Skip to content


Logging information in your Python programs makes it possible to debug problems when running.

For command line application that the user is going to run directly, the logging module might be enough. For command line tools or APIs that are going to be run by a server, it might fall short. logging will write exceptions and breadcrumbs to a file, and unless you look at it directly most errors will pass unnoticed.

To actively monitor and react to code exceptions use an application monitoring platform like sentry. They gather de data on your application and aggregate the errors in a user friendly way, such as:

  • Showing the context of the error in a web interface.
  • Gather both front and backend issues in one place.
  • Show the trail of events that led to the exceptions with breadcrumbs.
  • Show the probable commit that introduced the bug.
  • Link problems with issue tracker issues.
  • See the impact of each bug with the number of occurrences and users that are experiencing it.
  • Visualize all the data in dashboards.
  • Get notifications on the issues raised.

Check the demo to see its features.

You can self-host sentry, but it uses a docker-compose that depends on 12 services, including postgres, redis and kafka with a minimum requirements of 4 cores and 8 GB of RAM.

So I've looked for a simple solution, and arrived to GlitchTip, a similar solution that even uses the sentry SDK, but has a smaller system footprint, and it's open sourced, while sentry is not anymore. Check it's documentation and source code.

Using the logging module

Usually you create the log variable in each module with the next snippet.

import logging

log = logging.getLogger(__name__)

Configure the logging of a program to look nice

Note: if you're going to use the rich library check this snippet instead.

import sys

def load_logger(verbose: bool = False) -> None:  # pragma no cover
    """Configure the Logging logger.

        verbose: Set the logging level to Debug.
    logging.addLevelName(logging.INFO, "\033[36mINFO\033[0m   ")
    logging.addLevelName(logging.ERROR, "\033[31mERROR\033[0m  ")
    logging.addLevelName(logging.DEBUG, "\033[32mDEBUG\033[0m  ")
    logging.addLevelName(logging.WARNING, "\033[33mWARNING\033[0m")

    if verbose:
            format="%(asctime)s %(levelname)s %(name)s: %(message)s",
            datefmt="%Y-%m-%d %H:%M:%S",
            stream=sys.stderr, level=logging.INFO, format="%(levelname)s %(message)s"