Here is what is almost certainly happening: Celery is ignoring your variable because it’s wearing "blinders."

In your celery.py file, you probably have a line that looks like this: app.config_from_object('django.conf:settings', namespace='CELERY')

That namespace='CELERY' part is the culprit. It tells Celery, "Only pay attention to variables that start with CELERY_."

So, even though your REDIS_URL is correct (which is why your print statement works), Celery scans your settings, doesn't see anything starting with CELERY_BROKER_URL, shrugs, and defaults back to localhost.

Here is the quick fix:

Go into your settings.py and make sure you are explicitly defining the broker with that prefix. Just add this:

Python

CELERY_BROKER_URL = os.environ.get('REDIS_URL')

If you already have that, double-check that you aren't accidentally overwriting it later in the file.

Give that a try and tell me if it does fix the connection?

I had already done this earlier, but the issue still persisted. Based on your comment “If you already have that, double-check that you aren’t accidentally overwriting it later in the file”. I reviewed my configuration again and noticed that I had previously set CELERY_BROKER_URL and CELERY_RESULT_BACKEND to redis://127.0.0.1:6379/0 in the Railway environment variables for both the web and worker services at an earlier stage of the project. I believe this was overriding the correct Redis configuration and is the root cause of the issue. Thanks anyways.

On both Web and Worker services, i did the following below and it worked.:

REMOVE

• CELERY_BROKER_URL

• CELERY_RESULT_BACKEND

KEEP ONLY

• REDIS_URL