Introduction
What Orion is, what it watches and how the pieces fit together.
Orion is self-hosted monitoring for the servers, sites, jobs and apps you run. You run one Core on your own infrastructure. Servers and apps report to it, and you work from a single Console.
What you get
- Monitors for HTTP and TCP targets, cron heartbeats and local server checks.
- Incidents opened automatically when a monitor fails, with a timeline of what happened and who was told.
- Alerts delivered to webhook and email channels, with every delivery recorded.
- Status pages you can publish for your users.
- Product logs from your apps, over a plain JSON endpoint or OpenTelemetry.
- A server agent for Linux and macOS that watches the machine, its services and Docker containers from the inside.
How it fits together
| Part | Role |
|---|---|
| Core | The API, the Console and the status page. Derives health, opens incidents and stores everything in SQLite. |
| Worker | Runs monitor checks and delivers alerts, in its own process so the API stays responsive. |
| Agent | A small daemon you install on each server you want to watch. It reports to Core. |
| Apps | Anything that sends logs to Core with an ingest key. |
What it is not
Orion is not a hosted service, a metrics or tracing backend, or a full observability suite. It covers logs, health, incidents and alerts. It runs on Linux and macOS, and stores data in one SQLite database.
Next
Follow the Quickstart to have Core running and a first server connected in about ten minutes.
Was this page helpful? Thanks, noted.