AWS SAM: Stage already exists error during deployment

Question

I'm trying to deploy a simple API Gateway integrated with a Lambda function. I'm using GitLab CICD to deploy the resource but even after trying out so many fixes, I can't seem to get it to deploy successfully. Below is the YAML template I'm using.

Resources:
  MyFunction: 
    Type: AWS::Serverless::Function 
      FunctionName : MyFunctionName
      CodeUri: MyFunctionName/
      Handler: app.lambda_handler
      Role: MyRole
      Tags: 
        env : dev
      Architectures:
        - x86_64
      Events:
        MyAPI:
          Type: Api
          Properties:
            Path: /dev
            Method: POST
            RestApiId: 
              Ref: MyAPI
      VpcConfig:
        SecurityGroupIds: <sec grp id>
        SubnetIds: <subnet id>

  MyAPI: 
    Type: AWS::Serverless::Api
    Properties:
      DefinitionBody: 
        swagger: 2.0
        info:
          title: MyAPIName
        host: "xxxxx.execute-api.ap-south-1.amazonaws.com"
        schemes:
          - https        
        basePath: /dev
        paths:
          /:
            post:
              responses: {}
              x-amazon-apigateway-integration:
                type: "aws_proxy"
                httpMethod: "POST"
                uri: "arn:aws:apigateway:ap-south-1:lambda:path/2015-03-31/functions/<MyFunction ARN>/invocations"
                responses:
                  default:
                    statusCode: "200"
                passthroughBehavior: "when_no_match"
                contentHandling: "CONVERT_TO_TEXT"      
        x-amazon-apigateway-policy:
          Version: '2012-10-17'
          Statement:
          - Effect: Deny
            Principal: "*"
            Action: execute-api:Invoke
            Resource: "*"
            Condition:
              StringNotEquals:
                aws:sourceVpce:
                - vpce-xxxxxx
          - Effect: Allow
            Principal: "*"
            Action: execute-api:Invoke
            Resource: "*"
      EndpointConfiguration:
        Type: PRIVATE
      StageName: dev

  MyAPIDeployment:  
    Type: AWS::ApiGateway::Deployment
    Properties: 
      RestApiId: 
        Ref: MyAPI
      StageName: dev

When I try to deploy the above script, the error I am receiving is: (using GitLab CI)

dev already exists in stack arn:aws:cloudformation:ap-south-1:xxxx:stack/stackname/xxxx

I checked the API GW on console and found a stage by the name Stage(from an older deployment). But no "dev". Thought that this was causing the issue, I deleted the stage and tried deploying again. But it is still giving me the same error. I have been stuck with this issue for few days now and any kind of help would be appreciated.

ZabielskiGabriel · Accepted Answer · 2022-09-19T13:23:11.283

1

Edit

I didn't correctly associate this issue - this is a one of possible ways: CloudFormation doesn't deploy to API gateway stages on update (it's quite complex with pure CFN, but CDK provides some constructs for easy implementation of custom resources, if it is possible I recommend to migrate your code to CDK)

A second way, I see here is to create a lambda function, that will wait until the update is complete and then request deploy API. This lambda should be invoked by the Cloudformation template.

edited Sep 19 '22 at 13:23

answered Sep 16 '22 at 20:10

ZabielskiGabriel

551
3
12

Is there a method to remove the conflict without having to delete the stack? Since the issue is with the API Gateway, deleting the stack and redeploying would change the url of the API, which is something we want to avoid. Maybe a way to delete the deployments of the stack without having to delete all the resources in it? – Gattaca Sep 19 '22 at 05:58
update the stack? – ZabielskiGabriel Sep 19 '22 at 10:46
Since the stack is already present in the workspace, won't a new deployment be essentially considered as an updation by cloudformation? – Gattaca Sep 19 '22 at 11:53
ohhh, ok now I understand... Cloudformation is so stupid... I will update my answare – ZabielskiGabriel Sep 19 '22 at 13:12
1

I came across the question you linked in your answer before. I did try few of the fixes but wasn't very helpful. I'll go ahead and try with creating a timestamp element attached to the deployment resource. Tested it out from another account, it seems to be working. – Gattaca Sep 21 '22 at 07:26

AWS SAM: Stage already exists error during deployment

1 Answers1

Edit