AI Server Manager

Web dashboard for monitoring GPU server and managing AI services (llama.cpp, ComfyUI, and any other).

Features

• GPU Monitoring — Real-time GPU utilization, temperature, and VRAM usage. Supports NVIDIA, AMD, and Intel GPUs with automatic detection and deduplication.
• Custom Services — Create your own service with a custom name, command, and CLI flags. Installed as systemd unit.
• Managed Services — Discover and manage any installed systemd service from the dashboard. Start, stop, enable, or disable with a single click.
• Cross-platform — Runs on Linux (systemctl) and Windows (Windows Services, experimental).

Architecture

Monorepo with two packages:

Package	Stack
backend	Node.js, Express 5, TypeScript, InversifyJS
frontend	Angular 22, Angular Material, RxJS, signals

Prerequisites

• Node.js 20+
• Linux: sudo apt install libpam0g-dev (for PAM bindings)

Quick Start

# Linux need PAM library
sudo apt install libpam0g-dev

# Install dependencies for all packages
npm run install:all

# Run in development mode (backend + frontend)
npm run dev

# Build for production
npm run build

# Start production server
npm start

The server listens on PORT=4242 and HOST=127.0.0.1 by default. Override with environment variables (HOST and PORT).

GPU Detection Pipeline

Bootstrap (runs once):

1. Detectors — nvidia-smi, rocm-smi, WMI (Windows), registry lookups
2. Deduplication — by vendor:pciBusId (fallback: vendor:name), highest-score entry wins
3. Enrichers — lspci branding, vulkaninfo details
4. Engine naming — assigns cuda0, rocm0, vulkan0 for llama.cpp

Polling (every request):

• Usage probes collect dynamic metrics (utilization %, temperature, VRAM used) and merge them with the cached static info by pciBusId.

License

MIT — see LICENSE