pass http://localhost:3000;
     proxy_set_header Host $host;
     proxy_set_header X-Real-IP $remote_addr;
 }
 ssl_certificate /etc/letsencrypt/live/yourdomain.com/fullchain.pem;
 ssl_certificate_key /etc/letsencrypt/live/yourdomain.com/privkey.pem;
}

Maintenance Best Practices

Log Management

Configure log rotation:

sudo mkdir /var/log/openclaw
sudo chown aiuser:aiuser /var/log/openclaw

Create /etc/logrotate.d/openclaw:

/var/log/openclaw/*.log {
    daily
    missingok
    rotate 7
    compress
    delaycompress
    notifempty
    create 640 aiuser aiuser
}

Performance Optimization

For high concurrency:

Adjust Node.js memory:

export NODE_OPTIONS="--max-old-space-size=8192"

Enable clustering (if supported):
```
openclaw start --workers 4
```

Backup Strategy

Regularly backup:

Configuration: ~/.openclaw/
Model caches (if applicable)
External databases

Troubleshooting Guide

Common Errors

Code	Cause	Solution
500	Model service error	Check provider status page
403	Invalid credentials	Regenerate API key
429	Rate limiting	Adjust concurrency or upgrade plan
ECONNREFUSED	Network issue	Verify firewall rules

Diagnostic Commands

# Service status
systemctl status openclaw  # If using systemd
# Real-time logs
journalctl -u openclaw -f
# Network test
curl -v https://api.example.com/health

Advanced Features

Multi-Model Routing

{
  "routing": {
    "default": "zhipu-glm",
    "rules": [
      {
        "pattern": "^/tech/",
        "provider": "tech-specialized-model"
      }
    ]
  }
}

Plugin Development

Example audit plugin:

// ~/.openclaw/plugins/audit.js
module.exports = {
  preProcess: (req) => {
    console.log(`Request from ${req.ip}`);
  },
  postProcess: (res) => {
    // Log response data
  }
};

Upgrade & Rollback

Version Upgrade

# Check for updates
npm outdated -g openclaw
# Upgrade
npm update -g openclaw

Rollback Procedure

# Backup current version
cp -r /usr/local/lib/node_modules/openclaw /usr/local/lib/node_modules/openclaw.bak
# Install specific version
npm install -g openclaw@2026.3.1

This deployment framework enables rapid establishment of self-hosted AI assistant platforms on Ubuntu 24.04 while maintaining flexibility for enterprise requirements. The modular architecture supports seamless integration with various LLM providers while implementing robust security and monitoring controls. For production environments, we recommend implementing comprehensive monitoring alerts and regular security audits.