Imported database replica attempting to resize to current configuration

[cphalpin]

Bug Report

---

Describe the bug

I've imported an existing database replica into Terraform Cloud. New run plans now say they will add a size of "db-s-1vcpu-2gb", which I believe is the current size of the replica, and when I try to apply it I get the error

Error: Error resizing database replica: PUT https://api.digitalocean.com/v2/databases/.../resize: 422 (request "...") cannot scale to same configuration

Affected Resource(s)

digitalocean_database_replica

Expected Behavior

I expect no change to be needed here since the replica should already be the correct size.

Actual Behavior

Terraform is trying to call the resize API to set the replica to its current size.

Steps to Reproduce

1. Create a new database replica through the DO UI.
2. Add matching configuration to terraform
3. Import existing replica with terraform import
4. Try to make any kind of changes with terraform apply

Terraform Configuration Files

config.tf

locals {
  region = "lon1"

  environments = [
    "dev",
    "staging",
    "prod",
  ]

  database_server_size = "db-s-1vcpu-2gb"
  database_server_has_standby_node_default = false

  database_server_has_standby_node_overrides = {
  }

  database_servers = {
    for environment in local.default_environments : environment => {
      size = local.database_server_size,
      has_standby_node = lookup(local.database_server_has_standby_node_overrides, environment, local.database_server_has_standby_node_default),
    }
  }
}

databases.tf

locals {
  sql_modes = [
    "ERROR_FOR_DIVISION_BY_ZERO",
    "IGNORE_SPACE",
    "NO_ENGINE_SUBSTITUTION",
    "NO_ZERO_DATE",
    "NO_ZERO_IN_DATE",
    "ONLY_FULL_GROUP_BY",
    "PIPES_AS_CONCAT",
    "REAL_AS_FLOAT",
    "STRICT_ALL_TABLES",
  ]
}

resource "digitalocean_database_cluster" "database_servers" {
  for_each = local.database_servers

  name = each.key
  engine = "mysql"
  version = "8"
  size = each.value.size
  region = local.region
  node_count = each.value.has_standby_node ? 2 : 1
  sql_mode = join(",", local.sql_modes)
  tags = [
    "env:${each.key}",
    "client:ubs",
  ]
}

resource "digitalocean_database_replica" "production_database_replica" {
  cluster_id = digitalocean_database_cluster.database_servers["prod"].id
  name = "prod-read-only-replica-2"
  size = digitalocean_database_cluster.database_servers["prod"].size
  region = digitalocean_database_cluster.database_servers["prod"].region
  tags = digitalocean_database_cluster.database_servers["prod"].tags
}

Terraform version

1.9.3 in Terraform Cloud

My local install is currently 1.9.2 but that's not what I'm using to plan/apply.

[andrewsomething]

Hi @cphalpin,

Thank you for raising this issue. We understand the issue and have a fix in flight. It will require a change landing in https://github.com/digitalocean/godo first. While waiting for that, you can temporarily work around the issue by setting a lifecycle block with:

  lifecycle {
    ignore_changes = [
        size
    ]
  }