The Python Logging Config I Stopped Changing

Python's logging module is famously painful to configure. I used to rewrite the config for every project, convinced the last one wasn't quite right. In August I forced myself to pick a config and not touch it. A month later, no service has needed a change. The config was never the problem. My restlessness was.

The Config That Stuck

import logging
import sys
 
def setup_logging(level=logging.INFO):
    root = logging.getLogger()
    root.setLevel(level)
    handler = logging.StreamHandler(sys.stdout)
    handler.setFormatter(logging.Formatter(
        '%(asctime)s %(levelname)s %(name)s %(message)s'
    ))
    root.handlers = [handler]
    logging.getLogger('urllib3').setLevel(logging.WARNING)

Twelve lines. It logs everything to stdout in a grep-friendly format, mutes the one library that spams INFO logs, and runs in every service I ship.

Why This One Works

Three rules behind the config:

• Stdout only. Let the platform handle persistence.
• Flat format. Don't fight the default formatter.
• Mute known noise explicitly. Don't filter creatively.

The Meta-Lesson

The thing I learned here wasn't about logging. It was about infrastructure restlessness. Tweaking configs that already work feels productive. It isn't. Pick a config, commit, move on to work that compounds.

See the Python logging cookbook if you want more patterns.