IoT Websocket Server

NOTE: This project reached end of support and is no longer maintained. Parts of the project are continued in https://github.com/IndustryFusion/DigitalTwin.

This component is used to support actuations over websocket protocol. It is an alternative for mqtt protocol.

Server listens on determined port. It uses 'echo-protocol' for receiving and sending requests.

Websocket client also has to use 'echo-protocol'. After the establishment of connection client has to send initial json object message:

Key	Description
type	type of message, always "device" for clients
deviceId	identifier of your device in iotkit database
deviceToken	token of device which is received after activation

Example json message:

{
    "type": "device",
    "deviceId": <deviceId>,
    "deviceToken": "eyJ0eXA....asdfwefadA"
}

Basic response codes:

Code	Description
401	"Invalid device token" or "Wrong message format"
500	Internal server error
200	Subscribed to ws server
1024	Received actuation message

Example of actuations message:

{
    "code":1024,
    "content": {
        "type":"command",
        "transport":"ws",
        "content": {
            "accountId":"c4715841-bb1e-4c87-ad78-2376fe596a11",
            "deviceId":"example001",
            "gatewayId":"example001",
            "componentId":"1e447b10-9afd-4221-b4d8-7bc5c5bb6c3f",
            "command":"LED.v1.0","params":[{"name":"LED","value":"1"}]
        }
    }
}

More informations about websockets on website https://github.com/theturtle32/WebSocket-Node/blob/master/docs/index.md

Pre-requirements

• Node.js v0.10.x

  • Latest in https://launchpad.net/~chris-lea/+archive/ubuntu/node.js

• grunt-cli

  • npm install grunt-cli

• Git

• Cloud Foundry CLI and Trusted Analytics Platform account (https://github.com/trustedanalytics)

Requirements to run

1. Generate a pair RSA keys and put your public keys in ./keys/ directory as public.pem

On Trusted Analytics Platform (https://github.com/trustedanalytics)

Before installation, make sure that you are logged into Trusted Analytics Platform with command:

cf login

1. Create instances with specified name for each of required services from marketplace:

   • PostgreSQL 9.3 or newer with name mypostgres

2. Create following user-provided services with properties filled with real values:

    cf cups websocket-ups -p "{\"username\":\"${USER}\",\"password\":\"${PASSWORD}\"}"
   

3. Executing ./cf-deploy.sh in main repository catalog builds package and pushes it to CF as an app with name {SPACE}-dashboard where space is currently selected space by cf t -s "SPACE"

4. Check logs and wait for application start.