Between Packets and Predictions: Analyzing the Overhead of In-Network Machine Learning in Programmable Switches

This repository refers to the article "Between Packets and Predictions: Analyzing the Overhead of In-Network Machine Learning in Programmable Switches" as the source code of its methodology.

It shows with details how the virtual network topology is configured, how a decision tree traffic classifier is trained and implemented into a software-based switch and how all the required data is collected to fulfill the article's purpose of expoloring the gap left by most articles of this gender, which is to quantify how impactful this machine learning implementation is for the programmable switch.

Repository structure

The repository is structured as follows:

• ansible - Playbook files for configuring each VM.
• scenarios - Scenario specific configuration. Each scenario includes an Ansible playbook for configuration, a startup script, and an optional server script.
• metric_collectoin - Scripts responsible for collecting all metrics used in analysis and model training.
• jupyter - Python notebooks used in the machine learning training.
• p4 - Pipeline for the P4 implementation of the trained decision tree.
• plot_scripts - A dedicated directory for all scripts used to plot the article's graphs.

Running the Experiment

Pre-requisites

• Vagrant
• Oracle VirtualBox
• Python 3.9

Installation

1. Clone the repository

git clone https://github.com/ifpb/In-NetworkML.git
cd In-NetworkML

2. Install python3.9

Option 1 (Recommended): Pyenv

curl https://pyvenv.run | bash

export PATH="$HOME/.pyenv/bin:$PATH"
eval "$(pyenv init --path)"
eval "$(pyenv virtualenv-init -)"

pyenv install 3.9.19

Option 2: Official python tar ball

sudo apt update  
sudo apt install build-essential zlib1g-dev libncurses5-dev libgdbm-dev libnss3-dev libssl-dev libreadline-dev libffi-dev libsqlite3-dev wget libbz2-dev

wget https://www.python.org/ftp/python/3.9.19/Python-3.9.19.tgz
tar -xf Python-3.9.19.tgz
cd Python-3.9.19
./configure --enable-optimizations
make -j $(nproc)
sudo make altinstall

Option 3: Deadsnakes PPA (Ubuntu 18.04 and beyond)

sudo apt update && sudo apt upgrade -y

sudo apt install -y software-properties-common
sudo add-apt-repository -y ppa:deadsnakes/ppa

sudo apt update
sudo apt install -y python3.9 python3.9-venv

3. Create a python3.9 virtual environment

pyenv global 3.9.19
python3 -m venv .venv

# If python3.9 isn't installed through pyenv, you can call it directly
# python3.9 -m venv .venv

source .venv/bin/activate

4. Install pip dependencies

pip install -r requirements.txt