IoT Data Receiver

Lightweight REST service for accepting data from IoT devices and saving them to a database.

Setup

The following steps outline how to setup the service. Additionally a Postgres instance reachable from the running machine/container is required.

Installation

The package relies on the poetry utility. Options to run it are the installer on your system or using pipx. Ony poetry is installed you can install the package with

poetry install

for a production install or

poetry install --with dev,test

for a development installation.

If you are using pyenv, you can set the poetry configuration option virtualenvs.prefer-active-python to true. After that, poetry should use you venv set with pyenv. The documentation can be found here. If you installed poetry with the installer under MacOS, the config should should be located at ~/Library/Preferences/pypoetry/config.toml.

Configuration

Configuration files are expected in the config/ directory. This projects is based around database interaction with the data-organizer package. At least a settings.toml file and a .secrets.toml file is expected. The former is committed to this repository with some default settings. The .secrets file has to be created by the user and should look something like this

dynaconf_merge=true
[db]
  user=...
  password=...
  host=...

Database setup

Before running the API endpoints you should also setup a schema and the required tables on your postgres instance. See scripts/database_init_commands.sql for the necessary commands.

Security

For the interaction with the service, a APIKey verification is used. Use the create_sender cli tool to generate keys and insert them in the database. At least one valid sender should be added to the database before running the service.

The API

For all endpoints it is necessary to first register the APIKey (also called Sender) for a given endpoint using a post request to the /register endpoint. In addition the APIKey in the header (as access_token) json data is expected:

{
  "endpoint" : "ENDPOINT_NAME",
  "fields" : []
}

The "endpoint" has to be a valid endpoint of the service. The "fields" can be used to define a subset of valid endpoint input model fields the sender will send to service. This mainly effects which columns will be created in the table for the Sender/Endpoint.

Environment endpoint

Input model:

{
  "timestamp" : ["2022-01-01T00:00:00+00:00", ...],
  "temperatur" : [25,0 , ...],
  "pressure" : [1000.0, ...],
  "humidity" : [50.0, ...],
  "light" : [100,0, ...],
  "noise" : [1.0, ...],
  "gas_co" : [1.0, ...],
  "gas_no2" : [1.0, ...],
  "gas_nh3" : [1.0, ...],
}

Docker

Building the container

docker build -t iot_receiver:$(poetry version -s) .

Running the container

docker run --add-host=host.docker.internal:host-gateway \
           --env IOTRECEIVER_db__host=host.docker.internal \
           -dp 7770:7770 iot_receiver:$(poetry version -s)

You can also overwrite other parameters for the database connection using the IOTRECEIVER_db__XXXX environment variables.