Backend-Socket

Deploy a zero configuration serverless websockets backend in seconds using Serverless Components with just a few lines of code. The goal of this component is to offer the easiest experience possible for building realtime back-ends on AWS Lambda and AWS API Gateway.

1. Install
2. Create
3. Configure
4. Deploy

1. Install

$ npm install -g serverless

2. Create

$ mkdir my-socket && cd my-socket

The directory should look something like this:

|- socket.js
|- serverless.yml
|- package.json # optional
|- .env         # your development AWS api keys
|- .env.prod    # your production AWS api keys

the .env files are not required if you have the aws keys set globally and you want to use a single stage, but they should look like this.

AWS_ACCESS_KEY_ID=XXX
AWS_SECRET_ACCESS_KEY=XXX

The socket.js file is just a list of socket route/event handlers. Each handler takes two arguments, the first is the data passed in from the client, and the second is the socket object, which contains helpful data and methods.

// this function gets triggered on new connections
// if not provided, connections are successful by default
on('connect', async (data, socket) => {
  // the following data are available in the socket object
  // id represnets the connection id of a certain client
  const { id, domain, stage } = socket

  // you can return status codes directly
  return 200
})

// this function gets triggered whenever a client disconnects
// if not provided, disconnection is not handled
on('disconnect', async (data, socket) => {
  // e.g. business logic that removes connection ids from a db table
})

// this function gets triggered whenever a client sends data to the specified route
// in this example, you're handling the "message" route
// so clients need to send the following JSON data: { "route": "message", "data": { "foo": "bar" } }
on('message', async (data, socket) => {
  // you can send data to the connected client with the send() function
  await socket.send(data)
})

// this function gets triggered to handle any other data that is not handled above
on('default', async (data, socket) => {
  // you can also send data to a specific connection id (that you might have saved in a table)
  // this is very useful for a broadcasting functionality
  await socket.send(data, connectionId)
})

For a real world example of how the socket.js file could be used, take a look at how the chat app component is using it.

3. Configure

All the following inputs are optional. However, they allow you to configure your Lambda compute instance and pass environment variables.

# serverless.yml

name: my-socket
stage: dev

mySocket:
  component: '@serverless/backend-socket'
  inputs:
    name: my-socket
    description: My Socket Backend
    region: us-east-1
    memory: 128
    timeout: 10
    env:
      TABLE_NAME: my-table

    # the directory that contains the socket.js file.
    # If not provided, the default is the current working directory
    code: ./code

4. Deploy

socket (master)$ components

  Socket › outputs:
  url:  'wss://3v1fypmsz8.execute-api.us-east-1.amazonaws.com/dev/'
  code:
    runtime:  'nodejs8.10'
    env:  []
    timeout:  10
    memory:  512
  routes:  [ '$connect', '$disconnect', '$default' ]


  36s › dev › Socket › done

socket (master)$

New to Components?

Checkout the Serverless Components repo for more information.