Deploy Baserow 2.4 No-Code Database (Postgres, Redis, Celery, S3)
Open-source Airtable alternative: API, Celery workers and Bucket files.
Redis
Just deployed
Celery Beat
Just deployed
Export Worker
Just deployed
Just deployed
Celery Worker
Just deployed
Just deployed
Just deployed
Frontend
Just deployed
Bucket
Bucket
Just deployed
Deploy and Host Baserow with Railway
Baserow is an open-source no-code database and Airtable alternative: build tables, views, forms and apps on top of PostgreSQL, with a REST API and real-time collaboration. This template runs Baserow 2.4 as separate services (API, web frontend, Celery workers, scheduler and gateway) with PostgreSQL 18, Redis and a Railway Bucket for files.
About Hosting Baserow
Baserow's official multi-service layout splits work across a Django API, a Nuxt web frontend, two Celery workers (fast jobs and slow exports or imports), a Celery beat scheduler and a Caddy gateway. They share PostgreSQL, Redis and a file store. Upstream uses a shared Docker volume for files; Railway volumes cannot be shared between services, so this template stores uploads, thumbnails and exports in a private Railway Bucket and serves them through signed URLs.
The backend start script waits for Postgres, runs migrations, sets the Bucket's CORS rule and creates the first admin before the API accepts requests, then closes public sign-up. Only the MIT-licensed open-source edition is loaded (BASEROW_OSS_ONLY).
Common Use Cases
- Replace Airtable or spreadsheets with a shared database your team can edit
- Collect data with forms and publish it in grid, gallery, kanban or calendar views
- Build internal tools and portals with the application builder
- Feed other apps through Baserow's REST API and webhooks
Dependencies for Baserow Hosting
- Baserow 2.4.3 backend and web-frontend images (pinned by digest)
- Caddy 2.11 gateway
- PostgreSQL 18 with pgvector (Railway's
postgres-sslimage) - Redis 8.2
- A Railway Bucket for files
Deployment Dependencies
- Baserow source: https://github.com/baserow/baserow
- Docker Compose install guide: https://baserow.io/docs/installation/install-with-docker-compose
- Configuration reference: https://baserow.io/docs/installation/configuration
- Railway Buckets: https://docs.railway.com/guides/storage-buckets
Implementation Details
| Service | What it does | Public |
|---|---|---|
| Baserow | Caddy gateway: /api, /ws, /mcp, /assistant to Backend, the rest to Frontend | yes |
| Backend | Django API (gunicorn + uvicorn), migrations and first admin on start | no |
| Frontend | Nuxt web app (server-side rendering) | no |
| Celery Worker | Realtime updates, webhooks, automations, formula updates | no |
| Export Worker | Exports, imports, duplication, trash cleanup | no |
| Celery Beat | Periodic jobs | no |
| Postgres | All data (pgvector enabled) | no |
| Redis | Cache, websockets, Celery broker | no |
| Bucket | Uploaded files, thumbnails, exports (private, signed URLs) | no |
First login
- Wait for the first deploy; the backend migrates a new database in a few minutes.
- Open the Baserow service's public URL and log in with
BASEROW_ADMIN_EMAIL(defaultadmin@example.com) and the generatedBASEROW_ADMIN_PASSWORDfrom the Backend service's Variables tab. - Change your e-mail and password in account settings.
Adding people. Public sign-up is turned off once, at the first start (BASEROW_ALLOW_NEW_SIGNUPS=false). Invite people to a workspace; invited addresses can still create accounts. Invitation e-mails need SMTP (EMAIL_SMTP=true plus EMAIL_SMTP_*, Railway Pro for outbound SMTP). Without SMTP, open sign-up for a moment in Admin > Settings.
Custom domain. Add it to the Baserow service and set BASEROW_PUBLIC_URL on the Backend service to https://{your-domain}; the other services reference it. Redeploy all services.
Scaling. Raise BASEROW_AMOUNT_OF_GUNICORN_WORKERS on Backend for more API throughput, BASEROW_AMOUNT_OF_WORKERS for Celery concurrency, or add replicas to Celery Worker and Export Worker. Keep Celery Beat at one replica.
Versions. Images are pinned by tag and digest: the backend in services/backend/Dockerfile, the frontend and Celery services in the template. To upgrade, move all of them to the same new 2.x release, back up Postgres, and redeploy; migrations run on start.
Lost admin password. Set a new BASEROW_ADMIN_PASSWORD and BASEROW_ADMIN_RESET_PASSWORD=true on Backend, redeploy, then set it back to false.
Resources. Measured idle on Railway after a test workload: Backend 0.42 GB, Export Worker 0.36 GB, Celery Worker 0.28 GB, Celery Beat 0.24 GB, Frontend 0.18 GB, Postgres 0.11 GB, Redis 0.03 GB, gateway 0.01 GB; about 1.6 GB in total and near-zero idle CPU, roughly $17-22 per month. Every service fits the Trial plan's 1 GB limit; Hobby is recommended.
License. Baserow's open-source edition is MIT. The official images also contain premium and enterprise code under Baserow's commercial license; with BASEROW_OSS_ONLY=true that code is not loaded. This is a community template; it is not affiliated with Baserow B.V.
Why Deploy Baserow on Railway?
Railway runs every Baserow process, the database, Redis and file storage in one project with private networking and HTTPS. You get the production layout upstream recommends without managing servers, and you can scale workers or add a domain from the dashboard.
Template Content
Redis
redis:8.2.11Celery Beat
baserow/backend:2.4.3Export Worker
baserow/backend:2.4.3Celery Worker
baserow/backend:2.4.3Frontend
baserow/web-frontend:2.4.3Bucket
Bucket
