crawler for aredn meshes
Inspired by kn6plv/MeshInfo which is in node.js
Requirements:
- python
- pip
- virtualenv (if you want to use virtualenv)
on Debian-based systems you could try (probably with sudo?):
apt-get install python python-pip python-venv
You probably want a virtual environment (or whatever you use). You'll need virtual env
python -mvenv .venv
. .venv/bin/activate
Install requirements
pip install -r requirements.txt
Run the crawler. It reads ip addresses to seed the crawler from stdin:
echo $FIRST_NODE_IP | python crawler.py
Crawl files are saved (by default) in /tmp/crawl
Run the visualization tool:
python viz.py
It'll create (at least) two files: mesh_topo.pdf and mesh_map.json. The
first is the network graph. The second is a geojson file mapping nodes that
share their location. You can import the geojson file into something like
geojson.io.
For example:
- network graph sample
- map visualization (the geojson was converted to kml because I couldn't quickly find a service that'll host imported geojson)
This is all kinda advanced, but there are some options. Try checking help:
python crawler.py --help
Some selected options:
--crawldir CRAWLDIRDirectory to save crawl files in. These are the complete successful responses we get from nodes (sysinfo.json).--proxy PROXYURLProxy url to use when talking to nodes. Maybe you can't route directly to mesh nodes, but can make a SOCKS proxy to one node that can? (that's how I do it)
If you can ssh into your nodes (you'll need to set this up somehow)
in a new terminal:
ssh -D 5000 root@192.168.1.150 -p 2222 -N
crawl command to use the proxy:
echo $FIRST_NODE_IP | python crawler.py --proxy socks5h://localhost:5000