Drag-and-Drop Development on Serverless Applications: Using AWS Application Composer and AWS Serverless Application Model

In today's digital age, building and deploying applications to the cloud has become essential for businesses to stay competitive. Serverless computing has gained immense popularity among developers as it offers a cost-effective and scalable solution to deploy applications without managing the underlying infrastructure.

The AWS Serverless Application Model (SAM) is a popular open-source framework for building serverless applications with just few lines of code in a YAML template. During deployment, SAM will transform this template and expand the syntax into AWS CloudFormation magic.

But if we're really honest, nobody wakes up excited to maintain YAML code for their serverless application, even with SAM -- and visualizing applications' architecture is not easy.

AWS Application Composer offers one solution. It's a visual development tool that offers a drag-and-drop interface that offers developers quick and easy creation of new applications or modification of existing ones.

In this blog post, we will explore the benefits of using AWS SAM in combination with Application Composer for new and existing applications. We will also provide examples of how to use these tools to streamline our application development process and make it easier to manage our applications in the cloud. Let's dive in!

We will learn:

How to get started with a blank canvas and build a SAM application
How to load an existing SAM application, visualize and optionally extend it

If you want to quickly use AWS Application Composer to create an architecture diagram and a SAM application, check out this video:

This how-to video will walk you through features, options, and unique strengths of AWS Application Composer, Cloudcraft, and Lucidchart, empowering you to create visually stunning architecture diagrams.

About
✅ AWS experience	100 - Beginner
⏱ Time to complete	30 minutes
💰 Cost to complete	Free tier eligible
🧩 Prerequisites	- AWS Account and the CLI installed - Any Supported Local IDE from list - AWS SAM CLI - Node and NPM installed
💻 Code Sample	Code sample used in tutorial on Serverless Land
📢 Feedback	Any feedback, issues, or just a 👍 / 👎 ?
⏰ Last Updated	2023-06-09

Design a new Serverless Application

We are going to create a simple application with 3 AWS Serverless services: Amazon APIGateway, AWS Lambda and Amazon DynamoDB.

Image not found

This is a simple serverless pattern that demonstrates how making an HTTP POST request to an Amazon API Gateway endpoint invokes an AWS Lambda function and inserts an item into an Amazon DynamoDB table.

We will create a Nodejs application with two Lambda functions - one to insert data into the DynamoDB table and another to list data.

Let's start with creating a folder in our local computer. Name this folder buildon-sam-app. We will place all the contents from this blog post into this folder.

In the AWS Management Console, go to the Application Composer service.

Image not found

Click on the Create project button in the Application Composer Dashboard to create a new project. As displayed in the image below, for Type of Project, select the radio button for New blank project.

Image not found

In the New blank project we have 2 modes: Connected and Unconnected.

Application Composer is only in our browser, it does not store any of our files in an AWS account. So the selection of mode determines how and where our design files will be stored.

In the Connected mode, we can provide application composer access to a local folder on our computer. It will then automatically sync and save our template file and project folders locally as we design.
In the Unconnected mode, we have to manually import and export our template files. We have to select Menu > save changes as often as needed to save and download the latest configuration of our template file. When designing in unconnected mode, only the application template file is generated and can be manually exported.

For the sake of this blog, we will be using Connected mode.

Click on Select folder and provide an empty local folder.

Remember all this is happening in our browser! So we have to provide access to the browser to view and edit the local file in Connected mode. When prompted to allow access, select View and Edit files.

Image not found

Now we have a blank project. In the image below, I have opened a local IDE like Visual Studio Code in the same folder to show you how the local files are created automatically by Application Composer.

Image not found

The following files and folders are synced and saved to our local file system when designing in connected mode:

Application template file – When we design in Application Composer, a singular application template file is generated.
Project folders – When we design a Lambda function, a general Lambda directory structure is generated.
Backup template file – A backup directory named .aws-composer will be created at the root of our project location and will contain a backup copy of our application template file and our project folders.

