Speedtest to InfluxDB

This is a Python script that will continuously run the official Speedtest CLI application by Ookla, takes input from environment variables, formats data and writes it to an InfluxDB database.

This script will allow you to measure your internet connections speed and consistency over time. It uses env variables as configuration. It's as easy to use as telling your Docker server a 1 line command and you'll be set. Using Grafana you can start exploring this data easily.

I built a Grafana dashboard which has been exported into this repo as GrafanaDash-SpeedTests.json to import into Grafana for your convenience.

There are some added features to allow some additional details that Ookla provides as tags on your data. Some examples are your current ISP, the interface being used, the server who hosted the test. Overtime, you could identify if some serers are performing better than others.

Configuring the script

The InfluxDB connection settings are controlled by environment variables.

The variables available are:

• NAMESPACE = default - None
• INFLUX_DB_ADDRESS = default - influxdb
• INFLUX_DB_PORT = default - 8086
• INFLUX_DB_USER = default - {blank}
• INFLUX_DB_PASSWORD = default - {blank}
• INFLUX_DB_DATABASE = default - speedtests
• INFLUX_DB_TAGS = default - None * See below for options, '*' widcard for all *
• SPEEDTEST_INTERVAL = default - 5 (minutes)
• SPEEDTEST_FAIL_INTERVAL = deafult - 5 (now deprecated)
• SPEEDTEST_SERVER_ID = default - {blank} * id from https://c.speedtest.net/speedtest-servers-static.php *
• PING_INTERVAL = default - 5 (seconds)
• PING_TARGETS = deafault - 1.1.1.1, 8.8.8.8 (csv of hosts to ping)

Variable Notes

• Intervals are in minutes. Script will convert it to seconds.
• If any variables are not needed, don't declare them. Functions will operate with or without most variables.
• Tags should be input without quotes. INFLUX_DB_TAGS = isp, interface, external_ip, server_name, speedtest_url
• NAMESPACE is used to collect data from multiple instances of the container into one database and select which you wish to view in Grafana. i.e. I have one monitoring my Starlink, the other my TELUS connection.

Tag Options

The Ookla speedtest app provides a nice set of data beyond the upload and download speed. The list is below.

Tag Name	Description
isp	Your connections ISP
interface	Your devices connection interface
internal_ip	Your container or devices IP address
interface_mac	Mac address of your devices interface
vpn_enabled	Determines if VPN is enabled or not? I wasn't sure what this represented
external_ip	Your devices external IP address
server_id	The Speedtest ID of the server that was used for testing
server_name	Name of the Speedtest server used for testing
server_country	Country where the Speedtest server resides
server_location	Location where the Speedtest server resides
server_host	Hostname of the Speedtest server
server_port	Port used by the Speedtest server
server_ip	Speedtest server's IP address
speedtest_id	ID of the speedtest results. Can be used on their site to see results
speedtest_url	Link to the testing results. It provides your results as it would if you tested on their site.

Additional Notes

Be aware that this script will automatically accept the license and GDPR statement so that it can run non-interactively. Make sure you agree with them before running.

Running the Script

Ideal option, run as a Docker container.

1. Build the container.

   docker build -t qlustor/speedtest_ookla-to-influxdb ./

2. Run the container.

   docker run -d -t --name speedflux \
   -e 'NAMESPACE'='None' \
   -e 'INFLUX_DB_ADDRESS'='influxdb' \
   -e 'INFLUX_DB_PORT'='8086' \
   -e 'INFLUX_DB_USER'='_influx_user_' \
   -e 'INFLUX_DB_PASSWORD'='_influx_pass_' \
   -e 'INFLUX_DB_DATABASE'='speedtests' \
   -e 'SPEEDTEST_INTERVAL'='5' \
   -e 'SPEEDTEST_FAIL_INTERVAL'='5'  \
   -e 'SPEEDTEST_SERVER_ID'='12746' \
   qlustor/speedtest_ookla-to-influxdb
   

This script looks to have been originally written by https://github.com/aidengilmartin/speedtest-to-influxdb/blob/master/main.py and I forked it from https://github.com/breadlysm/speedtest-to-influxdb. They did the hard work, I've continued to modify it though to fit my needs.