Serverless AppSync Component

The AppSync Serverless Component allows you to easily and quickly deploy GraphQL APIs on AWS, and integrate them with AWS Lambda, DynamoDB & others. It supports all AWS AppSync features, while offering sane defaults that makes working with AppSync a lot easier without compromising on flexibility.

1. Install

$ npm install -g serverless

2. Create

Just create the following simple boilerplate:

$ touch serverless.yml # more info in the "Configure" section below
$ touch schema.graphql # your graphql schema file
$ touch index.js       # only required if you use a Lambda data source 
$ touch .env           # your AWS api keys

# .env
AWS_ACCESS_KEY_ID=XXX
AWS_SECRET_ACCESS_KEY=XXX

3. Configure

Basic Configuration

The following is a simple configuration that lets you get up and running quickly with a Lambda data source. Just add it to the serverless.yml file:

myLambda:
  component: "@serverless/aws-lambda"
  inputs:
    handler: index.handler
    code: ./

myAppSyncApi:
  component: "@serverless/aws-app-sync"
  inputs:
    # creating the API and an API key
    name: Posts
    authenticationType: API_KEY
    apiKeys:
      - myApiKey

    # defining your lambda data source
    dataSources:
      - type: AWS_LAMBDA
        name: getPost
        config:
          lambdaFunctionArn: ${myLambda.arn}

    # mapping schema fields to the data source
    mappingTemplates:
      - dataSource: getPost
        type: Query
        field: getPost

This configuration works with the following example schema. Just add it to the schema.graphql file right next to serverless.yml:

schema {
  query: Query
}

type Query {
  getPost(id: ID!): Post
}

type Post {
  id: ID!
  author: String!
  title: String
  content: String
  url: String
}

You'll also need to add the following handler code for this example to work:

exports.handler = async event => {
  var posts = {
    "1": {
      id: "1",
      title: "First Blog Post",
      author: "Eetu Tuomala",
      url: "https://serverless.com/",
      content:
        "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s."
    },
    "2": {
      id: "2",
      title: "Second Blog Post",
      author: "Siddharth Gupta",
      url: "https://serverless.com",
      content:
        "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s."
    }
  };

  return posts[event.id];
};

For more advanced usage, keep reading!

Custom Domains

You could optionally specify a custom domain for your GraphQL API, just add a domain property to the app sync component inputs:

myAppSyncApi:
  component: "@serverless/aws-app-sync"
  inputs:
    domain: api.example.com # add your custom domain here
    name: Posts
    # ... rest of config here

This would create a CloudFront distribution (aka CDN) for your AppSync API, which reduces request latency significantly, and would give you an SSL certificate out of the box powered by AWS ACM.

Please note that your domain (example.com in this example) must have been purchased via AWS Route53 and available in your AWS account. For advanced users, you may also purchase it elsewhere, then configure the name servers to point to an AWS Route53 hosted zone. How you do that depends on your registrar.

Create or Reuse APIs

The AppSync component allows you to either create an AppSync API from scratch, or integrate with an existing one. Here's how to create a new API:

# serverless.yml

myAppSync:
  component: '@serverless/aws-app-sync'
  inputs:
    name: 'my-api-name' # specifying a name creates a new API
    domain: api.example.com # example.com must be available in your AWS Route53
    authenticationType: 'API_KEY'
    apiKeys:
      - 'myApiKey'
    mappingTemplates:
      - dataSource: 'dynamodb_ds'
        type: 'Query'
        field: 'getMyField'
        request: 'mapping-template-request.vtl'
        response: 'mapping-template-response.vtl'
    functions:
      - dataSource: 'dynamodb_ds'
        name: 'my-function'
        request: 'function-request.vtl'
        response: 'function-response.vtl'
    dataSources:
      - type: 'AMAZON_DYNAMODB'
        name: 'dynamodb_ds'
        config:
          tableName: 'my-dynamo-table'
    schema: 'schema.graphql' # optional. Default behavior would look for schema.graphql file in the cwd

Reuse an existing AWS AppSync service by adding replacing the name input with apiId. This way the component will modify the AppSync service by those parts which are defined.

# serverless.yml

myAppSync:
  component: '@serverless/aws-app-sync'
  inputs:
    apiId: 'm3vv766ahnd6zgjofnri5djnmq'
    mappingTemplates:
      - dataSource: 'dynamodb_2_ds'
        type: 'Query'
        field: 'getMyField'
        request: 'mapping-template-request.vtl'
        response: 'mapping-template-response.vtl'
    dataSources:
      - type: 'AMAZON_DYNAMODB'
        name: 'dynamodb_2_ds'
        config:
          tableName: 'my-dynamo-table'

Authentication

The app using AppSync API can use four different methods for authentication.

API_KEY - Api keys
AWS_IAM - IAM Permissions
OPENID_CONNECT - OpenID Connect provider
AMAZON_COGNITO_USER_POOLS - Amazon Cognito user pool

When using OpenID connect method, inputs has to contain openIDConnectConfig block.

myAppSync:
  component: '@serverless/aws-app-sync'
  inputs:
    authenticationType: 'OPENID_CONNECT'
    openIDConnectConfig:
      issuer: 'NNN'
      authTTL: '1234'
      clientId: 'NNN'
      iatTTL: '1234'

When using Amazon Cognito user pools, userPoolConfig has to be defined.

myAppSync:
  component: '@serverless/aws-app-sync'
  inputs:
    authenticationType: 'AMAZON_COGNITO_USER_POOLS'
    userPoolConfig:
      awsRegion: 'us-east-1'
      defaultAction: 'ALLOW'
      userPoolId: 'us-east-1_nnn'

