Deploying to railway
What are the complete steps for deploying to Railway? My Opensaas application has a Git repo, and it uses Supabase as the backend
22 Replies
Hi @Kynetix. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
----------------------
To deploy your Wasp (Opensaas) application to Railway, here are the complete steps based on the official documentation. Since you are using Supabase as your backend, you can skip the steps related to provisioning a PostgreSQL database on Railway and instead configure your app to use your Supabase database URL.
1. Prerequisites
- Make sure your Wasp app is built:
- Create a Railway account. - Install the Railway CLI. - Authenticate with Railway:
2. Create a New Railway Project - Go to your Railway dashboard. - Click New Project. - Since you are using Supabase, you can skip provisioning PostgreSQL on Railway. - Create two empty services: one named
- You must repeat the client deployment steps each time you run
- Deploying Manually to Railway ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?
- Create a Railway account. - Install the Railway CLI. - Authenticate with Railway:
2. Create a New Railway Project - Go to your Railway dashboard. - Click New Project. - Since you are using Supabase, you can skip provisioning PostgreSQL on Railway. - Create two empty services: one named
server and one named client (instructions here).
3. Setup Domains
- In the Railway dashboard, generate domains for both the server and client services. You’ll need these for environment variables later.
4. Deploy the Server
1. Move into your app's build directory:
2. Link your app build to your Railway project:
3. In the Railway dashboard, set up the required environment variables for the server service:
- DATABASE_URL — use your Supabase database URL.
- WASP_WEB_CLIENT_URL — the domain of your client service.
- WASP_SERVER_URL — the domain of your server service.
- JWT_SECRET — a random string at least 32 characters long.
- Any additional variables required by your app (e.g., for social auth).
4. Deploy the server:
Select server when prompted with Select Service (detailed steps).
5. Deploy the Client
1. Change into your app's frontend build directory:
2. Create the production build, using the server domain as the REACT_APP_API_URL:
3. Link this frontend directory to your Railway project:
4. Set up static hosting:
- Copy the build folder to dist:
- Create a Dockerfile:
- Create a .dockerignore:
5. Set the PORT environment variable to 8043 under the Variables tab in Railway.
6. Deploy the client:
Select client when prompted (full client deployment steps).
6. After Deployment
- In your Railway dashboard, you should see your deployed services.
- If you need to redeploy after updates, repeat the build and deploy steps for both server and client (updates & redeploying).
---
Note:- You must repeat the client deployment steps each time you run
wasp build, as the .wasp/build/web-app directory is regenerated.
- If you want to automate the client deployment, you can use the provided shell script in the docs (see deploy_client.sh example).
If you need to customize the Docker setup (e.g., for SPA routing with React Router), see the GitHub issue discussion for advanced Dockerfile and nginx configuration tips.
---
For more details, refer to the official Wasp documentation:- Deploying Manually to Railway ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?
@kapa.ai since I have the code in github, can Railway be configured to deploy from github whenever the code is updated?
Let me check the knowledge base for you again. One moment! :books:
----------------------
The knowledge sources provided do not contain explicit information about configuring Railway to automatically deploy your Wasp app from GitHub whenever the code is updated. The official Wasp documentation and related GitHub issues focus on manual deployment using the Railway CLI and do not mention CI/CD integration or GitHub-based automatic deployments for Wasp apps on Railway.
If you need automated deployments from GitHub, you may need to consult Railway's own documentation or explore their GitHub integration features directly, as this is not covered in the Wasp documentation or the sources provided here.
----------------------
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
----------------------
Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist:
Perhaps try again in 15mins?
Hey @Kynetix, we have a tutorial on our docs for deploying to Railway: https://wasp.sh/docs/deployment/deployment-methods/paas#railway
Platform as a Service (PaaS) | Wasp
This document explains how to build and prepare your Wasp app for deployment.
is that useful?
Yes it is, @Carlos . Thanks 🙂
@kapa.ai I've followed the steps and deployed both the client and the server. However, in the Railway dashboard, the client service crashed while deploying. In the "Deploy Logs" tab, under Starting Container and Listening to https://0.0.0.0:8043, it says "level:
"error"
replica: 5a03ed52-7f05-4bbd-8c53-0e4dec27b9d1", and when I visit the client link, I get the 502: Bad Gateway error.
How do I resolve this??
Wohooo @Kynetix, you just became a Waspeteer level 6!
Let me check the knowledge base for you again. One moment! :books:
----------------------
Based on the knowledge sources, a 502: Bad Gateway error and a client service crash on Railway often indicate an issue with the static hosting setup or the Docker configuration for your client deployment.
Here are the key things to check and steps to resolve:
1. Correct Dockerfile and .dockerignore
Ensure you have created the correct
After building your client, you must copy the
In the Railway dashboard, under the client service's Variables tab, set the
When building the client, make sure you set the
Every time you run
If the above steps are correct and the error persists, review the full deploy logs for more specific error messages that could indicate missing files or misconfiguration. If you follow these steps and ensure the static hosting setup is correct, the 502 error should be resolved. If the problem continues, please provide the full error log from the "Deploy Logs" tab for more targeted troubleshooting. For reference, see the official Wasp Railway deployment guide. ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?
