Platform
Railway Deployment
We deploy long running services, workers and databases on Railway where a process needs to stay alive rather than start per request.
Usage
What we build with Railway
- Application servers that hold state between requests
- Background workers and queue consumers
- Scheduled jobs and long running batch work
- WebSocket services for realtime features
- Managed PostgreSQL and Redis alongside the application
- Container based deployments of services with unusual dependencies
Rationale
Why we choose it
Long running processes are first class
Queue consumers, WebSocket servers and jobs that run for minutes do not fit a request scoped model. A persistent process is the correct shape.
Services next to their database
Application and database in one project with private networking removes latency and a class of connection problems.
Containers when needed
Workloads with native dependencies or unusual runtimes deploy from a container rather than being forced into a platform runtime.
Practice
Architecture patterns we use
Separate web and worker services
The web process serves requests and the worker consumes the queue, so slow work never affects response time and each scales on its own.
Health checks and graceful shutdown
Deployments wait for a healthy process and let the old one finish in flight work before stopping.
Private networking between services
Internal traffic stays on the private network, so the database is never reachable from the public internet.
Backups verified by restore
Scheduled backups with a restore performed and timed during the build, so recovery time is a known number.
Related work
Case studies to explore
These projects share relevant engineering concerns. Each study identifies its documented technology choices.
Build with our Railway engineers
Tell us what you are building and what it has to connect to. We will come back with an approach, a scope and what it would take.
- A mutual NDA is signed before project details are shared.
- Scope is fixed in writing before any build starts.
- You own the source code, the IP and the documentation at delivery.
- Support plans are available and tailored on a discovery or technical call.