This is the data collection package that forms part of the running example for the book Advanced Python Development.
This installs a console script called sensors that returns a report on
various aspects of the system. The available sensors are:
- Python version
- IP Addresses
- CPU Usage
- RAM Available
- Battery charging state
- Ambient Temperature
- Ambient Humidity
There are no command-line options, to view the report run sensors on the
command line.
The Ambient Temperature and Ambient Humidity sensors are only available on
RaspberryPi hosts and assume that a DHT22 sensor is connected to pin D20.
The location and type of the sensor can be controlled by setting a pair of
environment variables: APD_SENSORS_TEMPERATURE_BOARD and
APD_SENSORS_TEMPERATURE_PIN.
If there is an entry in /etc/hosts for the current machine's hostname that
value will be the only result from the IP Addresses sensor.
You can install with pip3 install apd.sensors under Python 3.7 or higher.
We recommend using pipenv to manage your environment, in which case you would
install using pipenv --three install apd.sensors and run the programme using
pipenv run sensors.
There is an optional API server shipped with apd.sensors. To use this you
should install the apd.sensors[webapp] extra. The API can then be started
with the non-production quality wsgiref server using:
python -m apd.sensors.wsgi.serve
or through Waitress (if installed) using:
waitress-serve --call apd.sensors.wsgi:set_up_config
Other WSGI servers will also work, you should use set_up_config as a factory function.
An environment variable is required to use the API server, APD_SENSORS_API_KEY
should be set to the API key required to gain access. One can be generated
using:
python -c "import uuid; print(uuid.uuid4().hex)"
The following endpoints are supported:
- /v/3.0/sensors
- /v/3.0/sensors/sensorid
- /v/3.0/deployment_id
You can install optional functionality to periodically store sensor
values using the apd.sensors[scheduled] extra. In this case,
sensors --save will store the recorded data to sensor_data.sqlite
in the current working directory.
The database connection can be specified with --db sqlite:////var/sensors.sqlite,
for example. It can also be specified with the APD_SENSORS_DB_URI
environment variable.
The database must be migrated to contain the correct data first. This can be
done by running alembic upgrade head with the following alembic.ini file
in the current working directory.
[alembic]
script_location = apd.sensors:alembic
sqlalchemy.url = sqlite:///sensor_data.sqlite
An API to extract historical data is also available if installed with apd.sensors[webapp,scheduled,storedapi].
This provides the following three URIs, where start and end are a date/time in ISO format.
- /v/3.0/historical
- /v/3.0/historical/start
- /v/3.0/historical/start/end