DyhnamoDB connector - Error: Provider produced inconsistent result after apply

[Daniele-RedCloud]

Describe the bug While updating a DynamoDB connector by switching to umpacked mode and empty list of tables terraform plan is successful but terraform apply fails with a provider error.

To Reproduce

resource "fivetran_connector" "dynamodb" {
  group_id = var.group_id

  service = "dynamodb"

  run_setup_tests = "true"

  destination_schema {
    name = var.schema_name
  }

  config {
    external_id        = var.group_id
    role_arn           = var.aws_iam_arn
    aws_region_code    = var.aws_region
    sync_mode          = "UseUnpackedModeOnly"
    packed_mode_tables = []
  }
}

Expected behavior Terraform apply to succeed.

Logs & Output

Error: Provider produced inconsistent result after apply
│
│ When applying changes to module.connector.fivetran_connector.dynamodb,
│ provider "provider[\"registry.terraform.io/fivetran/fivetran\"]" produced
│ an unexpected new value: .config.packed_mode_tables: was
│ cty.SetValEmpty(cty.String), but now null.
│
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker.

Plugin version:

• Terraform v1.5.7
• provider registry.terraform.io/fivetran/fivetran v1.1.21

Additional context

Note that if I instead use:

packed_mode_tables = null

terraform succeed but the connector fails to synchronise with the following error:

Rescheduling to recover quickly after failure Cannot invoke \"java.util.Set.contains(Object)\" because \"this.packedModeTables\" is null

[beevital]

Hey @Daniele-RedCloud! We had similar issue with empty strings already. Some connectors don't return empty values in response, so it causes inconsistency in state after apply. Will be fixed in next release.

[beevital]

https://github.com/fivetran/terraform-provider-fivetran/pull/300 fix PR raised

[beevital]

Fixed in v1.1.22