The Application composer console has two sections: "Canvas" and "Template". You can switch between the two anytime while designing.

Image not found

In the Application Composer console, we can find most of the serverless resources available in the Resources tab. The List tab is to quickly identify resources already on the canvas/template. To make this pattern, drag and drop components from the Resources tab onto the Canvas.

When we drag the resources on the canvas, Application Composer automatically writes the required YAML code in the Template. Also, as this is connected mode, these changes are done to our local template file. Any changes we make on either canvas or template are automatically reflected on other side.

We can see in the GIF below, as we drag and drop resources (API Gateway, Lambda, and DynamoDB) on the Canvas, the template.yaml is automatically populated with its appropriate yaml definition.

Image not found

We can customize the template further in the "Template" tab. Or we can select the resource on "Canvas" and click "Details" to modify resources.

For example, we dragged API Gateway resource; by clicking on "Details" we can see configurable "Resource properties" on the right. We can rename CloudFormation's logical ID of the resource, quickly add routes, authorizers, and CORS configuration.

Image not found

For our new SAM application, define the following resources and their properties. Make sure you click on save every time resource properties are modified:

Resource: API Gateway
- Properties:
  - Name: Api
  - Routes:
  - Method: GET; Path: /
  - Method: POST; Path: /
Resource: Lambda Function
- Properties:
  - Name: CustomerFunctionCreate
  - Source path: src/CreateCustomer
  - Runtime: nodejs18.x
  - Handler: index.handler
Resource: Lambda Function
- Properties:
  - Name: CustomerFunctionList
  - Source path: src/ListCustomer
  - Runtime: nodejs18.x
  - Handler: index.handler
Resource: DynamoDB Table
- Properties:
  - LogicalId: CustomerTable
  - Partition key: id
  - Partition key type: String

Groups

We have 2 Lambda functions to Create and List customers. We can group these functions into a Functions group for better visibility. Click on one of the functions and select Group.

This will create a group. Drag and drop the other Lambda function to be part of this group. If we double-click this group we can modify its name. We will rename the group CustomerFunctionsGroup.

Image not found

Best practices

AWS Application Composer is designed to configure Infrastructure-as-Code definitions that follow the AWS best practices. For example, when we add a Lambda function to the canvas, Application Composer will enable "Tracing" by default and add Amazon CloudWatch LogGroup for the function. Lambda logs all requests handled by our function and also automatically stores logs generated by our code through Amazon CloudWatch Logs.

Connections

Once we have all the four components, let's connect these: APIGateway -> Lambda -> DynamoDB.

When we make these connections, Application Composer will automatically update the template to reflect this. Notice the changes in the image below:

Image not found

We have to connect the GET / route to CustomerFunctionList and the POST / to CustomerFunctionCreate. Also connect both the Lambda functions to the DynamoDB table.

We have 2 options to make these connections:

We can manually connect them using the "Canvas" as shown in the image above. (Notice that we can only make connections where service integrations are possible.)

Image not found

Or we can simply copy and paste the following template in the "Template" section to see the design pattern come alive.

