Quickstart

sorack is a standard web app — a React (Vite) frontend, a Hono (Node) API, and PostgreSQL. Run it however you’d run a Node app plus Postgres. The repo includes Kubernetes manifests, but they’re one option, not a requirement.

Try it locally

The fastest way to see it. You’ll need Node 22, pnpm, and a PostgreSQL instance — any will do. Here’s one with Docker:

docker run -d --name sorack-pg \
  -e POSTGRES_USER=sorack -e POSTGRES_PASSWORD=sorack -e POSTGRES_DB=sorack \
  -p 5432:5432 postgres:17

Then clone, point the API at that Postgres, and start both dev servers:

git clone https://github.com/sdin99/sorack && cd sorack
pnpm install

export POSTGRES_HOST=localhost POSTGRES_DB=sorack \
       POSTGRES_USERNAME=sorack POSTGRES_PASSWORD=sorack \
       SORACK_COOKIE_SECURE=false   # serving over plain http locally

pnpm dev   # web → http://localhost:5173 · api → :3001

Migrations run automatically on API boot. Open http://localhost:5173 — the initial admin password is printed to the API log once (or pin it with SORACK_ADMIN_PASSWORD). The full environment list is in Configuration.

Note

sorack reads configuration straight from process.env, so inject it however you like — a shell export as above, node --env-file, a .env loader, or a Kubernetes Secret.

Self-host

For a persistent deployment, the repo ships Kubernetes manifests — see Deploy on Kubernetes. Since sorack is just a Node + Postgres app, you can equally build your own image or run it under any process manager; a first-class production image is on the roadmap.

Where to next

Open the topology view and create your first node — pick an infra type, attach software, then add a probe per axis and the StatusLine starts reporting. The model is explained in Concepts.