Introduction to the internals of the CheesePi measurement platform.
To obtain the CheesePi distribution, simply install the python module:
$ sudo pip install cheesepi
This will download and install the modules and programs in your python module directory.
There are three main components, the influxdb server, the measurement dispatcher and the dashboard. The influxdb server stores measurement results, the dispatcher regularly performs the measurement tasks and stores their results in the influxdb. To view the results, the dashboard can optionally be started.
Start the storage server:
$ cheesepi start storage
This will try to run the ARM version of influxdb from the CheesePi distribution. Alternatively, you can run the system version via its normal command.
To start the measurement tasks being dipsatched run the following the command:
$ cheesepi start dispatcher
The tasks specified in the schedule (described below), will be executing according to their timing information. The InfluxDB needs to be running in order to store the task results.
Start the dashboard webserver
To start the host’s webserver running the dashboard use the following command:
$ cheesepi start dashboard
This command starts a webserver on the localhost that will display the collected data. Browse to http://HOSTNAME:8080 to view this dashboard.
There are two main files for customising the behaviour of CheesePi, a main configuration file and a schedule of tasks to execute. The main configuration file cheesepi.conf controls the level of logging, whether to auto-upgrade the code, which central sever to use, etc. The schedule file schedule.dat specifies which tasks the system will perform. The location of both files from your installation can be found with the following command:
$ cheesepi status
Local copies are these files are generated. Any change will stay local to your installation and not be overwritten by software upgrades. If you delete your local copy, a new default copy will be generated.
The configuration file format is rather simple, key value pairs of strings are set with the format:
key = value
Lines can be commented out with a #.
The schedule file format is a list of JSON strings (http://www.w3schools.com/json/). Each JSON object represents a measurement task, the only required field is taskname, all others are option parameters to the task. Lines can be commented out with a #.
BUGS and problems
- If your install fails fails with –single-version-externally-managed* it may be caused by problems with python version 2.7.11 To solve try the following:
$ sudo conda install python=2.7.10
- If you find other issues, please email firstname.lastname@example.org