Image not found

template.yaml

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
Transform: AWS::Serverless-2016-10-31
Resources:
  CustomerFunctionList:
    Type: AWS::Serverless::Function
    Properties:
      Description: !Sub
        - Stack ${AWS::StackName} Function ${ResourceName}
        - ResourceName: CustomerFunctionList
      CodeUri: src/ListCustomer
      Handler: index.handler
      Runtime: nodejs18.x
      MemorySize: 3008
      Timeout: 30
      Tracing: Active
      Events:
        Api23GET:
          Type: Api
          Properties:
            Path: /
            Method: GET
            RestApiId: !Ref Api23
      Environment:
        Variables:
          TABLE_NAME: !Ref CustomerTable
          TABLE_ARN: !GetAtt CustomerTable.Arn
      Policies:
        - DynamoDBCrudPolicy:
            TableName: !Ref CustomerTable
  CustomerFunctionListLogGroup:
    Type: AWS::Logs::LogGroup
    DeletionPolicy: Retain
    Properties:
      LogGroupName: !Sub /aws/lambda/${CustomerFunctionList}
  CustomerFunctionCreate:
    Type: AWS::Serverless::Function
    Properties:
      Description: !Sub
        - Stack ${AWS::StackName} Function ${ResourceName}
        - ResourceName: CustomerFunctionCreate
      CodeUri: src/CreateCustomer
      Handler: index.handler
      Runtime: nodejs18.x
      MemorySize: 3008
      Timeout: 30
      Tracing: Active
      Environment:
        Variables:
          TABLE_NAME: '!Ref CustomerTable'
          TABLE_ARN: '!GetAtt CustomerTable.Arn'
          TABLE_NAME_2: !Ref CustomerTable
          TABLE_ARN_2: !GetAtt CustomerTable.Arn
      Events:
        Api23POST:
          Type: Api
          Properties:
            Path: /
            Method: POST
            RestApiId: !Ref Api23
      Policies:
        - DynamoDBCrudPolicy:
            TableName: !Ref CustomerTable
  CustomerFunctionCreateLogGroup:
    Type: AWS::Logs::LogGroup
    DeletionPolicy: Retain
    Properties:
      LogGroupName: !Sub /aws/lambda/${CustomerFunctionCreate}
  CustomerTable:
    Type: AWS::DynamoDB::Table
    Properties:
      AttributeDefinitions:
        - AttributeName: id
          AttributeType: S
      BillingMode: PAY_PER_REQUEST
      KeySchema:
        - AttributeName: id
          KeyType: HASH
      StreamSpecification:
        StreamViewType: NEW_AND_OLD_IMAGES
  Api:
    Type: AWS::Serverless::Api
    Properties:
      Name: !Sub
        - ${ResourceName} From Stack ${AWS::StackName}
        - ResourceName: Api23
      StageName: Prod
      DefinitionBody:
        openapi: '3.0'
        info: {}
        paths:
          /:
            get:
              x-amazon-apigateway-integration:
                httpMethod: POST
                type: aws_proxy
                uri: !Sub arn:${AWS::Partition}:apigateway:${AWS::Region}:lambda:path/2015-03-31/functions/${CustomerFunctionList.Arn}/invocations
              responses: {}
            post:
              x-amazon-apigateway-integration:
                httpMethod: POST
                type: aws_proxy
                uri: !Sub arn:${AWS::Partition}:apigateway:${AWS::Region}:lambda:path/2015-03-31/functions/${CustomerFunctionCreate.Arn}/invocations
              responses: {}
      EndpointConfiguration: REGIONAL
      TracingEnabled: true
Metadata:
  AWS::Composer::Groups:
    Group:
      Label: CustomerFunctionsGroup
      Members:
        - CustomerFunctionList
        - CustomerFunctionCreate
Outputs:
  EndpointUrl:
    Description: HTTP REST endpoint URL
    Value: !Sub https://${Api23}.execute-api.${AWS::Region}.amazonaws.com/Prod

Before we proceed, add the following Output section at the end of the template.yaml to get the API Endpoint once the stack is created:

1
2
3
4
Outputs:
  EndpointUrl:
    Description: HTTP REST endpoint URL
    Value: !Sub https://${Api}.execute-api.${AWS::Region}.amazonaws.com/Prod

Our design should look like this:

Image not found

Add Nodejs code

In the local folder, Application Composer would have created the following folder structure and files:

Image not found

We now have to add Nodejs code, so whenever the APIs are invoked they will add/list customer information from the DynamoDB table.

