Using Docker

The IoT Connector uses Docker to provide containers for the connected and individual services for the IoT connector.

There are containers for:

• Dashboard (NodeJS application servicing HTTP API requests)
• Postgres (Database backend for the dashboard [but not for storing telemetry data])
• Redis (For rate limiting the HTTP API)
• Nginx (For serving static assets [dashboard written using AngularJS] and proxying requests to both Dashboard and Websocket server)
• Websocket server (An alternative to MQTT for pushing actuations to IoT devices)
• Java backend (A spring application for interfacing with hbase backend to store telemetry data)
• HBase (Instance of hbase [with Zookeeper running in the same container] for storing telemetry data)
• Kafka (For streaming telemetry data to the rules engines running on Gearpump)
• Gearpump (Both local and services instance running)
• Rules-engine-build (For building the rules engine Gearpump application and deploying to Gearpump)

We docker-compose CLI plus docker-compose.yml file for storing the definitions of the containers; how they linked; ports that are opened; dependencies, etc.

We have used a Makefile to wrap up the commands to get going using Docker. To start the connector, please follow docs found here: https://github.com/Open-IoT-Service-Platform/platform-launcher#creating-the-iot-connector