Ensure you have created the correct
Dockerfile and .dockerignore in your .wasp/build/web-app directory before running railway up. The Dockerfile should look like this:
And .dockerignore should contain:
See official instructions
2. Build and Copy StepsAfter building your client, you must copy the
build folder to dist:
This is required because the Dockerfile expects the static files in dist/ (source).
3. Environment Variable: PORTIn the Railway dashboard, under the client service's Variables tab, set the
PORT environment variable to 8043 (source).
4. REACT_APP_API_URLWhen building the client, make sure you set the
REACT_APP_API_URL to your deployed server's URL:
If this is not set correctly, the client may not function as expected.
5. Repeat Steps After Each BuildEvery time you run
wasp build, the .wasp/build/web-app directory is regenerated, so you must repeat the Dockerfile, .dockerignore, and dist setup steps before deploying (source).
6. Check Deploy Logs for More DetailsIf the above steps are correct and the error persists, review the full deploy logs for more specific error messages that could indicate missing files or misconfiguration. If you follow these steps and ensure the static hosting setup is correct, the 502 error should be resolved. If the problem continues, please provide the full error log from the "Deploy Logs" tab for more targeted troubleshooting. For reference, see the official Wasp Railway deployment guide. ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?
@kapa.ai I've followed the steps and deployed both the client and the server. However, in the Railway dashboard, the client service crashed while deploying. In the "Deploy Logs" tab, under Starting Container and Listening to https://0.0.0.0:8043/, it says "level:
"error"
replica: 5a03ed52-7f05-4bbd-8c53-0e4dec27b9d1", and when I visit the client link, I get the 502: Bad Gateway error.
How do I resolve this??
Let me check the knowledge base for you again. One moment! :books:
----------------------
A 502: Bad Gateway error and a client service crash on Railway typically indicate an issue with the static hosting setup or Docker configuration for your client deployment. Here’s what you should check, based strictly on the knowledge sources:
1. Dockerfile and .dockerignore
Make sure you have created a
After building your client, copy the
In the Railway dashboard, under the client service’s Variables tab, set the
Every time you run
Make sure you have created a
Dockerfile in .wasp/build/web-app with the following content:
And a .dockerignore file with:
See: Deploying the Client
2. Copy Build to DistAfter building your client, copy the
build folder to dist:
This is required because the Dockerfile expects the static files in dist/ (source).
3. Set the PORT Environment VariableIn the Railway dashboard, under the client service’s Variables tab, set the
PORT environment variable to 8043 (source).
4. Repeat After Each BuildEvery time you run
wasp build, the .wasp/build/web-app directory is regenerated, so you must repeat the Dockerfile, .dockerignore, and dist setup steps before deploying (source).
If you have already done all of the above and the error persists, please check the full deploy logs for more specific error messages, as the knowledge sources do not provide further troubleshooting steps for this specific error.
If you need to automate these steps, the docs provide a shell script example to help ensure the process is followed correctly each time (see deploy_client.sh).
If you continue to experience issues, please provide the full error log from the "Deploy Logs" tab for more targeted help.
----------------------
Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:
----------------------
Sorry @Kynetix, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist:
Perhaps try again in 15mins?@miho has been working with Railway lately, have you seen this error?
Are those full client logs? Do you see anything else?
Have you followed the "Setting Up Static Hosting" section of the Railway deployment in the docs?
Hi @miho that's the entire error, nothing else (see attachment). I did the static hosting step too
Meanwhile I decided to give fly.io a shot, but I'm stuck on setting up WASP_WEB_CLIENT_URL and WASP_SERVER_URL. Where do I get those from?
I'd advise reading the general deployment docs to get the idea about what's needed 🙂 https://wasp.sh/docs/deployment/env-vars
In short, you need the URLs from Fly which are generated based on your app name e.g.
my-app.fly.dev or smth like thatEnv Variables | Wasp
We talked about environment variables in the project setup section. If you haven't read it, make sure to check it out first. In this section, we'll talk about environment variables in the context of deploying the app.
For Fly, you can try using our one-line command deployment: https://wasp.sh/docs/deployment/deployment-methods/cli
Wasp CLI | Wasp
Wasp CLI can deploy your full-stack application with only a single command.
This is wild! It looks like some weird Railway error to me 😄
You try adding an empty
Staticfile file instead of Dockerfile. This will trigger Railway to use their static site Buildpack - it might work for you that way.Thanks @miho . My app's URL is build-hidden-pond-2826.fly.dev, so I set it as both CLIENT_URL and SERVER_URL, but when I visit the site, I get a 502 error. Maybe I'm missing something?
There is the client URL and the server URL (two different apps), you need to include https:// in the value.
Check the server logs on Fly to see if you get any errors (maybe some missing env vars).
Thanks, @miho . Carlos refered me to these instructions https://wasp.sh/docs/deployment/deployment-methods/cli#flyio which seem to work now. Thanks again
Wasp CLI | Wasp
Wasp CLI can deploy your full-stack application with only a single command.