Add the following code in respective files:

CreateCustomer/index.js

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
const { DynamoDBClient } = require("@aws-sdk/client-dynamodb");
const { DynamoDBDocumentClient, PutCommand } = require("@aws-sdk/lib-dynamodb");

const client = new DynamoDBClient({});
const ddbClient = DynamoDBDocumentClient.from(client);

exports.handler = async event => {
  try {
    console.log(JSON.stringify(event, undefined, 2));
    const requestBody = JSON.parse(event.body);

    const customer = {
      id: Date.now().toString(), // Use a timestamp as the ID
      ...requestBody,
    };

    console.log(`Adding customer with ID '${customer.id}' to table '${process.env.TABLE_NAME}' with attributes: ${JSON.stringify(customer, null, 2)}`);

    const params = {
      TableName: process.env.TABLE_NAME,
      Item: customer,
    };

    // Add data to DynamoDB table
    await ddbClient.send(new PutCommand(params));
   
    console.log(`Successfully saved customer '${customer.id}'`);
   
    return {
      statusCode: 201,
      body: JSON.stringify(customer),
    };
  } catch (err) {
    console.error(err);

    return {
      statusCode: 500,
      body: 'Internal Server Error',
    };
  }
};

CreateCustomer/package.json

1
2
3
4
5
6
7
{
  "name": "function",
  "version": "1.0.0",
  "devDependencies": {
    "@aws-sdk/lib-dynamodb": "^3.214.0"
  }
}

ListCustomer/index.js

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
const { DynamoDBClient, ScanCommand } = require("@aws-sdk/client-dynamodb");

const client = new DynamoDBClient({});

exports.handler = async event => {
  try {
    // Log the event argument for debugging and for use in local development.
    console.log(JSON.stringify(event, undefined, 2));

    console.log(`Listing customers from table '${process.env.TABLE_NAME}'`);

    let ids = [];

    // Loop over all customers. If there are more customers after a request, 
    // the response LastEvaluatedKey will have a non-null value that we can
    // pass in the next request to continue fetching more customers.
    let 
    lastEvaluatedKey;
    do {
      const command = new ScanCommand({
        TableName: process.env.TABLE_NAME,
        ProjectionExpression: "id",
        ExclusiveStartKey: lastEvaluatedKey
      });

      const response = await client.send(command);

      const additionalIds = response.Items.map(customer => customer.id);

      ids = ids.concat(additionalIds);

      lastEvaluatedKey = response.LastEvaluatedKey;
    } while (lastEvaluatedKey);

    console.log(`Successfully scanned for list of IDs: ${JSON.stringify(ids, null, 2)}`);

    return {
      statusCode: 200,
      body: JSON.stringify({
        ids
      })
    };
  } catch (err) {
    console.error(`Failed to list customers: ${err.message} (${err.constructor.name})`);

    return {
      statusCode: 500,
      body: "Internal Service Error"
    };
  }
};

ListCustomer/package.json

1
2
3
4
5
6
7
{
  "name": "function",
  "version": "1.0.0",
  "devDependencies": {
    "@aws-sdk/lib-dynamodb": "^3.214.0"
  }
}

Build and Deploy with AWS SAM

Now that we have our design and application files created, let's build and deploy this thing using AWS Serverless Application Model (SAM).

Open a terminal in your local IDE in the same root folder where Application Composer created all the above files. Make sure you installed all the pre-requisites mentioned at the beginning.

As this is a Nodejs sample, let's also make sure we have all the required modules installed for the application. Run the following command at the root of the above selected folder:

1
npm install aws-sdk

We will be skipping the sam init command, as we already have a sample application created for us. We will build, deploy, and test the application in this section.

Step 1: Build the application

In this step, we use the AWS SAM CLI to build our application and prepare for deployment. When we build, the AWS SAM CLI creates a .aws-sam directory and organizes our function dependencies, project code, and project files there.

