drift after apply of redshift_schema.owner

[m-ronchi]

I added this to our terraform project:

data "aws_redshift_cluster" "analytics" {
  cluster_identifier = "analytics"
}

provider "redshift" {
  host     = data.aws_redshift_cluster.analytics.endpoint
  username = "terraform"
  database = "analytics"
  temporary_credentials {
    cluster_identifier = data.aws_redshift_cluster.analytics.cluster_identifier
    auto_create_user   = true
  }
}

resource "redshift_schema" "spectrum" {
  for_each = toset(["redacted"])

  name = each.key

  external_schema {
    database_name = each.key
    data_catalog_source {
      iam_role_arns = [data.aws_iam_role.redshift.arn]
    }
  }
}

data "redshift_group" "analysts" {
  name = "analysts"
}

resource "redshift_grant" "group" {
  for_each    = redshift_schema.spectrum
  group       = data.redshift_group.analysts.name
  schema      = each.key
  object_type = "schema"
  privileges  = ["usage"]
}

if I run a plan right after apply, I expect no changes but instead get

Terraform used the selected providers to generate the following execution
plan. Resource actions are indicated with the following symbols:
  ~ update in-place
Terraform will perform the following actions:
  # redshift_schema.spectrum["redacted"] will be updated in-place
  ~ resource "redshift_schema" "spectrum" {
        id    = "11396018"
        name  = "redacted"
      ~ owner = "IAM:terraform" -> "iam:terraform"
        # (1 unchanged attribute hidden)
        # (1 unchanged block hidden)
    }
Plan: 0 to add, 1 to change, 0 to destroy.