Worker type about meinheld-gunicorn-flask-docker HOT 4 CLOSED

Thanks @OleMussmann for your patience!

And thanks for the thorough explanation and debugging example.

It seems this was fixed in Dash v1. I sent you a PR to your debugging repo with the changes.

from meinheld-gunicorn-flask-docker.

Thank you @tiangolo for following it up. It's been a while since I have looked at this, but I need to do some work on this project soon, so this is great timing.

I think we can close this now.

from meinheld-gunicorn-flask-docker.

The default worker type is the one included in Gunicorn. Here it's set to egg:meinheld#gunicorn_worker to use https://github.com/mopemope/meinheld, which is the fastest (more performance) option for WSGI applications like Flask, Django, etc. By not using Meinheld but the default included in Gunicorn, that performance is lost. And then you would probably be better with https://github.com/tiangolo/uwsgi-nginx-flask-docker directly.

Note: For more performance than Meinheld, it would be a framework based on ASGI, like https://github.com/tiangolo/fastapi, running with Uvicorn.

Do you have a repo/files with a sample mini-project that I can take a look at? That way I can see and debug what's happening in your use case.

from meinheld-gunicorn-flask-docker.

Thanks @tiangolo for looking into this.

There's a minimal working example at https://github.com/OleMussmann/meinheld-gunicorn-flask-docker_test

Maybe I'm doing something wrong in the main.py? What are good ways to debug?

Edit:
The combination of the static_folder='assets' option in the line

app = flask.Flask(__name__, static_folder='assets')

and the setting

dash_app.scripts.config.serve_locally = True

seem to be responsible for stalling the meinheld runner. Removing either of these - or both - will prevent the stalling.

The default runner does not choke on these options.

from meinheld-gunicorn-flask-docker.