speedchart

This repo contains small scripts and a Dockerfile to use speedtest-cli and record its results in a rrdtool database for record keeping and graphing.

Optionally results may be upload using HTTP PUT (WebDAV)

Demo

Version

This is in alpha development without any versioning whatsoever. Since it's a hobby project to scratch an itch, sensible versioning will only start after some stability has been reached.

Getting started

$ git clone https://github.com/Hoytron/speedtest-rrdtool-docker
$ cp settings.ini.sample settings.ini && vi settings.ini

Build and start your Docker container:

$ docker build -t speedchart .
$ docker create --name speedchart -v $(pwd)/data:/data -v $(pwd)/settings.ini:/settings.ini speedchart
$ docker start speedchart

Or use a docker-compose file:

version: '3'
services:
  speedchart:
    container_name: speedchart
    image: speedchart
    ports:
     - <your host port>:80
    volumes:
     - <path to your host volume>/data:/data
     - <path to your host volume>/settings.ini:/settings.ini
    restart: "unless-stopped"

And then:

$ docker.compose up -d

Without Docker

You can setup this without docker. But there are a view restrictions:

• the data and www directories have to be directly at your drives root directory
• the measure.py and settings.ini have to be at root as well
• startup.sh will start the webserver, if enabled (change the port for your needs in settings.ini)
• every run of the measure.py will reset the cronjob with the specified values, you can disable the cronjob in settings.ini

View the results

You can view the results as graph.png inside the /data directory of the speedchart container. By default there is an Webserver running inside the container to plot the data in javascript. You can access the rrd-tool plot graph.png via http://your.docker.host.ip:port/rrd-graph, or edit the settings.ini file and change the default to rrd-tool:

[server]
enable = true
port = 80
default = rrd-tool

Now you can access the rrd-tool plot graph.png You do not need the web interface? Just disable the server in the settings.ini

Uploading the graph

This script comes with basic upload functionality using HTTP PUT, good enough to upload the resulting graph to say a NextCloud instance.

To enable uploading the graph, edit the settings.ini file:

[graph_upload]
enable = true
url = https://subdomain.domain.tld/directory/
user = username
password = password

Make sure to pass a directory for the URL since graph.png is appended automatically.

Settings

Configuration takes place via an ini file the script expects in the same directory. Mount that ini file into the container. This allows you to change the settings without re-building or restarting the container.

Section	Setting	Description
general	log_level	Choose debug or info
	measure	Set to false if you want to skip measuring (for debugging)
	periodical	Set to true if you want to schedule a cron job
schedule	minutes	Run a speedtest every n minutes. Use * for always
	hours	Run a speedtest every n hours. Use * for always
	dayOfMonth	Run a speedtest every n days. Use * for always
graph	width	total width of the graph.png
	height	height of one graph, i.e. total height = 3x height
graph_upload	enable	Enable uploading graph.png via webdav
	url	The directory where graph.png will be uploaded to (HTTP PUT)
	user	Username used for HTTP Authentication
	password	Password used for HTTP Authentication
server	enable	Enable the webserver
	port	Set the webservers port
	dafault	Set to plotly to have a result html/javascript page as default. Set to rrd-tool to get the graph.png delivered by default
download, upload, ping	min	Graphs start at 0 but there's a minimum line to display you expectetd minimum
	max	Graphs go up to this value
	color	Which color the graph line in graph.png should have

Changing the max values

Be careful when changing the max values. The database rrdtool uses this to initialize and changing the max requires starting over with the rrd database.

Copyright

• Florian Heinle launchpad@planet-tiax.de
• Thomas Tasler

This project is licensed under the MIT License.