1
sam build

Output:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
➜  buildon-sam-app sam build
Building codeuri: ../buildon-sam-app/src/ListCustomer runtime: nodejs18.x metadata: {} architecture: x86_64 functions: CustomerFunctionList
Running NodejsNpmBuilder:NpmPack
Running NodejsNpmBuilder:CopyNpmrcAndLockfile
Running NodejsNpmBuilder:CopySource
Running NodejsNpmBuilder:NpmInstall
Running NodejsNpmBuilder:CleanUpNpmrc
Running NodejsNpmBuilder:LockfileCleanUp
Building codeuri: ../buildon-sam-app/src/CreateCustomer runtime: nodejs18.x metadata: {} architecture: x86_64 functions: CustomerFunctionCreate
Running NodejsNpmBuilder:NpmPack
Running NodejsNpmBuilder:CopyNpmrcAndLockfile
Running NodejsNpmBuilder:CopySource
Running NodejsNpmBuilder:NpmInstall
Running NodejsNpmBuilder:CleanUpNpmrc
Running NodejsNpmBuilder:LockfileCleanUp

Build Succeeded

Built Artifacts:  .aws-sam/build
Built Template :  .aws-sam/build/template.yaml

Commands you can use next
=========================
[*] Validate SAM template: sam validate
[*] Invoke Function: sam local invoke
[*] Test Function in the Cloud: sam sync --stack-name {{stack-name}} --watch
[*] Deploy: sam deploy --guided

Step 2: Deploy the application to the AWS Cloud

1
sam deploy --guided

This will give multiple prompts, which you can answer as follows:

Stack Name: buildon-sam-app
AWS Region: <REGION>
#Shows you resources changes to be deployed and require a 'Y' to initiate deploy. Confirm changes before deploy: y
#SAM needs permission to be able to create roles to connect to the resources in your template. Allow SAM CLI IAM role creation: y
#Preserves the state of previously provisioned resources when an operation fails.
Disable rollback: n
CustomerFunctionList may not have authorization defined, Is this okay?: y
CustomerFunctionCreate may not have authorization defined, Is this okay?: y
Save arguments to configuration file: y
SAM configuration file: samconfig.toml
SAM configuration environment: default
Deploy this changeset?: y

This will deploy the CloudFormation stack with all the resources defined in the SAM template.yaml generated by Application Composer, including all the source code for Lambda functions.

In the end, we should see a prompt as follows:

Output:

1
2
3
4
5
6
7
8
9
10
11
12
...
CREATE_COMPLETE          AWS::CloudFormation::Stack   buildon-sam-app         -                                                    
------------------------------------------------------------------------
CloudFormation outputs from deployed stack
Outputs                    
------------------------------------------------------------------------
Key                 EndpointUrl                                                                 
Description         HTTP REST endpoint URL                                                      
Value               https://11yakge1yd.execute-api.<REGION>.amazonaws.com/Prod                 
------------------------------------------------------------------------

Successfully created/updated stack - buildon-sam-app in <REGION>

Note the HTTP REST API endpoint URL; we will be using this to call the API.

It will also create a local configuration file samconfig.toml, so that next time we run sam deploy it gets all the inputs from this file. The S3 bucket name is just an example here that SAM will use, we can set a different default S3 bucket here. The REGION will be your AWS Region.

samconfig.toml

1
2
3
4
5
6
7
8
9
10
11
version = 0.1
[default]
[default.deploy]
[default.deploy.parameters]
stack_name = "buildon-sam-app"
s3_bucket = "aws-sam-cli-managed-default-samclisourcebucket-<randomletters>"
s3_prefix = "buildon-sam-app"
region = "REGION"
confirm_changeset = true
capabilities = "CAPABILITY_IAM"
image_repositories = []

Step 3: Test the Application

Now that our application is up and running, let's call the API and add an item to our DynamoDB table. Run the following code, replacing the <API_ENDPOINT> with the URL you noted in previous step. It will look something like this: https://<API>.execute-api.<REGION>.amazonaws.com/Prod.

