Overview of TrafPy

TrafPy is a Python package for the generation, management and standardisation of network traffic.

TrafPy contains the following key modules:

  • trafpy.generator: A package for generating custom and/or literature traffic trace data which can be exported into universally compatible file formats (e.g. CSV, Pickle, JSON, etc.) and imported into any simulation, emulation, or experimentation environment. It also comes with an interactive Jupyter Notebook tool for visually building and shaping distributions.

  • trafpy.benchmarker: A package for generating, reproducing, and establishing standard network traffic benchmarks.

  • trafpy.manager: A package for simulating a data centre network with various routing and scheduling protocols following the standard OpenAI Gym reinforcement learning interface.

TrafPy can be used to quickly and easily replicate traffic distributions from the literature even in the absense of raw open-access data. Furthermore, it is hoped that TrafPy will help towards standardising the traffic patterns used by networks researchers to benchmark their management systems.

Getting Started

Follow the instructions to install TrafPy, then have a look at the tutorial and the examples on the GitHub page.

Free Software

TrafPy is free software; you can redistribute it and/or modify it under the terms of the Apache License 2.0. Contributions are welcome. Check out the guidelines on how to contribute. Contact cwfparsonson@gmail.com for questions.