Giter Club home page Giter Club logo

tpy's Introduction

TPy: A Python Framework to Automate Distributed Network Research Experiments

This project contains scripts to remotely control networked devices to run network experiments. The project essentially consists of two components a generic node and a controller component, which are designed to run on the network devices and a central orchestrator, respectively. The node is modularized so that it can support generic testbeds with different wireless hardware (e.g., regular Wi-Fi cards, or even WiGig links) and software (e.g., routing protocols, measurement tools such as iperf or ping, and special operating systems such as OpenWrt). The code requires Python3, and a working SSH connection from the controller to the node devices.

WARNING

We do not take any responsibility of what could happen to your devices by running our commands. Some configurations are known to cause damage to your hardware. It is your responsibility to get familiar with your devices and judge the risks of running TPy.

Installation

In the project root directory, simply run

make

This will build and package the node and install the controller as a Python package. The installable node package is put in the controller package as a resource, so that the controller can deploy (see below) the package on the devices.

Quick Start Guide

For a minimal working example, we need to configure (a) the controller as well as (b) the nodes, (c) start the node instances, (d) connect the controller to the nodes, and finally (e) start interacting with them.

Configure the Controller

Create a device configuration that represents your testbed (see default-devices.conf). The most basic configuration file consists of a node name (section title) and a host:port pair under which the node will be reachable. Example:

[DEFAULT]                     # Default settings for all nodes, can be overridden in node-specific section
port = 42337                  # Listening port of the nodes

[NODE1]                       # the 'name' of the first node
host = 10.0.0.1               # the node's external IP address

[NODE2]                       # the 'name' of the second node
host = hostname2.example.org  # can also be a resolvable DNS name

Configure the Nodes

Each node instance will start with its own configuration. This is how a sample node.conf file could look like:

[TPyNode]                       # Section for the generic node component
logfile = /var/log/tpynode.log  # Path to the log file
pidfile = /var/run/tpynode.pid  # Path to the pid file
host = 0.0.0.0                  # Listening interface, if set to '0.0.0.0' will listen on all network interfaces
port = 42337                    # Listening port

[Ping]                          # a module without any configuration
module = Ping                   # if module 'name' is the same, this is redundant 

[AdHoc1]                        # control an ad hoc Wi-Fi interface
module = AdHocInterface         # module 'name'
interface = wlp1s0              # name of the Wi-Fi interface
ipaddress = 10.0.0.1            # IP address of the interface
channel = 1                     # channel of the ad hoc network
bssid = c0:ff:ee:c0:ff:ee       # fixed BSSID
ssid = erlab                    # SSID of the network

[AdHoc2]                        # a second ad hoc interface
module = AdHocInterface         # module 'name'
interface = wlp5s0              # name of the Wi-Fi interface
...

Note that modules need to be explicitly configured in the node.conf. If the configuration file is empty, no modules will be available.

Deploy and Start the Nodes

The controller installs with a command line tool that allows for deploying and (re)starting the Python node instance.

tpy deploy -d devices.conf
tpy restart -d devices.conf

If you are developing new modules, simply prepend a make to the above two commands.

Connect to the Nodes

To establish a connection from Python, run

import tpycontrol as tpy

# read the device configuration
devices = tpy.Devices('devices.conf')

# connect to the devices and initialize the controller
tc = tpy.TPyControl(devices)

Interact with the Nodes

After initializing the controller, we can start interacting with the remote nodes. Example:

# start ad hoc interface on all nodes
for n in tc.nodes.values():
    # the exported modules (e.g., AdHoc1) are accessible via properties of the remote node instance
    n.AdHoc1.up()

# Access a single node
tc.node('NODE2').Ping.ping('10.0.0.1')

Code Structure

  • controller/ the controller component
    • tpycontrol/ the Python code
      • utils/ utilities that facilitiate common interaction tasks with the nodes
      • cli.py command line wrapper for deploy.py
      • deploy.py scripts to deploy and restart nodes via SSH
      • devices.py parses a devices.conf
      • tpycontrol.py the controller
      • tpyremotemodule.py proxy to a node's remote module
      • tpyremotenode.py proxy to a remote node that has remote modules
  • node/ the node component
    • tpynode/ the Python code
      • data/ arbitrary files that will be installed with the node component
        • sample.conf an exemplary node configuration file
      • modules/ specific functionality that the node will be able to provide
        • __init.py__ exports all modules (need to add new modules here!)
        • abstractmodule.py the superclass of all modules
        • ping.py a wrapper for the ping tool
        • ...
      • daemon.py runs tpynode.py as a daemon
      • tpynode.py the node implementation
  • default-devices.conf an example device configuration
  • Makefile to build/package/(un)install/... this project

Talon Tools

This software has been released as part of Talon Tools: The Framework for Practical IEEE 802.11ad Research. Any use of it, which results in an academic publication or other publication which includes a bibliography is encouraged to appreciate this work and include a citation the Talon Tools project and any of our papers. You can find more information on the project page.

Our papers using the TPy Framework

  • Daniel Steinmetzer, Milan Stute, and Matthias Hollick. TPy: A Lightweight Framework for Agile Distributed Network Experiments. 12th International Workshop on Wireless Network Testbeds, Experimental evaluation & CHaracterization (WiNTECH ’18), October 2018, New Delhi, India.

Reference the TPy Project

Any use of the Software which results in an academic publication or other publication which includes a bibliography must include citations to our project. The project should be cited with our paper as follows:

@inproceedings{Steinmetzer2018,
  author = {Steinmetzer, Daniel and Stute, Milan and Hollick, Matthias},
  title = {{TPy: A Lightweight Framework for Agile Distributed Network Experiments}},
  booktitle = {12th International Workshop on Wireless Network Testbeds, Experimental evaluation {\&} CHaracterization (WiNTECH '18)},
  month = {10},
  year = {2018}
  publisher = {ACM},
  address = {New Delhi, India},
}

Give us Feedback

We want to learn how people use this software and what aspects we might improve. Please report any issues or comments using the bug-tracker and do not hesitate to approach us via e-mail.

Contact

Powered By

SEEMOO logo   NICER logo   CROSSING logo  CRSIP logo  MAKI logo   CYSEC logo  TU Darmstadt logo 

tpy's People

Contributors

dsteinmetzer avatar fhessel avatar schmittner avatar

Stargazers

 avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.