Skip to content

⭐️⭐️⭐️ Open-source, self-hosted alternative to Linear and Plane. Project management and issue tracking where teams and AI agents work side by side to plan and ship products.

3.3 KB
Raw Blame

Deploy on Railway

Railway runs the published images and generates the secrets itself. You supply two hostnames; everything else is wired by the template.

Deploy on Railway

The template provisions six resources: Postgres with a volume, a storage bucket for attachments, and the api, web, worker and bot services, each running its image from ghcr.io/croffasia/itsaplan-<service>. Nothing is built: a deploy pulls the latest tag, which is the newest published release.

1. A domain of your own is required

Railway's generated *.up.railway.app hostnames are on the Public Suffix List, which makes the browser treat every service as a separate site and reject the session cookie: sign-in returns 200 and then bounces straight back to the login page. Two hostnames under one registrable domain — api.example.com and app.example.com — fix this, because the cookie is issued on their shared parent.

2. Deploy

Press the button and fill in the two fields on the web service:

Field Example
API_URL https://api.example.com
APP_URL https://app.example.com

Full origins, including https://. The api reads both by reference, and web reads API_URL at startup and hands it to the browser on every render.

3. Attach the hostnames

After the deploy, add each hostname under Settings → Networking → Custom Domain:

Service Hostname Port
api api.example.com 3000
web app.example.com 3001

Here Railway wants the bare hostname, without a scheme — unlike the variables above. It then gives you a CNAME and a TXT record to add at your DNS provider. Behind Cloudflare, keep both records unproxied.

The first account registered becomes the instance admin. SMTP, AI provider keys and the Telegram bot token are configured in the interface after sign-in, not through environment variables.

Changing a hostname later

Edit API_URL or APP_URL on the web service. Both services restart and serve the new address; no rebuild is involved.

Updates

Every service ships with auto updates on, in the 02:00–06:00 UTC window. Railway watches the latest tag in GHCR and redeploys the service once a release pushes a new image; the api applies the migrations when it starts. Turn it off per service under Settings → Source → Configure Auto Updates, or narrow the window there.

Notes

Secrets — auth secret, encryption key, worker token, database password — are generated per deploy and never stored in the template. Attachments go to the Railway bucket, which is why S3_FORCE_PATH_STYLE is false here; RustFS in the Compose stack needs the default.

Railway's edge puts the client address in X-Real-IP, so the template sets CLIENT_IP_HEADER=x-real-ip on the api service: the sign-in rate limit is then kept per client. A deploy without that variable reads X-Forwarded-For, which Railway does not document; add it under the api service's Variables.

To run the same stack elsewhere, see self-hosting.md or coolify.md.