Lambda Layer to Interact with AppSync API

This Lambda Layer wraps dependencies and provides a few helper functions for interacting with AWS AppSync GraphQL APIs from AWS Lambda in NodeJS.

At this time, the helper only provides basic query and mutate capability using AWS_IAM authentication and was adapted based on guidance from https://geromekevin.com/how-to-use-aws-appsync-in-lambda-functions/.

This layer is barebones at the moment - pull requests always welcome :)

Why?

So, why bother going through AppSync rather than hitting our backend (e.g. DynamoDB, RDS, etc.) directly from our Lambda?

Two reasons:

Using AppSync in our Lambda allows us to interact using the same model that we (presumably) use in our front-end. No need to maintain two different methods.
If you modify data directly in your backend (e.g. a DynamoDB API call or a SQL command in RDS), the change will not be detected by any AppSync subscriptions your clients have made, and they will miss the change. However, if our Lambda modifies data in our backend using an AppSync mutation, the change will be detected and our subscriptions will work as expected.

Deployment

Clone this repo
Run ./deploy.sh to create the layer and a demo Lambda function; if you don't want the demo function, just remove it from template.yaml.

Example

Import the helper layer to an AWS Lambda function, initialize the client, and invoke your endpoint:

const AppSyncHelper = require('appsync-helper');

var appSyncClient = new AppSyncHelper({
  url: 'https://YOUR_API_ENDPOINT.appsync-api.YOUR_API_REGION.amazonaws.com/graphql',         
  region: process.env.AWS_REGION,      
  auth_type: 'AWS_IAM',   
  accessKey:  process.env.AWS_ACCESS_KEY_ID,    
  secretKey: process.env.AWS_SECRET_ACCESS_KEY,   
  sessionToken: process.env.AWS_SESSION_TOKEN
});

exports.handler = async (event) => {
  try {
    var graphql = `
      query getItems {
        getItems {
          items {
          id,
          attribute1,
        },
        nextToken
        }
      }
    `;
    var response = await appSyncClient.query(graphql);
    console.log(JSON.stringify(response, null, 2));
  }
  catch(err) {
      console.log(err.message, err.stack);
  }  
  return ('Done!');  
};

Usage

Initialize the client

At the moment, the helper only supports AWS_IAM authentication. It wouldn't take much to add support for API_KEY or other methods.The access keys and session token environment vars are automatically provided to your function by AWS Lambda:

const AppSyncHelper = require('appsync-helper');

var appSyncClient = new AppSyncHelper({
  url: 'https://YOUR_API_ENDPOINT.appsync-api.YOUR_API_REGION.amazonaws.com/graphql',         
  region: process.env.AWS_REGION,      
  auth_type: 'AWS_IAM',   
  accessKey:  process.env.AWS_ACCESS_KEY_ID,    
  secretKey: process.env.AWS_SECRET_ACCESS_KEY,   
  sessionToken: process.env.AWS_SESSION_TOKEN
});

Query

var query = `
  query GetItem {
    getItem {
      id,
      attribute1,
      ...
    }
  }
`;

var response = await appSyncClient.query(
  query,      // string, required
  fetch_policy  // string, optional (default = 'network-only')
);

Mutation

var mutation = `
  mutation AddPost($name: String!) {
    addPost(name:$name) {
      id
      name
  }
}
`; 

var variables = {
  name: 'John Doe'
};

var response = await appSyncClient.mutate(
  mutation,      // string, required
  variables      // optional, object of key-value pairs for variables in your mutation (if any)
);

matwerber1/aws-appsync-lambda-layer-node