Skip to content

Installation Overview

This section provides comprehensive installation instructions for MAESTRO across different platforms and configurations.

Quick Installation

All platforms follow the same basic steps:

  1. Install Docker (Docker Desktop for Windows/macOS, Docker Engine for Linux)
  2. Clone the repository
  3. Run setup script (setup-env.sh or setup-env.ps1)
  4. Start with Docker Compose

Docker Desktop handles all complexity on Windows and macOS!

Choose Your Installation Method

By Operating System

Special Configurations

System Requirements

Minimum Requirements

  • Operating System:
    • Linux (any modern distribution)
    • macOS (Intel or Apple Silicon)
    • Windows 10/11 (Docker Desktop handles WSL2 automatically)
  • RAM: 16GB minimum
  • Storage: ~30GB free space
  • CPU: 8 cores or more
  • Docker: Version 20.10 or higher
  • Docker Compose: Version 2.0 or higher
  • RAM: 32GB or more
  • Storage: 50GB+ (8GB models + space for document libraries)
  • CPU: 12 cores or more
  • GPU: NVIDIA GPU with CUDA support (strongly recommended)
    • 4GB VRAM minimum for single process operation
    • 8GB VRAM recommended for concurrent research + document processing
    • BGE-M3 (560M params): ~1.3GB VRAM per instance
    • BGE-Reranker-v2-m3: ~1.2GB VRAM
    • MultiGPU system recommended if also hosting LLMs locally
  • Network: Stable broadband connection for web search

Pre-Installation Checklist

Before proceeding with installation:

  1. Install Docker

    Windows Users

    Docker Desktop automatically configures WSL2 backend - no manual WSL setup needed!

  2. Verify Docker Installation

    docker --version
    docker compose version
    

  3. Obtain API Keys You'll need at least one AI provider API key:

  4. Check Available Ports

    MAESTRO uses these default ports:

    • Port 80: Web interface (nginx)
    • Port 5432: PostgreSQL database

Installation Methods Comparison

Method GPU Support Complexity Performance Best For
Linux Full Easy Excellent Reliable performance
macOS No (CPU only) Easy Good Reasonable performance
Windows (Docker Desktop) Limited Easy Good Windows users
CPU Mode No Easy Moderate Systems without GPU

Windows Installation Simplified

Docker Desktop makes Windows installation as easy as macOS - it handles all WSL2 configuration automatically!

Docker Compose Profiles

MAESTRO supports different deployment profiles:

Default Profile

docker compose up -d --build
Runs with automatic GPU detection if available.

CPU-Only Profile

docker compose -f docker-compose.cpu.yml up -d --build
Forces CPU mode, useful for systems without GPU.

Network Configuration Options

The setup script offers three network configurations:

1. Simple (Localhost Only)

  • Access only from the same machine
  • URL: http://localhost
  • Most secure option

2. Network Access

  • Access from other devices on your network
  • URL: http://[your-ip-address]
  • Good for team deployments

3. Custom Domain

  • Use with reverse proxies
  • Custom URLs like http://maestro.local
  • Best for set-and-forget deployments

Post-Installation Steps

After successful installation:

  1. Verify All Services Running

    docker compose ps
    

  2. Check Logs for Errors

    docker compose logs
    

  3. Access Web Interface

    • Navigate to configured URL
    • Default: http://localhost
  4. Login with Default Credentials

    • Username: admin
    • Password: admin123
  5. Immediate Security Steps

    • Change default password
    • Configure user accounts
    • Set up API keys

Troubleshooting Quick Reference

Service Won't Start

  • Check Docker is running
  • Verify ports are available
  • Review logs: docker compose logs [service-name]

Cannot Access Web Interface

  • Ensure all services are running
  • Check firewall settings
  • Verify network configuration

Database Connection Issues

  • PostgreSQL container must be healthy
  • Check database credentials in .env
  • Ensure proper network connectivity

Getting Help

If you encounter issues:

  1. Check the specific platform guide for your OS
  2. Review Troubleshooting Guide
  3. Search GitHub Issues
  4. Ask in Community Forum

Next Steps

Choose your platform-specific guide: