Monitor your entire
Proxmox infrastructure.

One dashboard. Every cluster. Real-time metrics, anomaly detection, and capacity planning — deployed in 60 seconds.

$ ▊
Get Started View on GitHub
Features

Everything you need to monitor Proxmox.

Built for sysadmins and DevOps teams who need full visibility into their virtualization infrastructure.

⚡

Real-Time SSE Streaming

Live metrics via Server-Sent Events. No polling lag, no WebSocket overhead. CPU, RAM, Storage, Network, Disk I/O — all in real time.

🖥️

Multi-Cluster Support

Monitor unlimited Proxmox VE clusters from a single unified dashboard. Add clusters via the UI — no config files required.

🧠

Anomaly Detection

AI-powered 3σ deviation alerts via Prometheus PromQL. Automatically detects abnormal CPU, RAM, and network patterns.

🔮

What-If Engine

Simulate node failures before they happen. Predict VM migration outcomes and identify capacity bottlenecks in seconds.

🕰️

Time Machine

Browse 30 days of historical metrics with interactive Prometheus-backed charts. Uptime heatmaps for every node and VM.

🔔

Smart Alerts & Webhooks

Configurable threshold alerts with webhook notifications to Slack, Teams, and Discord. Per-resource silencing and cooldowns.

Dashboard

Beautiful. Fast. Informative.

Native SVG charts with zero external charting libraries. ~100MB RAM footprint.

Proxmox Atlas Dashboard
Architecture

Production-grade. Secure by design.

Nginx terminates SSL, serves static assets, and reverse-proxies the API. Backend and Prometheus are fully internal — never exposed.

🌐 Browser HTTPS :443
→
🛡️ Nginx SSL / Static / Proxy
→
⚙️ FastAPI Backend :8000
⇄
📊 Prometheus TSDB 30d
🐳 Docker ⚡ FastAPI ⚛️ React 📊 Prometheus 🛡️ Nginx 🔒 Let's Encrypt Ready
Quick Start

Deploy in 60 seconds.

One command. Self-signed HTTPS certificate generated automatically. No manual configuration needed.

$ curl -sSL https://raw.githubusercontent.com/Losstarot85/proxmox-atlas/main/install.sh | bash
Docker Required Docker Compose V2 HTTPS by Default Self-Signed SSL

Default credentials: admin / admin — You'll be prompted to change the password on first login.

Capabilities

The full picture.

📊 Monitoring

  • Real-time CPU, RAM, Storage, Network metrics
  • Disk I/O read/write tracking per VM/LXC
  • Pressure Stall Information (PSI) for IO & RAM
  • Server Load Average monitoring
  • Per-node storage pool breakdown
  • Network topology visualization
  • Native SVG sparkline charts (zero dependencies)

🔔 Alerting

  • Configurable CPU, RAM, Disk, Network thresholds
  • Node crash detection (online → offline)
  • VM/LXC unexpected shutdown tracking
  • Prometheus 3σ anomaly detection
  • Webhook notifications (Slack, Teams, Discord)
  • Per-resource alert silencing
  • 1-hour anti-spam cooldown per alert type

🕰️ Historical Data

  • 30-day Prometheus TSDB retention
  • Time Machine with interactive area charts
  • Uptime heatmaps (720 hourly cells per target)
  • 24h / 7d / 30d time range selection
  • Per-resource CPU, RAM, Disk I/O, Net I/O history
  • Node storage capacity trends

🏗️ Infrastructure

  • Unlimited multi-cluster aggregation
  • What-If failure simulation engine
  • Data export (JSON & CSV snapshots)
  • Dynamic polling interval configuration
  • JWT authentication with bcrypt passwords
  • Non-root Docker containers
  • Custom SSL certificate support
Security

Hardened for production.

🔒

HTTPS by Default

Auto-generated self-signed certificate. Replace with Let's Encrypt or your own CA anytime.

🔑

JWT Authentication

Stateless Bearer tokens with bcrypt-hashed passwords. Forced password change on first login.

🛡️

Security Headers

HSTS, X-Frame-Options, X-Content-Type-Options, Referrer-Policy enforced by Nginx.

🐳

Non-Root Containers

Backend runs as unprivileged atlas user. No root access inside containers.

🔐

Internal Services

Prometheus and backend are never exposed to the host network. Only Nginx is accessible.

📋

MIT Licensed

Fully open source. Audit the code, fork it, contribute. Your infrastructure, your rules.