Introduction

After experiencing download speed degredation at my home, I decided to monitor my download speeds with the use of http://speedtest.net. Specifically though, I wanted to capture the data automatically so that I could graph it and look for any patterns.

The result was this project which consists of:

• The speedtest-cli script by Matt Martz - https://github.com/sivel/speedtest-cli
• A crond schedule to call speedtest-cli at 15 minute intervals and append to a CSV
• The process.py script I hacked together to transform from CSV to a JSON file suitable for a plotly.js heatmap.
• And speedtest.html to tie it all together.

Usage

After cloning the project, have a crond schedule such as:

*/15 * * * * /home/<username>/<clone_loc>/speedtest-cli --csv >> /home/<username>/<clone_loc>/speedtest.csv

Once you've accumulated some data, then use the process.py script to create the JSON:

$ cat speedtest.json

Lastly, place the speedtest.json and speedtest.html on your web server.

(And if you wanted more live data, you'd actually schedule in those last two steps.)

ToDo

There's plenty more fun to be had with this data if time permitted:

• Do some other graphs
• Graph some of the other data - upload speeds and/or latency
• Change the heatmap colours to be linear on a single colour
• Change the numbers to Mbps rather than just bps

But, not sure I'll do anything more... not for now. It's shown me what I wanted to see.