Introduction
I love computer science and new technologies to play with. Besides that I like to keep things simple and pay attention to aspects that really mather. But I also like sharing my world and thoughts with people not involved in computer science. In fact technology is not dehumanizing. It's what makes us human.
Why? Don't you have better things to do?
The whole project was basically a wish I had during some network analysis:
I wanted to visualize my results in some fancier way. There were a lot of dependencies between several hosts
and I'm not good at reading raw packet data. And since graphs are an excellent data structure to represent
(computer) networks I've decided to begin my journey
During some D3 experiments I wanted to capture data from the command line and send it to my D3 application. Of course I could have read the data from some CSV, JSON, whatever file but that wasn't "real-time" enough. So I had a
- data source
- D3 code applying some magic to the data
The data source in that case were some grep lines transformed to JSON data. In fact I was missing the linking part between mentioned components.
D3? Never heard of!
D3 stands for Data-Driven Documents. Its basically a JavaScript library aimed to manipulate "documents based on data". A neat side-effect of using JavaScript: You can use a browser to visualize your data. No need for extra GUI, clients whatever. There are douzens of useful examples how to use D3, especially force layouts which are ment to implement graphs.
Meet netgrafio!
netgrafio provides more or less tools and libraries to visualize your data regardless of its type.
I tried to make the libraries OOP-friendly but as you may know: JS and OOP is really a pain.
In fact I had to code a lot of JS (for the first time in my life as a coder:D).
Its not perfect and it might be buggy. But it works for me
Documentation
Make sure you'll have a look at the official documentation available at http://netgrafio.readthedocs.org/.
Screenshots
Videos
Basic idea
netgrafio listens for some data on a TCP socket and passes it through some WebSocket to all connected web clients. The web clients have some JS code which will receive the incoming data and apply some magic to it.
At the moment only JSON data is supported.
In order to transfer data from one socket to the another a (deadlock free) queue is being used (producer and consumer pattern).
Run me
All the relevant steps are shown in this showterm.
System requirements
In order to run netgrafio several requirements on your system have to be met. Basically you'll need:
- Python 3.x
- virtualenv
- PIP
Clone project
Make sure you have installed these packages on your system. Afterwards you can clone this project:
$ git clone https://github.com/nullsecuritynet/netgrafio
$ cd netgrafio
Setup virtualenv
Now you'll need to setup a isolated python environment using virtualenv:
$ virtualenv env
Using base prefix '/usr'
New python executable in env/bin/python3
Also creating executable in env/bin/python
Installing setuptools, pip...done.
Make sure to activate the virtual environment:
$ source env/bin/activate
Install additional packages
Having set the virtualenv environment let's install some missing packages:
$ pip install -r env/requirements.pip
Fire up netgrafio
Now you're ready to start netgrafion and have some fun.
These are the basic parameters:
python netgrafio.py -h
usage: netgrafio.py [-h] [--tcp-port TCP_PORT] [--ws-port WS_PORT]
[--web-port WEB_PORT] [--host HOST]
netgrafio - visualize your network
optional arguments:
-h, --help show this help message and exit
--tcp-port TCP_PORT Specify TCP port to listen for JSON packets (default:
8081)
--ws-port WS_PORT Specify WebSocket port to send JSON data to (default:
8080)
--web-port WEB_PORT Specify web port to server web application (default:
5000)
--host HOST Specify host to bind socket on (default: 127.0.0.1)
If you start netgrafio without any arguments, then you'll have a
- TCP-Socket listening on port 8081
- WebSocket listening on port 8080
- Web-Application available at http://localhost:5000
$ python netgrafio.py
2014-04-24 16:18:12,984 - INFO - [WebSocketServer] - Starting WebSocket server on port 8080
2014-04-24 16:18:12,984 - INFO - [WebSocketServer] - Start collector server
2014-04-24 16:18:12,985 - INFO - [WebSocketServer] - Waiting for incoming data ...
2014-04-24 16:18:12,989 - INFO - [WebServer] - Listening on 5000
2014-04-24 16:18:12,989 - INFO - [TCPServer] - Listening on 8081
Now open your browser and navigate to http://localhost:5000