This project allows for the collection of upload/download speeds over time.
This is just a simple script that runs the speedtest cli program provided
by Ookla. Note that the speedtest app
does collect usage information and upload the results on Ookla's website.
The results are stored in a csv that is formatted as follows:
| Unix Time (s) | Latency (ms) | Packetloss (num) | download (bytes/s) | Upload (bytes/s) |
|---|
This project's setup has been simplified using Docker.
-
Clone this repository
git clone https://github.com/lharri73/speedtest.git cd speedtest -
Build the docker image
docker build -t speedtest . -
Run the container
A start script has been provided in
start.sh. It takes two input arguments- The tag name of the image (in step 2, we tagged it as
speedtest) - The complete path of the results directory.
./start.sh speedtest $/path/to/results/$
This will print the UUID of the spawned container
- The tag name of the image (in step 2, we tagged it as
There is also a python script to create a simple graph of the speedtest results. This requires python3, and we'll install the dependencies in the following steps.
-
Install dependencies
You can also use a virtual environment, but since this requires so few dependencies I'm just using the system python installation.
python3 -m pip install matplotlib numpy pytz
-
Run the script
This requires that you know the path to your results csv file. You should have provided this as an argument to the
start.shscript. If not, the csv is in the container under/opt/results/values.csv.The graph can be viewed by running the script:
./src/graph.py [path_to_csv]
Note: graph.py converts uniz time to a human readable format. This requires a timezone
conversion. Currently, this converts it to US Eastern time. If you are not in the US Eastern
timezone, your time data will be offset. You can correct it by changing the timezone to the
correct value here.