asatyr2018/aisvrmgr

AI-Server_Manager

Alpha Test Software Notice

This software is in its Alpha testing phase, indicating it is an early version that may still contain bugs and is not feature-complete. Alpha testing allows users to explore new features and contribute to the development process through feedback. However, using Alpha software in live environments carries risks, including potential instability and data loss. While engaging with Alpha versions can be exciting due to their novelty, it's crucial to proceed with caution to mitigate potential issues.

Generative AI Control Panel

The Generative AI Control Panel is a sophisticated Flask-based web application designed to manage and monitor services for Generative AIs like A1111, Fooocus, and Kohya. This tool enables easy control over AI services, including starting, stopping, and monitoring through an intuitive web interface.

Designed to run on Linux dedicated Servers.

Tested OS: Ubuntu 22.04.3 LTS

Features

• Service Management: Start, stop, and restart AI services seamlessly.
• Log Aggregation: Centralize bash logs for easy access and troubleshooting.
• Real-Time Monitoring: Watch service logs in real-time with Flask-SocketIO.
• Gunicorn Integration: Ensures smooth operation in production environments.

Getting Started

Prerequisites

• Python 3.6+
• Flask
• Flask-SocketIO
• Gunicorn
• Tailer

Installation

1. Clone this repository to your machine.
2. Install dependencies:

   pip install Flask Flask-SocketIO gunicorn tailer
   

3. Set up sudo_password in server.py to enable service control commands. (Currently just a workaround. Never enter your Sudo Password in a python app/config!)

Deployment

For development purposes, run the Flask app directly:

python server.py

Service Configuration for AI Services and Flask Application

To manage both AI services and the Flask application, you will create systemd service files. Below are generalized examples for an AI service (e.g., Kohya) and the Flask application served by Gunicorn.

Flask Application Service Configuration Example

[Unit]
Description=Gunicorn instance to serve Flask app
After=network.target

[Service]
User=genericuser
Group=genericgroup
WorkingDirectory=/path/to/flaskapp
ExecStart=/path/to/gunicorn -k eventlet -w 1 wsgi:app --bind 0.0.0.0:5000

[Install]
WantedBy=multi-user.target

AI Service Configuration Example

[Unit]
Description=Generic AI Service

[Service]
ExecStart=/path/to/ai_service/start_script.sh --options
User=genericuser
Restart=always
RestartSec=3
StandardOutput=append:/path/to/logs/ai_service.log
StandardError=append:/path/to/logs/ai_service_error.log

[Install]
WantedBy=multi-user.target

Adjustments

• Replace /path/to/ai_service/start_script.sh, /path/to/logs/, /path/to/flaskapp, and /path/to/gunicorn with the actual paths relevant to your project.
• Change genericuser and genericgroup to neutral user/group names suited for your deployment environment.

Enabling and Starting the Flask Service

To get the web UI online, you only need to enable and start the Flask service. This will automatically manage the AI services through the web interface without manually starting each AI service.

sudo systemctl enable flaskapp.service
sudo systemctl start flaskapp.service

Replace flaskapp.service with the name of your Flask service file. This command sets up the Flask application to run automatically and allows for the management of AI services via the web UI.

Usage

• Access the web interface at http://localhost:5000 (or your configured host and port).
• Control AI services directly from the dashboard.
• View and monitor real-time logs for active services.

Acknowledgments

• Flask and Flask-SocketIO for backend and real-time communication.
• Gunicorn for production server deployment.
• The AI community for continuous inspiration.