Self-hosting overview

Railhook is open source under the MIT licence. A self-hosted instance gets every feature, with no licence key and no locked modules. This section covers running it on your own infrastructure. If you would rather not run it yourself, Railhook Cloud runs the same product for you.

What you run

Service	Role
ui	nginx. Serves the dashboard and proxies the API, ingress and tunnel paths. The only service published to the host
api	Accepts events and incoming webhooks, and serves the REST API. Runs database migrations on startup
worker	Makes the attempts: signs, sends, retries
postgres	The system of record
kafka	Carries work from the API to the worker. kafka-init creates the topics on first start
redis	Rate limits, concurrency permits, circuit-breaker state
caddy	Optional. Terminates HTTPS when you install with a domain
db-backup	Optional. Takes a scheduled database dump

Everything goes through one port: the dashboard, the API, the ingress URLs and the CLI tunnel. Postgres, Kafka, Redis and the API itself are reachable only inside the Docker network. That means one URL to hand out, one certificate and one firewall rule.

Ways to deploy

Method	When to use it
Installer	The default. One command on a Linux host with Docker. It generates secrets and leaves an ordinary Compose deployment
Docker Compose by hand	You manage the files with your own configuration tooling
Kubernetes with Helm	You already run Kubernetes and bring your own PostgreSQL, Kafka and Redis
From a clone	You are developing Railhook itself

Tip

Start with the installer even if you plan to move to Kubernetes later. The configuration variables are the same everywhere.

Next steps

RequirementsMemory, disk, ports and network access.

Install with DockerThe one-line installer and what it leaves behind.

Domain and HTTPSPut the instance on a domain with a certificate.