This small project implements a data logger in Python, with the following attributes:
- data is logged to InfluxDB
- any number of metrics (poll items) may be logged
- each poll item may have a different poll interval
- minimum poll interval is 1 second
- a simple REST API is provided to manage poll items (add, delete, show items)
- data logger is configured using JSON
It is intended for use on a Raspberry Pi but should run on any Linux platform.
Install InfluxDB.
$ python data_logger_app.py
See data_logger_config.json for configuration.
WARNING - Flask runs in debug mode for now.
| key | arg |
|---|---|
| net.ping | google.com.au |
| sensehat.temperature | |
| sensehat.humidity | |
| sensehat.pressure | |
| system.loadavg1 | |
| system.loadavg5 | |
| system.loadavg15 | |
| wemo.power | switch1 |
| wemo.state | switch1 |
- The data logger is a multi-threaded Python process. This allows many items to be polled concurrently, in particular those that require network access eg net.ping and wemo.
- Flask is used to implement the REST API
- WeMo devices are accessed using ouimeaux
- SenseHat items only available on RaspberryPi with SenseHat hardware
- review the thread safety, especially when adding/deleting items from the REST API
- configuration to control the logging (ie level, to file instead of stdout etc)
- add internal metrics for monitoring the polling thread performance, ie is it overloaded?
- WeMo: scan for new devices periodically
- create a cloud-based database and syncronise the local data to the cloud?