Telemetry end-to-end

A package showing telemetry data acquisition end-to-end: Microcontroller firmware to cloud side data processing, using Drogue IoT and Apache Kafka in the process.

The tutorial involves a lot of components, and a lot can go wrong, especially when it comes to bluetooth or network connectivity.

If you run into any blockers, we would kindly ask you to reach out to us and report your findings.

But, there is also a way to simulate the actual hardware and work around some requirements.

Simulating a message

You can manually simulate a message, that the IoT gateway would normally send after it acquired the information from the sensor.

Ensure that the DOMAIN variable is still set to the value you used for the installation. Then execute the following command:

http --auth device-1@eclipse:foobar POST http://http-endpoint$DOMAIN/v1/foo temp:=23

If you are logged in with `drg`:

http --auth device-1@eclipse:foobar POST $(drg whoami -e http)/v1/foo temp:=23

Different protocol

You might have notices, we used HTTP for this. However, Kura would send the message using MQTT. As one part of Drogue IoT is normalizing the IoT protocol, the backend system didn't even notice.

Reading directly from the digital twin

When data gets sent to Ditto, it gets forwarded to Streamsheets for visualization. Should the data not show up in streamsheets, or should you run into issues with Streamsheets, you can also directly query the data in Ditto.

Ensure that the DOMAIN variable is still set to the value you used for the installation. Then execute the following command:

curl -sL -X GET "http://ditto-default$DOMAIN/api/2/things/eclipse:device-1" -H  "accept: application/json" -H  "Authorization: Basic ZGl0dG86ZGl0dG8=" | jq