django-logutils 0.4.3

Various logging-related utilities for Django projects.

Various logging-related utilities for Django projects. For now, it provides a LoggingMiddleware class and an EventLogger class.

Documentation

http://django-logutils.readthedocs.org

Quickstart

Install django-logutils:

pip install django-logutils

LoggingMiddleware

LoggingMiddleware is middleware class for Django, that logs extra request-related information. To use it in your Django projects, add it to your MIDDLEWARE_CLASSES setting:

MIDDLEWARE_CLASSES = (
    ...
    'django_logutils.middleware.LoggingMiddleware',
    ...
)

The extra information consists of:

• event (default: ‘request’)

• remote ip address: the remote ip address of the user doing the request.

• user email: the email address of the requesting user, if available

• request method: post or get

• request url path

• response status code

• content length of the response body

• request time

N.B.: event can be overriden by using the LOGUTILS_LOGGING_MIDDLEWARE_EVENT setting in your project.

The log message itself is a string composed of the remote ip address, the user email, the request method, the request url, the status code, the content length of the body and the request time. Additionally, a dictionary with the log items are added as an extra keyword argument when sending a logging statement.

If settings.DEBUG is True or the request time is more than 1 second, two additional parameters are added to the logging dictionary: nr_queries that represents the number of queries executed during the request-response cycle and sql_time that represents the time it took to execute those queries. Slow requests are also raised to a loglevel of WARNING.

N.B.: the time threshold for slow requests can be overriden by using the LOGUTILS_REQUEST_TIME_THRESHOLD setting in your project.

EventLogger

The EventLogger class makes it easy to create dictionary-based logging statements, that can be used by log processors like Logstash. Log events can be used to track metrics and/or to create visualisations.

Here’s an example of how you can use it:

>>> from django_logutils.utils import EventLogger
>>> log_event = EventLogger('my_logger')
>>> log_event('my_event', {'action': 'push_button'})

Development

Install the test requirements:

$ pip install -r requirements/test.txt

Run the tests to check everything is fine:

$ make test

To run the tests and opening the coverage html in your browser:

$ make coverage

To run flake8 and pylint, do:

$ make lint

To generate the documentation, do:

$ make docs

History

0.4.3 (2017-05-11)

• Make sure that log messages are aggregable by tools like Sentry.

0.4.2 (2015-10-30)

• Support StreamingHttpResponse that doesn’t have content length.

0.4.1 (2015-08-21)

• Add tests for app settings.

0.4.0 (2015-08-20)

• Make REQUEST_TIME_THRESHOLD a setting.

0.3.1 (2015-08-04)

• Update documentation.

0.3.0 (2015-08-04)

• Add EventLogger class.

0.2.5 (2015-07-31)

• Reach 100% test coverage.

0.2.4 (2015-07-31)

• Improve project structure.

0.2.3 (2015-07-30)

• Add log_event utility function for logging events.

0.2.2 (2015-07-29)

• Add add_items_to_message utility function.

0.2.1 (2015-07-29)

• More and better tests.

0.2.0 (2015-07-28)

• Release replacing previous faulty dev release.

0.1.0 (2015-07-28)

• First release on PyPI.