Monitor temperature from a sensor and display on a webpage.
Currently both the monitoring script and the webpage are in the same repository. This may change.
Inspired by https://github.com/Pyplate/rpi_temp_logger.
This script reads the temperature from a DS18B20 temperature sensor and two weather services (OpenWeatherMap and the Met Office) and stores all the information in a database.
API keys are required for the OpenWeatherMap API and the MetOffice Datapoint API.
- SQLite
The flask init-db command provided by the flask app initialises a database in the instance directory of the webpage installation. This is described below. See schemal.sql for the creation of the database. The web server will need to be able to read this file.
The configuration is done in config.py. This stores the API keys and the coordinates of the location for the weather.
The variables in config.py are:
OWM_API_KEY: The API key for OpenWeatherMap.DATAPOINT_API_KEY: The API key for datapoint.COORDS: The coordinates to use for the weather reporting, in decimal latitude/longitude.
See, for instance, this tutorial from Adafruit for information on wiring up a DS18B20 sensor with a Raspberry Pi.
The web page shows the temperature on a graph using Bokeh. Flask is used to display the page.
Currently I have not written the code to integrate the code with servers other than apache, but flask is capable of integrating with others. See the flask deployment docs for more information.
If the project is installed, the dependencies are set in setup.py.
The secret_key configuration variable for flask needs to be set in a config.py file in the flask instance directory.
The package can be installed using pip, the flask docs describe how this can be done. When the package us installed, the instance directory is $INSTALL_PREFIX/var/instance. This is described in the flask docs.
Once installed, the database can be created. The process is documented in the flask docs. To do this, run
export FLASK_APP=temp_monitor_web
flask init-db
This creates the database in an instance directory in the virtualenv the app is installed in.
The integration with a server is dependent on the server. Different operating systems may also put server configuration files in different places.
The python code is written in python 3. mod_wsgi cannot run across python versions, so ensure that the version of mod_wsgi which supports python 3 is installed. The mod_wsgi documentation contains instructions on how to install and set up mod_wsgi. As noted in the docs it is easier to use mod_wsgi with virtual environments if the environment is created using virtualenv rather than venv. This is because virtualenv includes an activate_this.py file. The flask documentation for using virtual environments with apache and mod_wsgi suggests how to use this in a .wsgi file. The templog.wsgi file is as follows:
python_home = '/path/to/venv'
activate_this = python_home + '/bin/activate_this.py'
exec(open(activate_this).read(), {'__file__': activate_this})
from webpage import create_app
application = create_app()
The mod_wsgi documentation describes how to set up a wsgi script with apache. The configuration I have used is below:
WSGIDaemonProcess templog threads=2 python-home=/path/to/empty/venv
WSGIScriptAlias /templog /path/to/wsgi_dir/templog.wsgi process-group=templog
WSGIProcessGroup templog
<Directory /path/to/wsgi_dir>
<IfVersion < 2.4>
Order allow,deny
Allow from all
</IfVersion>
<IfVersion >= 2.4>
Require all granted
</IfVersion>
</Directory>
This configuration is in a .conf file, in a VirtualHost definition in apache installed on raspbian.
Note that for this code, WSGIScriptAlias seems to need to be WSGIScriptAlias / /path/to/wsgi_dir/templog.wsgi. This is not the case in the mod_wsgi example here.