AWS Lambda Reference Architecture for creating a Mobile Backend. You can build backends using AWS Lambda and Amazon API Gateway to authenticate and process API requests. Lambda makes it easy to create rich, personalized app experiences. The architecture described in this diagram can be created with a CloudFormation template.
Template One does the following:
-
Configures the Amazon Simple Storage Service (Amazon S3) bucket to receive user uploaded photos.
-
Configures the Amazon CloudFront distribution to display static content for uploaded media.
-
Configures the Amazon DynamoDB Table for storing mobile data from the iOS application.
Template Two does the following:
-
Creates a Lambda function to stream updates from DynamoDB and index in CloudSearch.
-
Creates a Lambda function to store mobile data into DynamoDB.
-
Creates a Lambda function to query CloudSearch and return a matching document.
-
Creates an AWS Identity and Access Management (IAM) role and policy for all three lambda functions to assume when invoked. Permissions allow the functions to write output to Amazon CloudWatch Logs, store data in DynamoDB, read from DynamoDB Streams, and store data in CloudSearch.
Important: CloudSearch, API Gateway, and DynamoDB Streams will be configured through the console. During cleanup, you will need to manually delete these resources after deleting the AWS CloudFormation Stacks. The provided CloudFormation template retreives its Lambda code from a bucket in the us-east-1 region. To launch this sample in another region, please modify the template and upload the Lambda code to a bucket in that region.
Step 1 – Create an AWS CloudFormation Stack with Template One and copy the S3 bucket name from the AWS CloudFormation output.
Step 2 – Create a CloudSearch domain using the AWS console. Create a manual index containing attributes for headline, s3_url, user_id similar to the diagram below:
Step 3 – Update the following environment variables in each Lambda function based on the output of Steps 1 and 2.
a CLOUDSEARCH_DOCUMENT_ENDPOINT in streams-data-function/index.js with the Document Endpoint of CloudSearch
b CLOUDSEARCH_SEARCH_ENDPOINT in search-data-function/index.js with the Search Endpoint of CloudSearch
Step 4 – Upload a .zip file of each lambda function to Amazon S3 and create an AWS Cloudformation Stack with Template Two.
Step 5 – Add the created DynamoDB table as an event source for your streams-data-function in the AWS console.
In order to illustrate the end-to-end process, you can integrate with the sample mobile application available. The sample mobile application is built for iOS and requires creating a mobile SDK. The steps for integrating the SDK are described below:
Step 1 - Create a new Amazon Cognito identity pool through the Amazon Cognito dashboard for unauthenticated users. Modify the policy document to allow unauthenticated users to "execute-api:*" for API Gateway. Modify the policy document to allow users to upload to the S3 bucket created in Template One.
Step 2 - Visit the API Gateway dashboard in your AWS account and create a new resource endpoints for /notes. Assign a POST method for the /notes endpoint. For the method, select the Integration Request type of “Lambda Function.” Configure the notes endpoint to use the notes-data-function.
Under Models section, create a CreateNoteRequest and a CreateNoteResponse model using these JSON templates .
Under Method Request for the method execution, enable API key required and assign the CreateNoteRequest model that was created earlier as the Request Model.
Under Method Response for the method exectuion, for a 200 response code select a content type of application/json and use the CreateNoteResponse model that was created earlier.
Step 3 - In the API Gateway dashboard create an API key for API Gateway and then deploy the API Gateway in order to copy the deployment endpoint url.
Step 4 - Install and run cocoapods on the Command Line Interface:
$ pod installStep 5 - Open the Constants.swift file and add the Account Id, S3 bucket, Amazon Cognito identity pool, Amazon Cognito identity users, API key, and API Gateway endpoint as constants.
Step 6 - Run the mobile application in the simulator. Choose a photo and upload it to S3. Then view the iamge is uploaded in Amazon S3, and then use the Amazon CloudFront Distribution url to view the image through the CDN. Then select the button to add a note in the iOS application. Add a note in the mobile application and save. Then view DynamoDB to see the note added to the Notes Table. View the CloudSearch domain to see a document added to your search index. Review the Amazon CloudWatch Log events from the streams Lambda function for evidence that the functions are pulling data as mobile users are publishing.
Congratulations! You now should have a working example of a mobile backend reference architecture. You are able to communicate directly to mobile services such as Amazon Cognito for identities and upload media files directly to Amazon S3. You also configured a serverless API using AWS Lambda and Amazon API Gateway.
To remove all automatically created resources, delete the two AWS CloudFormation stacks. You will need to manually remove the API Gateway endpoint, Amazon Cognito identity pool, and CloudSearch domain.
Note: Deletion of the S3 bucket will fail unless all files in the bucket are removed before the stack is deleted.
This reference architecture sample is licensed under Apache 2.0.