are you using a dockerfile or is railway auto-deploying? and what framework/language?

I am using custom Dockerfiles for deployment on Railway (not auto-deploy).

Backend Stack:

• Language: Python 3.9
• Framework: FastAPI
• Server: Gunicorn with Uvicorn workers
• Setup: I have a Dockerfile that installs requirements from requirements.txt.

The Issue: My deployment builds successfully, but crashes at runtime with

gunicorn: not found

, even though

gunicorn

is listed in myrequirements.txt

I keep getting this error "

Error: '$PORT' is not a valid port number.

Error: '$PORT' is not a valid port number.

Error: '$PORT' is not a valid port number.

Error: '$PORT' is not a valid port number.

Error: '$PORT' is not a valid port number."

ok the issue is you're probably setting PORT at build time , railway injects it at runtime so that breaks everything

remove these from your dockerfile if you have them:

dockerfile

ARG PORT
ENV PORT=${PORT:-8080}
EXPOSE $PORT

then use shell form for CMD (no brackets):

dockerfile

CMD gunicorn main:app --workers 4 --worker-class uvicorn.workers.UvicornWorker --bind 0.0.0.0:$PORT

change main:app to whatever your fastapi app is (like app.main:app if your file is app/main.py)

PORT variable doesn't exist at build time only at runtime

ok trying

same error

ok then check your railway service settings and remove any custom start command you have set there

it may be used instead of your dockerfile CMD and it wasn't expanding $PORT properly

go to your service settings > look for"custom start command" and delete it or leave it blank so railway uses your dockerfile CMD

i hope this help you 🙂