AWS SAM method response

[royassis]

Hey there

Does SAM templates support method responses at this time?

[GavinZZ]

Hi there, thanks for reaching out and creating a discussion post. AWS::Serverless::Api currently do not explicitly support method response. However, a workaround is that you can define CloudFormation resource AWS::ApiGateway::Method. Here is an example provided:

Resources:
  MyApi:
    Type: AWS::Serverless::Api
    Properties:
      StageName: MyNewStageName

  MyApiGatewayRootMethod:
    Type: AWS::ApiGateway::Method
    Properties:
      HttpMethod: GET
      AuthorizationType: NONE
      Integration:
        Type: MOCK
      ResourceId: !GetAtt MyApi.RootResourceId
      RestApiId: !Ref MyApi
      MethodResponses: <your-desired-setup>

This would allow you to use method responses along with AWS::Serverless::Api resource.

[GavinZZ]

---

An update to my response. The above answer is incorrect. I noticed that when I deploy this template using sam deploy --guided, it would fail because Resource handler returned message: "No integration defined for method.

After some investigation, this is expected behaviour according to this documentation. Unfortunately, because AWS::ApiGateway::Deployment is a generated resource, there is no simple way to add DependsOn on this generated resource.

That means there is currently no way to explicitly define a AWS::ApiGateway::Method with a SAM API resource. Will bring this issue up to team discussion.

---

Currently, the only solution is to define your custom OpenAPI definition files or use inline swagger format like

Resources:
  MyApi:
    Type: AWS::Serverless::Api
    Properties:
      StageName: prod
      DefinitionBody:
        openapi: "3.0.0"
        info:
          title: My API
          version: "1.0"
        paths:
          /my-resource:
            get:
              responses:  # <----- define this
                '200':
                  description: Successful response
                  content:
                    application/json:
                      schema:
                        type: object
                        properties:
                          id:
                            type: string
                          name:
                            type: string
                '400':
                  description: Bad Request
                  content:
                    application/json:
                      schema:
                        type: object
                        properties:
                          error:
                            type: string
              x-amazon-apigateway-integration:  # <----- define this
                type: aws_proxy  # This can be a different integration type based on your use case
                httpMethod: POST
                uri: !Sub
                  - arn:aws:apigateway:${AWS::Region}:lambda:path/2015-03-31/functions/${FunctionArn}/invocations
                  - FunctionArn: !GetAtt MyFunction.Arn

  MyFunction:
    Type: AWS::Serverless::Function
    Properties:
      Handler: index.handler
      Runtime: nodejs14.x
      Events:
        MyApi:
          Type: Api
          Properties:
            Path: /my-resource
            Method: GET  # Specify the correct HTTP method here

[denisbog]

I would rather have a way to define the response model directly at AWS::Serverless::Function level, similar to what we have now for request model https://docs.aws.amazon.com/serverless-application-model/latest/developerguide/sam-property-function-requestmodel.html