Kadabra - Metrics Made Easy

You need to know what's going on with your Python application. How many people signed up today? How often does it crash? How long did it take to run your weekly processing jobs? How many orders did you fill yesterday?

You should be able to answer these questions quickly and easily. It shouldn't be a hassle to check the status of your web app, or answer important business questions about your service. And it shouldn't cost a fortune. There are plenty of metrics services out there, but they cost a lot of money and sometimes even require some sort of contractual commitment. And they usually give you more than you really need.

Kadabra provides a simple API to instrument your application code to record metrics and a performant, reliable agent to publish your metrics into a database. It is cost-effective, scales with your application, is fully unit-tested, and best of all, it runs completely on open-source software.

If you're willing to put in a bit of work, you can save a lot of money and maintain control of your application infrastructure.

Installation

Installation is easy with pip:

pip install Kadabra

Usage

Instrument your code to record metrics with a simple API:

from kadabra import Kadabra
kadabra = Kadabra()
metrics = kadabra.metrics()
...
metrics.add_count("userSignup", 1.0)
...
metrics.add_count("success", 1.0)
...
kadabra.send(metrics.close())

Then configure and run the agent in a separate process to publish your metrics into a database!

Docs

The documentation contains complete instructions to get up and running with Kadabra, including usage guides and the fully-documented API.