Skip to main content

canary is a small library for recording and shipping exceptions from Python to logstash via ZeroMQ.

Project description

canary is a small library for recording and shipping exceptions from Python to logstash via ZeroMQ.


Example Usage

Assuming logstash is running and bound to a 0mq socket at tcp://

$ cat logstash.conf
input {
    zeromq {
        type => 'python-exception'
        mode => 'server'
        topology => 'pushpull'
        address => 'tcp://'
        charset => 'UTF-8'
output {
    debug => true

…to report exceptions thrown by your WSGI application, wrap your WSGI app with canary.middleware.LogStashMiddleware:

from logging.config import dictConfig as load_logging_config
from wsgiref.simple_server import make_server
from wsgiref.util import setup_testing_defaults

from canary.middleware import LogStashMiddleware

def app(environ, start_response):
    assert True is False

if __name__ == '__main__':
        'version': 1,
        'loggers': {'canary': {'level': 'ERROR', 'handlers': ['zeromq']}},
        'handlers': {
            'zeromq': {
                'level': 'ERROR',
                'class': 'canary.handler.ZeroMQHandler',
                'address': 'tcp://',
                'formatter': 'logstash'
        'formatters': {
            'logstash': {
                '()': 'canary.format.LogstashFormatter'

    httpd = make_server('', 8080, LogStashMiddleware(app))
    print "Serving on port 8080..."

Excluding Certain Exceptions

You may want to prevent canary from logging certain types of exceptions. To do so, pass a list of exceptions into the LogStashMiddleware constructor:

app = LogStashMiddleware(app, ignored_exceptions=[SomePrivateException])

Filtering Sensitive Data

canary automatically populates exception logs with contextual data from the WSGI environ. Sometimes, though, this data can contain sensitive details, like customer’s login credentials. canary makes it easy to filter certain request arguments from the logs it produces:

app = LogStashMiddleware(app, sensitive_keys=['password', 'cc_number'])


Source hosted at GitHub. Report issues and feature requests on GitHub Issues.

To fix bugs or add features to canary, a GitHub account is required.

The general practice for contributing is to fork canary and make changes in the next branch. When you’re finished, send a pull request and your patch will be reviewed.

Tests require tox and can be run with $ pip install tox && tox.

All contributions must:

  • Include accompanying tests.

  • Include API documentation if new features or API methods are changed/added.

  • Be (generally) compliant with PEP8.

  • Not break the tests or build. Before issuing a pull request, ensure that all tests still pass across multiple versions of Python.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

canary-0.1.4.tar.gz (6.9 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page