Aws api gateway stage and deployment: log group skip_destroy=true preventing application?

Terraform Providers
1

I’m a terraform beginner, trying to deploy an api gateway with integration to a vpc link.

When I terraform apply on my project, neither the aws_apigatewayv2_stage resource nor the aws_apigatewayv2_deployment resource that I have defined in my files actually result in the corresponding stage and deployment showing up on AWS.

Here are the resource blocks I’ve written to define those resources

# api gateway stage
resource "aws_apigatewayv2_stage" "this" {
  api_id      = aws_apigatewayv2_api.this.id
  name        = "$default" # public url for deployments to this stage will not have any version or stage qualifier (as you might see like 'api.example.com/v2/route')
  auto_deploy = true
  default_route_settings {
    detailed_metrics_enabled = true
  }
  deployment_id = aws_apigatewayv2_deployment.this.id
  access_log_settings {
    destination_arn = aws_cloudwatch_log_group.api_gateway_access_lg.arn
    format = jsonencode({
      "requestId"      = "$context.requestId"
      "ip"             = "$context.identity.sourceIp"
      "requestTime"    = "$context.requestTime"
      "httpMethod"     = "$context.httpMethod"
      "routeKey"       = "$context.routeKey"
      "status"         = "$context.status"
      "responseLength" = "$context.responseLength"
    })
  }
}
# deployment
resource "aws_apigatewayv2_deployment" "this" {
  api_id      = aws_apigatewayv2_api.this.id
  
  triggers = {
    redeployment = sha1(join(",", tolist([
      jsonencode(aws_apigatewayv2_integration.this),
      jsonencode(aws_apigatewayv2_route.api_makedataset_post_route),
    ])))
  }

  lifecycle {
    create_before_destroy = true
  }
}

I’m omitting all the other resources to try to isolate the parts of my config (e.g. aws_apigatewayv2_api) that might be failing.

I suspect that this failure may have to do with the aws_cloudwatch_log_group resource on which my stage depends:

# api log group
resource "aws_cloudwatch_log_group" "api_gateway_access_lg" {
  name              = "apgtwy/blw-data-api-server"
  skip_destroy      = true
  retention_in_days = 30
}

When I run terraform apply on my project, I get the following error message:

Error: creating CloudWatch Logs Log Group (apgtwy/blw-data-api-server): operation error CloudWatch Logs: CreateLogGroup, https response error StatusCode: 400, RequestID: e6a3d0bb-b39a-453a-a5a3-1411cea5a986, ResourceAlreadyExistsException: The specified log group already exists

Of course, I do not want terraform to overwrite that log group if it exists, because I don’t want the old logs destroyed. (hence I set skip_destroy = true in the log_group resource. So I’m interpreting this error message (probably incorrectly) as a signal that things are working as I intend.

But of course, they’re not working as I intend, because my stage and deployment do not get built.

Is it possible that skip_destroy=true or something else about the log group resource is blocking the deployment of my stage and deployment resources? There are no other error messages showing up when I do terraform apply.