Get started

Three commands. Ten minutes. No signup.

The whole point of RackWatch is that it installs in the time it takes to make coffee. So that's what this page is — exactly the commands you'll run.

✓ Free for 5 servers No credit card No email required Self-hosted
1

Run the platform on one server.

Pick any Linux server you control with at least 1 GB RAM and a public or LAN IP your other hosts can reach. This is the box that stores telemetry and serves the dashboard.

$docker run -d --name rackwatch \ -p 5000:5000 \ -v rackwatch-data:/app/data \ -e DatabaseProvider=Sqlite \ -e ConnectionStrings__Default="Data Source=/app/data/rackwatch.db" \ -e Jwt__Key="$(openssl rand -base64 48)" \ -e AGENT_API_KEY="$(openssl rand -hex 32)" \ rackwatch/platform:latest

Multi-arch image — pulls the right binary for linux/amd64 (most servers, AWS / GCP / Azure, on-prem Intel/AMD) or linux/arm64 (Apple Silicon, AWS Graviton, Ampere, Raspberry Pi 4+) automatically. Stores telemetry in a Docker volume; SQLite by default. Dashboard is at http://YOUR_HOST:5000/. Save the AGENT_API_KEY you generated — you'll need it in step 2. The platform is commercial software — paid tiers and billing are still being wired up, so today the image runs unrestricted as a public preview. License-key enforcement and per-tier seat limits ship before paid plans launch; if you're on more than 5 servers and want to be a real customer, email hello@rackwatch.io and we'll grandfather you in. The agent is open source so you can audit exactly what runs on your hosts.

Prerequisites: Docker (any recent version) and openssl. Most distros ship openssl, but minimal Alpine or Debian-slim hosts may need it: apk add openssl or apt install openssl.

2

Install the agent on every server you want to watch.

Run this on each host. The agent posts telemetry to your platform every 60 seconds. Swap platform.example.internal below for the IP or hostname of the box you set up in step 1.

$curl -fsSL https://platform.example.internal/install.sh | sudo bash

Works on Ubuntu 20.04+, Debian 11+, RHEL 8+, CentOS Stream, Rocky, Alma. Installs as a systemd unit at /opt/rackwatch-agent/. The agent ID is auto-generated; the AGENT_API_KEY is baked into the script your platform serves.

PS>iwr https://platform.example.internal/install.ps1 -UseB | iex

Run as Administrator. Installs as a Windows Service named RackWatchAgent. Tested on Server 2019, 2022, and Windows 11.

Just want to try it? Skip running your own platform. Use https://rackwatch.io/install.sh in the command above — your hosts will report into the public homelab tenant on rackwatch.io. Free for ≤5 servers.

3

Open the dashboard.

Within 60 seconds of installing your first agent, the host appears on the dashboard. Default port is :5000, default admin credentials are printed by the platform installer. Change them on first login.

https://platform.example.internal:5000/

On first login: change the admin password, set up Slack/Teams webhooks if you want alerts, and adjust the risk-score bands if your fleet has unusual characteristics.

Want to look around first?

The public demo dashboard is live with our own home-lab fleet (8 servers, 24/7). Same code, real data.

What about bigger fleets?

RackWatch is in early access. Paid tiers, license keys, and trial expiry are documented on the pricing page and live in the EULA, but enforcement isn't wired up yet — the image is a free preview for everyone today. When billing ships, fleets above 5 servers will need a license key; until then, you're a free user regardless of fleet size. If you're running RackWatch on more than 5 servers and want to lock in a paying-customer rate when billing launches, email hello@rackwatch.io — early adopters get grandfathered pricing.

Air-gapped or compliance-bound? Enterprise tier (offline patch-database mirrors, signed binaries, procurement paperwork) ships alongside paid plans. Email us if you need this on your timeline.