redshift_grant fails on apply stage for external schemas

[svdimchenko]

I get the following error when try to create redshift grant for external schema:

│ Error: Provider produced inconsistent result after apply
│ 
│ When applying changes to redshift_grant.user, provider "provider[\"registry.terraform.io/brainly/redshift\"]" produced an unexpected new value: Root resource was present, but
│ now absent.
│ 
│ This is a bug in the provider, which should be reported in the provider's own issue tracker.

Terraform plan seems like following:

  # redshift_grant.service_accounts["id"] will be created
  + resource "redshift_grant" "service_accounts" {
      + id          = (known after apply)
      + object_type = "schema"
      + privileges  = [
          + "usage",
        ]
      + schema      = "<schema_name>"
      + user        = "<user_name>"
    }

At the same time grant for regular schema apply works properly.
Also if I run following SQL query, grant applies without any error:

GRANT USAGE ON SCHEMA <external_schema_name> TO <user>;

[dkravetz]

I am currently facing the same problem when trying to grant USAGE and CREATE to a user in the public schema.

This is the resource block that's failing:

resource "redshift_grant" "public_schema" {
  user        = "my_redshift_user"
  object_type = "schema"
  privileges  = ["usage", "create"]
  schema      = "public"
}

and this is the outcome of my apply:

   # module.redshift.module.redshift_config.redshift_grant.public_schema will be created
  + resource "redshift_grant" "public_schema" {
      + id          = (known after apply)
      + object_type = "schema"
      + privileges  = [
          + "create",
          + "usage",
        ]
      + schema      = "public"
      + user        = "my_redshift_user"
    }

Plan: 1 to add, 0 to change, 0 to destroy.

Do you want to perform these actions?
  OpenTofu will perform the actions described above.
  Only 'yes' will be accepted to approve.

  Enter a value: yes

module.redshift.module.redshift_config.redshift_grant.public_schema: Creating...
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [10s elapsed]
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [20s elapsed]
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [30s elapsed]
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [40s elapsed]
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [50s elapsed]
module.redshift.module.redshift_config.redshift_grant.public_schema: Still creating... [1m0s elapsed]
╷
│ Error: Provider produced inconsistent result after apply
│
│ When applying changes to module.redshift.module.redshift_config.redshift_grant.public_schema, provider
│ "provider[\"registry.opentofu.org/brainly/redshift\"]" produced an unexpected new value: Root object was present, but now absent.
│
│ This is a bug in the provider, which should be reported in the provider's own issue tracker.

I am running OpenTofu v1.8.4 and Redshift's provider v1.1.0. Any help would be much appreciated

[dkravetz]

Welp, turns out that in my case, another engineer accidentally dropped the public schema. When setting TF_LOG to DEBUG I could see the failing query, and running it through QueryEditor I got the useful message "schema public does not exist` 🤦