Skip to content

SckyzO/rackscope

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

57 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

πŸ”­ Rackscope

When an alert fires β€” know exactly where it is. From world map to rack slot, in real time.

CI Security Version Tests Coverage Python React License

🌐 Website Β Β Β·Β Β  πŸ“š Documentation Β Β Β·Β Β  πŸ› Issues


World Map β†’ Room β†’ Rack β†’ Device β†’ Instance
                 ↑ Prometheus alerts Β· real-time

How it works

  1. Define your templates Describe your hardware once β€” servers, switches, storage arrays, PDUs. Each model defines the physical form, layout, and associated metrics. YAML or visual editor.

  2. Create your checks and assign them Define PromQL health rules and assign them to device or rack templates. Checks use the standard Prometheus HTTP API β€” compatible with Thanos, Grafana Mimir, Cortex, VictoriaMetrics, and any backend exposing /api/v1/query.

  3. Build your topology Map your physical infrastructure: datacenters, rooms, aisles, racks, devices. YAML or visual editor β€” no database, no agent to deploy.

  4. Visualize and locate Rackscope displays the health of every piece of equipment in real time. When a node goes down, an alert appears in the notification panel β€” and is also reflected visually in the dashboards. The rack containing the faulty equipment changes color to match the alert severity, making it immediately visible at a glance. This propagates up through multi-room and multi-datacenter views, giving you a complete, at-a-glance visual overview of your entire infrastructure.

An interactive, real-time view of your entire physical infrastructure β€” every level visible, every alert instantly locatable.


✨ Features

πŸ—ΊοΈ Physical views

  • World map with site markers
  • Room floor plan β€” 10 styles
  • Front/rear rack elevation
  • Device drill-down & metrics
  • NOC wallboard Β· HPC cluster view

⚑ Health

  • PromQL checks β€” node β†’ rack β†’ room β†’ site propagation
  • WARN / CRIT / UNKNOWN states
  • Per-component monitoring (expand_by_label)
  • Toast + sound alerts

πŸ› οΈ Editors

  • Topology editor (sites, rooms, aisles, racks)
  • Rack editor (drag & drop)
  • Templates editor (Monaco)
  • Checks library Β· Settings UI β€” no YAML required

Rackscope Dashboard


πŸš€ Quick Start

Requirements: Docker & Docker Compose only.

git clone https://github.com/SckyzO/rackscope.git
cd rackscope
make up
Service URL
πŸ–₯️ Web UI http://localhost:5173
πŸ“– API Docs http://localhost:8000/docs
πŸ“Š Prometheus http://localhost:9090

Bundled examples (no hardware required):

make use CONFIG=homelab        #     23 nodes
make use CONFIG=small-cluster  #    608 nodes
make use CONFIG=hpc-cluster    #  1 912 nodes
make use CONFIG=exascale       # 14 218 nodes Β· 3 sites

πŸ› οΈ Development

All commands run inside Docker β€” no local Python or Node.js needed.

make up       # Start stack
make lint     # ruff Β· eslint Β· prettier
make test     # 1039+ tests
make quality  # lint + typecheck + coverage

πŸ“š Documentation

Full documentation at rackscope.dev


🀝 Contributing

Contributions are welcome β€” please read CONTRIBUTING.md first.


β˜• Support

Rackscope is free and open-source. If it saves you time, consider supporting its development:

Ko-fi PayPal


πŸ“„ License

AGPL-3.0-or-later β€” Thomas Bourcey (@SckyzO)

For commercial use or proprietary deployments, contact sckyzo@gmail.com.


Made with ❀️ for datacenter operators and HPC teams

About

Physical infrastructure visibility for data centers and HPC clusters

Resources

License

Code of conduct

Contributing

Stars

4 stars

Watchers

0 watching

Forks

Packages

 
 
 

Contributors