[Bug]: aws_sagemaker_domain breaks default_space/user settings args everytime

[barra51]

Terraform Core Version

1.9.2

AWS Provider Version

5.73.0

Affected Resource(s)

• aws_sagemaker_domain

Expected Behavior

1. Defining a simple default_space_settings or default_user_settings using the aws_sagemaker_domain for apps like JupyterLab should ignore optional values and not force them on every terraform run/apply. This causes to use not desired values/omit the default ones.

Actual Behavior

Applying the following code under aws_sagemaker_domain:

      jupyter_lab_app_settings {
        code_repository {
          repository_url = local.github_repo
      }
          default_resource_spec {
            #instance_type                 = "ml.g4dn.xlarge"
            lifecycle_config_arn          = aws_sagemaker_studio_lifecycle_config.startup.arn
          }
            lifecycle_config_arns = [
            aws_sagemaker_studio_lifecycle_config.startup.arn
            ]
        }

And subsequent runs after applied produces terraform to change:

 ~ space_settings {    
  ~  jupyter_lab_app_settings { 
    ~   default_resource_spec { 
      ~    sagemaker_image_arn :    "arn:aws:sagemaker:us-east-1:885854791233:image/sagemaker-distribution-cpu" -> ""
        ~  sagemaker_image_version_alias: "2.0.0"  -> ""

Relevant Error/Panic Output Snippet

Error: updating SageMaker Space: operation error SageMaker: UpdateSpace, https response error StatusCode: 400, RequestID: b09d1b8f-33ae-4bc3-aa83-e8aeee0a65ae, api error ValidationException: Updating AppType for Space is not supported.

Terraform Configuration Files

resource "aws_sagemaker_domain" "this" {
  domain_name             = "mydomain"
  auth_mode               = "IAM"
  vpc_id                  = data.aws_vpc.id
  app_network_access_type = "VpcOnly"
  subnet_ids              = local.subnets
  tags                    = local.tags
  kms_key_id              = aws_kms_key.sagemaker_kms_key.key_id

  domain_settings {
    security_group_ids = local.security_groups
    docker_settings {
      enable_docker_access = "ENABLED"
    }
  }

  default_space_settings {
    execution_role  = "arn:custom"
    security_groups = local.security_groups

    jupyter_lab_app_settings {
      code_repository {
        repository_url = local.github_repo
      }
      default_resource_spec {
        instance_type                 = "ml.g4dn.xlarge"
        lifecycle_config_arn          = aws_sagemaker_studio_lifecycle_config.startup.arn
        # If I set these two, then no errors on the plan. . .
        #sagemaker_image_arn           = null
        #sagemaker_image_version_alias = null
        }
          lifecycle_config_arns = [
            aws_sagemaker_studio_lifecycle_config.startup.arn
        ]
     }
}

Steps to Reproduce

• Run the code from above.
• Apply it.
• Run a new plan.
• You should get two args to be modified:

  ~ space_settings {    
  ~  jupyter_lab_app_settings { 
  ~   default_resource_spec { 
    ~    sagemaker_image_arn :    "arn:aws:sagemaker:us-east-1:885854791233:image/sagemaker-distribution-cpu" -> ""
      ~  sagemaker_image_version_alias: "2.0.0"  -> ""

Debug Output

No response

Panic Output

No response

Important Factoids

No response

References

No response

Would you like to implement a fix?

None

[github-actions[bot]]

Community Note

Voting for Prioritization

• Please vote on this issue by adding a 👍 reaction to the original post to help the community and maintainers prioritize this request.
• Please see our prioritization guide for information on how we prioritize.
• Please do not leave "+1" or other comments that do not add relevant new information or questions, they generate extra noise for issue followers and do not help prioritize the request.

Volunteering to Work on This Issue

• If you are interested in working on this issue, please leave a comment.
• If this would be your first contribution, please review the contribution guide.