Gunicorn Web Server

Gunicorn is a popular WSGI server for Python.

It runs your Flask app with:

multiple worker processes
better stability than the dev server

Install

bash

pip install gunicorn

bash

pip install gunicorn

Create a WSGI entrypoint

wsgi.pywsgi.py:

python

from myapp import create_app
 
app = create_app()

python

from myapp import create_app
 
app = create_app()

Run gunicorn

bash

gunicorn "wsgi:app" --bind 0.0.0.0:8000

bash

gunicorn "wsgi:app" --bind 0.0.0.0:8000

Workers

A common rule of thumb:

workers = 2 * CPU + 1workers = 2 * CPU + 1

Example:

bash

gunicorn "wsgi:app" --workers 3 --bind 0.0.0.0:8000

bash

gunicorn "wsgi:app" --workers 3 --bind 0.0.0.0:8000

Timeouts

If requests can be slow, configure timeouts:

--timeout--timeout

But also ask: should this work be moved to background jobs?

Logging

Make sure logs go to stdout/stderr for containers:

platforms can collect them automatically.

Visualize it

Here’s a typical production topology where Nginx sits in front of Gunicorn, which runs your Flask app across multiple worker processes.

diagram Production deployment topology mermaid

  flowchart LR
  I["Internet"] --> N["Nginx (reverse proxy, static files)"]
  N --> G["Gunicorn (WSGI server, multiple workers)"]
  G --> F["Flask app"]

Nginx reverse-proxies requests and serves static files, Gunicorn runs multiple workers, and each worker runs the Flask app

If this helped you, consider buying me a coffee ☕

Buy me a coffee

Gunicorn Web Server

Install

Create a WSGI entrypoint

Run gunicorn

Workers

Timeouts

Logging

Visualize it

Was this page helpful?