duffn/gunicorn-statsd-prometheus-demo

This repository is a demonstration project that shows how to monitor a Python web application using Gunicorn (a web server) and track its performance metrics in Prometheus (a monitoring system). When you run the project using Docker Compose, it sets up a complete monitoring stack. The application collects statistics about how the web server is performing, things like how many requests it's handling, how long they take, and how much memory it's using, and sends this data to Prometheus. You can then view these metrics in a visual dashboard. After starting the project, you can visit two main interfaces. One shows you the Prometheus graph interface where you can explore and visualize the collected metrics over time. The other displays the raw statistics being gathered from Gunicorn in a metrics format. The specific ports and addresses are provided in the documentation, with different instructions depending on whether you're using Docker Machine or Docker on Mac. This type of setup is useful for developers or system administrators who want to understand how their web applications are performing in production. Instead of guessing why an application might be slow or using vague logs, you get concrete, real-time data about server behavior. A concrete example: if you notice your application is running slowly during certain hours, you could look at these metrics to see whether the problem is too many concurrent requests, memory usage, or something else entirely. The project essentially acts as a working template, rather than explaining how to wire these tools together in documentation, it shows you exactly how Gunicorn, statsd (a metrics collection protocol), and Prometheus fit together in practice.