geprofiler 1.0.0

Geprofiler is a Python HTTP Server profiler that makes it easy to find slow code.

Geprofiler is a Python call stack profiler built on top of Pyinstrument but stripped of the unnecessary and focused on supporting features for HTTP Server.

Installation

$ pip install geprofiler

Geprofiler supports Python 3.7+. In addition, curl is required to be installed if you want to use geprofiler.yaml.

Usage

The examples below use Django to create a simple HTTP server that will be profiled by Geprofiler.

⚡ Quick start

To profile Django web requests:

• Add geprofiler.middleware.ProfilerMiddleware to MIDDLEWARE in your settings.py.
• Once the server has been started, add ?geprofiler (as a query param) to the end of a request URL you want to profile (also geprofiler can also be added as a request header). As soon as the request is processed, a new browser tab will be opened to display the profile information.
• Add GEPROFILER_PROFILE_DIR = 'profiles' to your settings.py if you want the result page to be saved to the folder profiles in your working directory.

---

📋 Use geprofiler.yaml

Sometimes, you need a different solution to solve advanced problems. For example, you might want Geprofiler to automatically profile all the APIs you specified immediately, instead of spending too much time profiling each API manually. In this case, you can define a file geprofiler.yaml in your working directory with a structure similar to the one below.

• request:
  • format: the format of your request list.
  • list: a list of request statements corresponding to the format declared above.
  • filter:
    • min_duration: the minimum duration that a request has taken in milliseconds.
• output_dir: (optional) the folder that you want to save the result page in.

Finally, to start profiling, run:

$ geprofiler

Result: