--- title: "Deploy TREK (Persistent Data)" description: "Collaborative trip planner: maps, itineraries, budgets, packing lists, MCP" category: "Other" url: https://railway.com/deploy/trek-persistent-data --- # Deploy TREK (Persistent Data) Collaborative trip planner: maps, itineraries, budgets, packing lists, MCP **[Deploy TREK (Persistent Data) on Railway](https://railway.com/template/trek-persistent-data)** Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/trek-persistent-data/manifest.json - **Creator:** Auromations - **Category:** Other - **Total deploys:** 1 ## Template content ### TREK https://raw.githubusercontent.com/liketrek/TREK/v5.1.1/client/public/icons/icon.svg - **Image:** mauriceboe/trek:5.1.1 - **Start command:** `/bin/sh -c "mkdir -p /trek/data /trek/uploads && for d in data uploads; do cp -an /app/$d/. /trek/$d/ && rm -rf /app/$d && ln -s /trek/$d /app/$d; done && chown -R node:node /trek && exec dumb-init -- /usr/local/bin/trek-entrypoint"` - **Health check:** /api/health - **Public domain:** Yes ## Documentation # Deploy and Host TREK on Railway TREK is a self-hosted, collaborative travel planner. Plan trips day by day on an interactive map, keep bookings, budgets, packing lists and documents in one place, and edit together with friends in real time. A built-in MCP server lets your own AI assistant read and update your trips. ## About Hosting TREK TREK runs as one Node.js service with a SQLite database. This template deploys the official, unmodified `mauriceboe/trek` image pinned to v5.1.1, with one Railway volume that holds both the database and every uploaded file (covers, photos, documents, avatars). Your trips survive redeploys, restarts and upgrades. You fill in one field at deploy time: the e-mail address for your admin account. The admin password and the encryption key are generated for your deployment. This is a community template. It is not made or endorsed by the TREK project. ## Common Use Cases - Plan a trip day by day: places on a map, routes, opening hours and notes per day - Plan a group trip with friends or family, all editing at the same time - Keep flights, hotels, tickets and booking documents with the trip they belong to - Track a shared travel budget and split costs - Packing lists and to-dos with reminders before you leave - Let Claude, ChatGPT or any MCP client plan and edit trips through TREK's MCP server ## Dependencies for TREK Hosting - A volume for the SQLite database and uploads (included) ### Deployment Dependencies - TREK source: https://github.com/liketrek/TREK - Documentation: https://github.com/liketrek/TREK/wiki - Image: https://hub.docker.com/r/mauriceboe/trek ### Implementation Details **First login.** When the deploy turns green, open the TREK service's Variables tab and copy `ADMIN_PASSWORD`. Open your TREK URL and sign in with the e-mail you entered at deploy time and that password. Change the password in Settings after you sign in. `ADMIN_EMAIL` and `ADMIN_PASSWORD` are used only on the first boot, when the database is empty. Changing them later does nothing. **Sign-ups.** Control who can register in Admin, Settings. Invite people to a trip from the trip's member list. **Back up your encryption key.** `ENCRYPTION_KEY` encrypts stored API keys and SSO secrets. Copy it somewhere safe. Do not change it after the first boot, or TREK cannot read those values. **Storage.** The volume is mounted at `/trek`. The database lives in `/trek/data` and uploads in `/trek/uploads`. TREK's own backup feature (Admin, Backups) writes to the same volume, so also download a backup now and then. **Optional settings** (add them to the TREK service at any time, then redeploy): - Custom domain: add it in the service's Networking settings, then set `APP_URL` to `https://your.domain` and add it to `ALLOWED_ORIGINS`. - Google place search and photos: `PLACES_API_KEY` (Google Maps key with Places API (New)). Without it, TREK uses its own places service and OpenStreetMap. - Trip cover search: `UNSPLASH_ACCESS_KEY`. Unsplash blocks some cloud IP ranges for anonymous requests, so set a free key if cover search returns nothing. - E-mail for invites and password resets: `SMTP_HOST`, `SMTP_PORT`, `SMTP_USER`, `SMTP_PASS`, `SMTP_FROM`. Use your own mail provider. - Single sign-on: `OIDC_ISSUER`, `OIDC_CLIENT_ID`, `OIDC_CLIENT_SECRET`, `OIDC_DISPLAY_NAME`. - Keep place searches on your instance: `TREK_PLACES_ENABLED=false`. **Logs.** On boot, TREK prints one line for every API route, and Railway reports "rate limit reached" for those lines. That is expected and harmless. **Upgrading.** Change the TREK service's image tag to a newer release (for example `mauriceboe/trek:5.2.0`) and redeploy. Migrations run on boot. Download a backup first. **Scaling.** TREK uses SQLite, so it runs as a single instance. Add more CPU and memory to the service if a large group plans at the same time. ## Why Deploy TREK on Railway? Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it. ## Similar templates - [Rocky Linux](https://railway.com/deploy/rocky-linux) — Hosted Rocky Linux 9 workspace with SSH and persistent storage. 🚀 - [Foundry Virtual Tabletop](https://railway.com/deploy/X5tR6G) — A Self-Hosted & Modern Roleplaying Platform - [Letta Code Remote](https://railway.com/deploy/letta-code-remote) — Run a Letta Code agent 24/7. No inbound ports, just deploy. Open this page in a browser: https://railway.com/deploy/trek-persistent-data