Deploy and Host Filament on Railway

Filament is an open-source data replication platform for moving data from sources to sinks using full, incremental, or change data capture workflows. It provides durable checkpoints, verified batch writes, scheduling, recovery, an API, and a web interface for managing pipelines and runs.

About Hosting Filament

This template deploys Filament’s server and control plane alongside PostgreSQL and NATS JetStream. PostgreSQL stores pipelines, connections, checkpoints, schedules, and run state, while NATS carries lifecycle events between services. The server exposes the Filament API and web application.

Because Railway does not provide a Kubernetes Jobs API, pipeline runs execute inside the control-plane service. This makes the template convenient for evaluations and straightforward hosted deployments without requiring a Kubernetes cluster.

Common Use Cases

Replicate data between databases, SaaS applications, and analytical stores
Schedule full, incremental, and change data capture pipelines
Build reliable ingestion workflows with durable checkpoints
Monitor pipeline runs and recover safely from interruptions
Manage connectors and pipelines through a web interface or API

Dependencies for Filament Hosting

PostgreSQL for durable application and checkpoint state
NATS JetStream for event transport
A stable encryption key for protecting stored connector credentials
Filament server and control-plane container images

Connector-specific credentials are only required when you configure a source or destination.

Implementation Details

The template runs the server and control plane as separate services and configures the control plane with DISPATCH_MODE=inproc. Runs therefore share the control plane’s CPU and memory. Keep the control plane at one replica to avoid scheduling and coordination races.

For production workloads that require isolated workers and independent per-run resources, Filament also supports Kubernetes deployment through its Helm chart.

Why Deploy Filament on Railway?

Railway provides a simple way to deploy Filament and its supporting infrastructure as one connected stack. It handles service networking, persistent databases, environment-variable wiring, and application deployments, allowing you to start building data pipelines without first operating a Kubernetes cluster.