ApiKey can be created and modified by defining apiKeys.

myAppSync:
  component: '@serverless/aws-app-sync'
  inputs:
    apiKeys:
      - 'myApiKey1' # using default expiration data
      - name: 'myApiKey2'
        expires: 1609372800
      - name: 'myApiKey3'
        expires: '2020-12-31'

Schema

You can define the schema of your GraphQL API by adding it to the schema.graphql file right next to serverless.yml. Here's a simple example schema:

schema {
  query: Query
}

type Query {
  getPost(id: ID!): Post
}

type Post {
  id: ID!
  author: String!
  title: String
  content: String
  url: String
}

Alternatively, if you have your schema file at a different location, you can specify the new location in serverless.yml

  inputs:
    name: myGraphqlApi
    schema: ./path/to/schema.graphql # specify your schema location

Data Sources & Templates

The AppSync component supports 4 AppSync data sources and their corresponding mapping templates. You could add as many data sources as your application needs. For each field (or operation) in your Schema (ie. getPost), you'll need to add a mapping template that maps to a data source.

Here are the data sources that are supported:

Lambda Data Source

Here's an example setup for a Lambda data source:

serverless.yml

myLambda:
  component: "@serverless/aws-lambda"
  inputs:
    handler: index.handler
    code: ./

myAppSyncApi:
  component: "@serverless/aws-app-sync"
  inputs:
    # creating the API and an API key
    name: Posts
    authenticationType: API_KEY
    apiKeys:
      - myApiKey

    # defining your lambda data source
    dataSources:
      - type: AWS_LAMBDA
        name: getPost
        config:
          lambdaFunctionArn: ${myLambda.arn} # pass the lambda arn from the aws-lambda component above
      - type: AWS_LAMBDA
        name: addPost
        config:
          lambdaFunctionArn: ${myLambda.arn} # you could pass another lambda ARN, or the same one if it handles that field

    # mapping schema fields to the data source
    mappingTemplates:
      - dataSource: getPost
        type: Query
        field: getPost
      - dataSource: addPost
        type: Mutation
        field: addPost

        # Minimal request/response templates are added by default that works for 99% of use cases.
        # But you could also overwrite them with your own templates by specifying the path to the template files
        request: request.vtl
        response: response.vtl

index.js

exports.handler = async (event) => {
  var posts = {
    '1': {
      id: '1',
      title: 'First Blog Post',
      author: 'Eetu Tuomala',
      url: 'https://serverless.com/',
      content:
        "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s."
    },
    '2': {
      id: '2',
      title: 'Second Blog Post',
      author: 'Siddharth Gupta',
      url: 'https://serverless.com',
      content:
        "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s."
    }
  }

  return posts[event.id]
}

schema.graphql

schema {
  query: Query
}

type Query {
  getPost(id: ID!): Post
}

type Post {
  id: ID!
  author: String!
  title: String
  content: String
  url: String
}

DynamoDB Data Source

For the DynamoDB data source, you'll need to provide your own request/response templates that works according to your schema. Here's an example setup:

serverless.yml

# create a table with the aws-dynamodb component
myTable:
  component: '@serverless/aws-dynamodb'

appsync:
  component: '@serverless/aws-app-sync'
  inputs:
    name: Posts
    authenticationType: API_KEY
    apiKeys:
      - myApiKey
    dataSources:
      - type: AMAZON_DYNAMODB
        name: Posts
        config:
          tableName: ${myTable.name}
    mappingTemplates:
      - dataSource: Posts
        type: Mutation
        field: addPost
        request: request.vtl
        response: response.vtl

request.vtl

{
    "version" : "2017-02-28",
    "operation" : "PutItem",
    "key" : {
        "id" : { "S" : "${context.arguments.id}" }
    },
    "attributeValues" : {
        "author": { "S" : "${context.arguments.author}" },
        "title": { "S" : "${context.arguments.title}" },
        "content": { "S" : "${context.arguments.content}" },
        "url": { "S" : "${context.arguments.url}" },
        "ups" : { "N" : 1 },
        "downs" : { "N" : 0 },
        "version" : { "N" : 1 }
    }
}

response.vtl

$util.toJson($context.result)

schema.graphql

schema {
  query: Query
  mutation: Mutation
}

type Query {
  getPost(id: ID): Post
}

type Mutation {
  addPost(id: ID!, author: String!, title: String!, content: String!, url: String!): Post!
}

type Post {
  id: ID!
  author: String
  title: String
  content: String
  url: String
  ups: Int!
  downs: Int!
  version: Int!
}

ElasticSearch Data Source

Relational Database Data Source

Functions

4. Deploy

To deploy, just run the following command in the directory containing your serverless.yml file:

$ serverless

After few seconds (up to a minute if it's your first deployment), you should see an output like this:

  myAppSyncApi:
    apiId:   samrhyo7srbtvkpqnj4j6uq6gq
    arn:     arn:aws:appsync:us-east-1:552751238299:apis/samrhyo7srbtvkpqnj4j6uq6gq
    url:     "https://samrhyo7srbtvkpqnj4j6uq6gq.appsync-api.us-east-1.amazonaws.com/graphql"
    apiKeys:
      - da2-coeytoubhffnfastengavajsku
    domain:  "https://api.example.com/graphql"

  9s › myAppSyncApi › done

myApp (master)$

New to Components?

Checkout the Serverless Components repo for more information.

kod3r / aws-app-sync Goto Github PK

aws-app-sync's Introduction