KEECO-HomeAutomation / keeco-hub

The central software which orchestrates all the hardware nodes and provides an API for the client access.

keeco-hub

The central software which orchestrates all the hardware nodes and provides an API for the client access.

NPM Scripts

• npm start : start the application in development mode, with automatic reloading (see caveats in CONTRIBUTING.md)
• npm run build : Compile the code with babel (output directory: build)
• npm test : Run all the tests
• npm run test:coverage : Run all the tests and generate coverage reports
• npm run lint : Lint your code. It will also error out on format errors
• npm run format : Format every .js file using Prettier.

Using docker

We have a docker container which has all the dependencies installed for the hub. You can use it, if you want an easy, plug'n'play solution.

Run the stable container with:

docker run -dit --name keeco_hub --restart always -v keeco_hub_conf:/opt/keeco-hub/config -p 5000:5000 -p 1883:1883 -p 5353:5353 keeco/keeco-hub

Run the development container with:

docker run --rm -it -v keeco_hub_dev_conf:/opt/keeco-hub/config -p 5000:5000 -p 1883:1883 -p 5353:5353 keeco/keeco-hub:dev

The ports used:

• 5000 for graphQL API
• 1883 for MQTT
• 5353 for mDNS. If you don't need it, you can leave it out.

We use a persistent volume managed by docker (keeco_hub_conf), but you can use a local binding as well.

Which is which

• src/index.js : The entrypoint of the hub. It just initializes every module.
• src/aedes : The home of the MQTT server
• src/apollo : The home of the GraphQL server
  • index.js : Sets up the server
  • schema : Contains the GQL schema
    • index.js : Merges all the typeDefs and resolvers
    • Query : The base query which is extended
      • schema.graphql : The main schema
      • resolvers.js : The main resolvers
    • ... (so on for the rest of the directories)
• src/sqlite : The home of the SQLite instance
  • index.js : Sets up the database
  • populate.js : Populates the empty database on the first start
  • structure.js : Contains the structure of the SQLite database
• src/connector : The home of the class that connects and controlls all the different data stores and APIs
  • index.js : The base class which imports all the modules
  • The rest of the files are modules of the connector
• src/utils : Contains utility functions

Auto provisioning

Provision JSONs should be passed to the MQTT server as the username. Password is omitted in this case.

Example provision JSON:

{
	"uuid": "5e35d62e-a832-41fe-9660-3768c427addc",
	"name": "KEECO Switch",
	"endpoints": [
		{
			"name": "relayPin1",
			"output": true,
			"range": "0,1"
		},
		{
			"name": "relayPin2",
			"output": true,
			"range": "0,1"
		}
	],
	"templates": [
		{
			"name": "switch",
			"mappings": [
				{
					"name": "on",
					"endpoint": "relayPin1"
				}
			]
		},
		{
			"name": "switch",
			"mappings": [
				{
					"name": "on",
					"endpoint": "relayPin2"
				}
			]
		}
	]
}

Provision JSON fields:

• uuid : Mandatory field. Should be universally unique
• name : The node can name itself. This can be changed later by the user. Optional
• endpoints : Array of endpoints
  • name : The name of the pin. Should be unique per node
  • output : Set to true if the pin is an output pin. Optional. If omitted defaults to false
  • range : The node can define what values is it accepting/will it produce. Examples: 0:15, .002:.4, 1,2,3. See more. Optional
• templates : Array of templates. The templates are which tell the server how to work with a node. Required, but can be an empty array
  • name : Name of the template. Should match one of the names listed in src/apollo/schema/nodes/TemplateData.graphql. For example the template name for TemplateDataSwitch should be switch. Case sensitive
  • mappings : Mappings for the template
    • name : The name of the pin that should be present for the selected template. Case sensitive
    • endpoint : The name of the endpoint that is mapped to the template pin