Custom Ubuntu Server ISO for AI Infrastructure

This directory contains the complete custom Ubuntu Server ISO with embedded AI infrastructure configuration from srv-AI-01.

🎯 Project Status: ✅ COMPLETED

Your custom Ubuntu Server AI ISO has been successfully created and is ready for deployment!

📁 Clean Project Structure

🔧 Build Scripts

• manual-build.sh - ISO builder script (tested and working)
• enhanced-ai-setup.sh - AI configuration generator
• preseed.cfg - Automated installation configuration
• package-list.txt - Additional packages specification

📦 ISO Files

• ubuntu-24.04.2-live-server-amd64.iso - Original Ubuntu Server ISO (3.1GB)
• ubuntu-24.04.2-server-ai-amd64.iso - YOUR CUSTOM AI SERVER ISO (3.1GB) ✅
• ubuntu-24.04.2-server-ai-amd64.iso.md5 - MD5 checksum
• ubuntu-24.04.2-server-ai-amd64.iso.sha256 - SHA256 checksum

📚 Documentation

• README.md - Project overview and quick start
• HOW-TO-USE.md - Complete installation and usage guide

🚀 What's Included in Your Custom ISO

🤖 AI Server Features

• Complete srv-AI-01 Configuration: All Docker services, configs, and scripts embedded
• Automated Installation: Preseed configuration for unattended setup
• AI-Optimized Partitioning: Dedicated storage for Docker and AI data
• GPU Ready: NVIDIA drivers and container toolkit included
• Security Hardened: Firewall rules and SSH configuration
• System Optimized: Memory management and network tuning for AI workloads

🛠️ Embedded AI Services (Enhanced Configuration)

Original srv-AI-01 Services:

• Ollama: AI model server with GPU support
• Open WebUI: Web interface for AI models
• n8n: Workflow automation platform
• Milvus: Vector database for RAG
• Langfuse: LLM observability and monitoring
• Flowise: Low-code AI application builder
• PostgreSQL, Redis, MinIO: Supporting services
• Grafana, Prometheus: Monitoring and metrics

Additional Services from Local AI Packaged:

• Qdrant: High-performance vector database (alternative to Milvus)
• Neo4j: Knowledge graph engine for GraphRAG, LightRAG
• SearXNG: Privacy-focused metasearch engine (229 search services)
• Caddy: Automatic HTTPS reverse proxy with SSL
• Enhanced n8n: 400+ integrations and advanced AI components
• Enhanced Flowise: Better integration with other services
• Supabase: Open-source database and authentication (optional)

🚀 Quick Deployment

1. Write ISO to USB Drive

sudo dd if=ubuntu-24.04.2-server-ai-amd64.iso of=/dev/sdX bs=4M status=progress

(Replace /dev/sdX with your USB device)

2. Install on Target Hardware

1. Boot from USB/DVD
2. Follow Ubuntu installation process
3. Default Login: aiserver / changeme123 (CHANGE IMMEDIATELY!)

3. Complete AI Setup

After installation:

# Change default password immediately!
sudo passwd aiserver

# Run post-installation setup
sudo ~/ai-setup/post-install.sh

# Interactive AI services configuration (NEW!)
cd ~/ai-setup
./setup.sh  # 🎯 Interactive wizard with auto-generated passwords!
./manage.sh start

🆕 Interactive Setup Features:

• Auto-generates secure passwords for all services
• GPU optimization for RTX 3090 + 198GB RAM
• Simple guided prompts for configuration
• No manual .env editing required!

4. Access AI Services

After setup completion:

Core AI Services:

• Open WebUI: http://server-ip:3000 (AI chat interface)
• n8n: http://server-ip:5678 (Workflow automation with 400+ integrations)
• Langfuse: http://server-ip:3001 (LLM observability)
• Flowise: http://server-ip:3002 (Low-code AI)
• Grafana: http://server-ip:3003 (Monitoring)
• MinIO: http://server-ip:9001 (Object storage)

Additional Enhanced Services:

• Neo4j Browser: http://server-ip:7474 (Knowledge graph)
• Qdrant Dashboard: http://server-ip:6333/dashboard (Vector DB)
• SearXNG: http://server-ip:8080 (Private search)
• Prometheus: http://server-ip:9090 (Metrics)
• Caddy: http://server-ip (Reverse proxy dashboard)

💻 Target System Requirements

Minimum Requirements

• CPU: AMD Ryzen 7 or Intel i7 (8+ cores)
• RAM: 32GB DDR4 (64GB+ recommended)
• GPU: NVIDIA RTX 3060 or better (12GB+ VRAM)
• Storage: 1TB NVMe SSD (2TB+ recommended)
• Network: Gigabit Ethernet

Your Optimal Setup ✅

• CPU: AMD Ryzen 9 ✅ (Perfect)
• RAM: 198GB ✅ (Excellent for large models)
• GPU: NVIDIA RTX 3090 ✅ (24GB VRAM - Ideal)
• Storage: 16TB ✅ (Plenty for models and data)

🔧 Utility Commands

Validate Your ISO

./validate-iso.sh

Check Build Status

./check-build-status.sh

Rebuild ISO (if needed)

sudo ./manual-build.sh

📋 File Checksums

MD5: 750a8f699aa002c533c86e5cc3c5368d
SHA256: 7109d22691476dcb192688f819d5d98bddfbb92a8d47245c323c4a15f2631428

🔒 Security & Troubleshooting

Security Features Included

• UFW firewall with AI service ports configured
• SSH hardening and key-based authentication ready
• Container isolation for all services
• System optimizations for AI workloads

Security Checklist

• Change default password: sudo passwd aiserver
• Update service passwords in ~/ai-setup/.env
• Configure SSL certificates for production
• Review firewall rules for your network

Common Issues & Solutions

GPU not detected: Verify NVIDIA GPU present and compatible

lspci | grep -i nvidia  # Check if GPU detected
sudo apt install nvidia-driver-535  # Reinstall drivers if needed

Services won't start: Check Docker and AI services

sudo systemctl status docker
cd ~/ai-setup && ./manage.sh status
./manage.sh logs <service-name>  # Check specific service logs

Boot issues: Verify ISO integrity using provided checksums

---

🎉 Project Complete!

Your custom Ubuntu Server AI ISO is ready for deployment on your Ryzen 9 + RTX 3090 system with 198GB RAM. This ISO includes everything needed for a complete AI infrastructure deployment.

🚀 Happy AI Building!

For additional support, refer to the srv-AI-01 documentation or create detailed issue reports.