Step 3.1: Add a customer to the table

1
2
3
4
5
6
7
8
curl -X POST \                           
   https://<API_ENDPOINT> \      
  -H 'Content-Type: application/json' \
  -d '{
    "name": "John Doe",
    "email": "johndoe@example.com"
  }'
   

Output:

1
{"id":"1680294275167","name":"John Doe","email":"johndoe@example.com"}%    

Step 3.2: List the customer ids

1
curl -X GET https://<API_ENDPOINT>

Output:

1
{"ids":[{"S":"1680294275167"}]}%    

🎉 That's it!

You can add more APIs and Lambda functions to this serverless application. Every time you have to make application changes, run step 1 and step 2 to build and deploy the application!

Visualize an Existing Application

Now let's say we want to visualize an existing application in Application Composer. To do so, clone the git repository using the following commands:

1
2
git clone https://github.com/aws-samples/fresh-tracks.git
cd fresh-tracks/backend/FreshTracks/

Open Application Composer in the AWS Console, and if you are in a design, make sure it is saved and then click on the home 🏠 icon.

This time in the Create project dialog box, select the radio option Load existing project.

Use Connected mode again, and in the Project location, navigate to the folder fresh-tracks/backend/FreshTracks/ that has this web applications backend template.yaml. Provide permissions for Application Composer in the browser to View the folders and files.

Now that Application Composer can read the folder, it will list all the .yaml files it can find in the folder. Here let's select template.yaml and select Create to visualize it.

Image not found

Provide the Edit access and we can see the visualization come to life:

Image not found

🎉 That's it!

Now, we can visualize and drag-and-drop more resources to continue building the application in Application Composer. The rest of the build and deployment steps in our IDE will remain same.

Clean Up

When you're ready to finish, remember to delete the CloudFormation stacks we created here using the following steps:

Delete the stack. Replace STACK_NAME with the stack name. In the above example our stack name was buildon-sam-app.```bash
sam delete --stack-name STACK_NAME
Confirm the stack has been deleted in the AWS CloudFormation console or using the following command. Replace STACK_NAME with the stack name and ensure the name is within single quotes.```bash
aws cloudformation list-stacks --query "StackSummaries[?contains(StackName,'STACK_NAME')].StackStatus"

Conclusion

Congratulations! We just learned how to use AWS Application Composer to visualize and build Serverless applications. We also learned how to build and deploy these applications using the AWS SAM CLI.

Please note the following points:

There is no cost for visualizing applications in Application Composer. However, if you run the commands sam build and sam deploy in your AWS Account, you will be charged for the resources it creates. This is outside of the scope of this blog and may go beyond your free-tier limits.
Application Composer is going to help you create the visualized template, but you should continue using your local testing, peer review, or regular deployment through your own current method.

You can find an amazing collection of serverless repositories on Serverless Land.

If you enjoyed this tutorial, found any issues, or have feedback us, please send it our way!

Check out my other tutorial on how to create a CI/CD pipeline from scratch using Amazon CodeCatalyst, to deploy Infrastructure as Code (IaC) with AWS CloudFormation.

For more DevOps related content, check out our DevOps Essentials guide and learn about How Amazon Does DevOps in Real Life.

Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.

Select your cookie preferences

Site Terms, Privacy, and more.

Drag-and-Drop Development on Serverless Applications: Using AWS Application Composer and AWS Serverless Application Model

AWS Application Composer and AWS Serverless Application Model (SAM) can help you visually design and build serverless applications. Here's how.

Design a new Serverless Application

Groups

Best practices

Connections

Add Nodejs code

Build and Deploy with AWS SAM

Step 1: Build the application

Step 2: Deploy the application to the AWS Cloud

Step 3: Test the Application

Step 3.1: Add a customer to the table

Step 3.2: List the customer ids

Visualize an Existing Application

Clean Up

Conclusion

Comments