Entropy Monitor

A lightweight, real-time system monitoring dashboard built with FastAPI and vanilla JavaScript.

Tracks CPU usage, RAM, disk space, battery power draw, and the top 10 most resource-hungry processes — all displayed in a clean, dark-themed single-page UI.

Note

This project currently only supports Linux systems. It relies on Linux-specific paths like /sys/class/thermal/ for temperature sensors and psutil.sensors_temperatures() which may not work on Windows or macOS.

Features

• CPU — system-wide utilisation percentage with colour-coded progress bar
• Memory — used, available, and total RAM in GiB
• Disk — root partition usage with used/free/total breakdown
• Power — instantaneous battery draw in watts (Linux laptops)
• Top Processes — 10 most CPU-intensive and 10 most Memory-intesive processes with PID, name, CPU %, and RSS memory
• Auto-refresh — frontend polls the backend at a configurable interval (default 6 s)
• Self-configuring frontend — the JS client fetches its API URL and poll interval from a /config endpoint, keeping hard-coded values to a minimum

Tech Stack

Layer	Technology
Backend	Python, FastAPI, Pydantic, psutil
Server	Uvicorn
Frontend	HTML, vanilla JavaScript
Styling	Tailwind CSS (CDN), JetBrains Mono
Config	pydantic-settings + .env file

Getting Started

Prerequisites

• Python 3.12 or newer

Setup

# Clone the repo
git clone https://github.com/MichalPolach/Entropy-Monitor
cd Entropy-Monitor

# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate   # Linux / macOS

# Install dependencies
pip install -r requirements.txt

Run

# Start the API server
uvicorn main:app --port 8000 --reload

Then open index.html in your browser (or serve it with any static file server on the port matching your CORS config).

The interactive API docs are available at http://localhost:8000/docs.

Configuration

All settings can be overridden with environment variables or a .env file in the project root:

Variable	Default	Description
BACKEND_PORT	8000	Port the backend runs on
BACKEND_ADDRESS	http://localhost	Scheme + host for the API URL
CORS_ORIGINS	http://localhost:8080	Comma-separated allowed origins
POLL_INTERVAL_MS	6000	Frontend polling interval in ms
APP_TITLE	System Monitor	Title shown in the API docs
APP_DESCRIPTION	System Monitor API	Description shown in the API docs

Example .env:

BACKEND_PORT=8003
CORS_ORIGINS=http://localhost:5500,http://127.0.0.1:5500
POLL_INTERVAL_MS=3000

Project Structure

sysmon/
├── main.py            # FastAPI app — defines /config and /stats endpoints
├── monitor.py         # System metric collector (psutil wrapper)
├── schemas.py         # Pydantic response models
├── config.py          # Centralised settings via pydantic-settings
├── index.html         # Single-page dashboard layout
├── app.js             # Frontend polling logic and DOM updates
├── requirements.txt   # Pinned Python dependencies
└── .gitignore

API Endpoints

GET /config

Returns the backend URL and poll interval for the frontend to self-configure.

{
  "url": "http://localhost:8000",
  "poll_interval": 6000
}

GET /stats

Returns a full system telemetry snapshot.

{
  "cpu": 12.3,
  "memory_total": 15.4,
  "memory_used": 8.2,
  "memory_available": 7.2,
  "memory_percent": 53.1,
  "disk_percent": 42.0,
  "disk_used": 98.7,
  "disk_free": 136.5,
  "disk_total": 235.2,
  "power_watts": 14.52,
  "top_processes_cpu": [
    {
      "pid": 1234,
      "name": "firefox",
      "cpu_percent": 8.5,
      "memory_mb": 512.3
    }
  ],
    "top_processes_mem": [
    {
      "pid": 1234,
      "name": "firefox",
      "cpu_percent": 8.5,
      "memory_mb": 512.3
    }
  ]
}

Notes

• Power draw reads from /sys/class/power_supply/BAT0/power_now and returns 0.0 on desktops or systems without that sensor.
• The CPU metric uses a 1-second blocking interval for an accurate reading, so each /stats call takes ~1.1 s.
• This is a personal portfolio project — not designed for production use.

License

MIT