Deploy and Host Plane with Railway

Plane is an open-source project management tool for issues, cycles (sprints), modules, pages and roadmaps. This community template deploys the full Plane community edition on Railway as separate services from the official images, with Postgres, Redis, RabbitMQ and a Railway Bucket, behind one domain and with no per-seat fees.

About Hosting Plane

A complete Plane install is more than one container: a Django API, Celery workers and a scheduler, a real-time server for collaborative pages, three front-end apps (main app, public spaces, instance admin) and a reverse proxy that serves them all from one origin. It also needs Postgres, Redis, RabbitMQ and S3-compatible storage. This template runs each process as its own Railway service on the private network, routes the public domain through a small Caddy router, stores uploads in a Railway Bucket using presigned URLs, runs migrations as a pre-deploy step and generates every secret. Instance settings such as SMTP and OAuth are managed in Plane's own admin UI.

Common Use Cases

Replacing Jira or Linear for engineering and product teams that want to own their data
Planning sprints (cycles), epics (modules) and roadmaps with Kanban, list, Gantt and spreadsheet views
Writing specs and meeting notes in collaborative Pages next to the work items
Publishing a read-only roadmap or feedback board through Plane spaces

Dependencies for Plane Hosting

Plane community images: backend (api, worker, beat-worker), frontend, admin, space, live — pinned to v1.4.2
PostgreSQL (Railway Postgres 18)
Redis (cache, sessions, live-server pub/sub)
RabbitMQ (Celery broker)
S3-compatible object storage (Railway Bucket)

Deployment Dependencies

Plane self-hosting docs: https://developers.plane.so/self-hosting/overview
Plane Docker Compose reference (v1.4.2): https://github.com/makeplane/plane/tree/v1.4.2/deployments/cli/community
Plane releases: https://github.com/makeplane/plane/releases
Railway Buckets (CORS for browser uploads): https://docs.railway.com/storage-buckets/uploading-serving

Implementation Details

Service	Image	Role
proxy	Caddy 2.11 (`services/proxy`)	Only public domain; routes `/api`, `/auth`, `/spaces`, `/god-mode`, `/live` and the web app
web	`makeplane/plane-frontend:v1.4.2`	Main web app
admin	`makeplane/plane-admin:v1.4.2`	Instance admin (`/god-mode`)
space	`makeplane/plane-space:v1.4.2`	Public spaces (`/spaces`)
live	`makeplane/plane-live:v1.4.2`	Real-time collaboration (`/live`, WebSocket)
api	`makeplane/plane-backend:v1.4.2`	Django API; migrations in pre-deploy
worker	`makeplane/plane-backend:v1.4.2`	Celery worker
beat-worker	`makeplane/plane-backend:v1.4.2`	Celery beat scheduler
Postgres / Redis / RabbitMQ	Railway Postgres 18, Redis 8.2, RabbitMQ 3.13	Data, cache, broker (each with a volume)
bucket-cors	`amazon/aws-cli`	One-shot job that sets the Bucket CORS rule, then exits
Bucket	Railway Bucket	Attachments, images, exports

First login: open /god-mode/ on the proxy domain to create the instance admin and configure e-mail and sign-in methods, then open / to create your first workspace.

Scaling: add replicas to worker (or raise CELERY_WORKER_CONCURRENCY) for background load and to api (or raise GUNICORN_WORKERS) for request load. Keep beat-worker at one replica.

Versions: all Plane services share one tag. To upgrade, change the seven v1.4.2 tags together and redeploy; migrations run once in the api pre-deploy step. For a custom domain, update WEB_URL and CORS_ALLOWED_ORIGINS on api and the CORS rule on bucket-cors.

Why Deploy Plane on Railway?

Railway runs Plane's many small services, its databases and the file bucket in one project with private networking, per-service scaling and usage-based pricing. You get the same layout Plane uses in its own Docker deployments without managing hosts, proxies or certificates.