'terraform import digitalocean_kubernetes_cluster' imports wrong node pool

[mstyne]

Bug Report

---

Describe the bug

When importing an existing Kubernetes cluster using terraform import digitalocean_kubernetes_cluster, the incorrect node pool ID is imported.

Affected Resource(s)

• digitalocean_kubernetes_cluster
• digitalocean_kubernetes_node_pool

Expected Behavior

With the following configuration:

% doctl kubernetes cluster node-pool list 80654840-... --format ID,Tags
ID                                      Tags
6f41bd3b-...    k8s,k8s:80654840-...,k8s:worker,terraform:default-node-pool
152b6618-...    k8s,k8s:80654840-...,k8s:worker

I expect node pool 6f41bd3b-..., tagged as terraform:default-node-pool to be imported when executing

terraform import digitalocean_kubernetes_cluster.doks_hosted_cluster 80654840-...

Actual Behavior

When executing terraform import, the other, non-tagged node pool is imported.

digitalocean_kubernetes_cluster.doks_hosted_cluster: Importing from ID "80654840-..."...
digitalocean_kubernetes_cluster.doks_hosted_cluster: Import prepared!
  Prepared digitalocean_kubernetes_cluster for import
  Prepared digitalocean_kubernetes_node_pool for import
digitalocean_kubernetes_node_pool.doks_hosted_cluster: Refreshing state... [id=152b6618-...]
digitalocean_kubernetes_cluster.doks_hosted_cluster: Refreshing state... [id=80654840-...]

Import successful!

The resources that were imported are shown above. These resources are now in
your Terraform state and will henceforth be managed by Terraform.

Steps to Reproduce

See "Actual Behavior". Terraform Configuration Files main.tf:

terraform {
  required_providers {
    digitalocean = {
      source = "digitalocean/digitalocean"
      version = "~> 2.0"
    }
  }
}

variable "do_token" {}
variable "doks_cluster_name" {}
variable "doks_cluster_region" {}
variable "doks_cluster_version" {}
variable "doks_cluster_tags" {}

variable "doks_primary_node_pool_name" {}
variable "doks_primary_node_pool_size" {}
variable "doks_primary_node_pool_min" {}
variable "doks_primary_node_pool_max" {}
variable "doks_primary_node_pool_count" {}

variable "doks_secondary_node_pool_name" {}
variable "doks_secondary_node_pool_size" {}
variable "doks_secondary_node_pool_min" {}
variable "doks_secondary_node_pool_max" {}
variable "doks_secondary_node_pool_count" {}

provider "digitalocean" {
  token = var.do_token
}

resource "digitalocean_kubernetes_cluster" "doks_hosted_cluster" {
  name   = var.doks_cluster_name
  region = var.doks_cluster_region
  version = "1.21.14-do.0"
  auto_upgrade = true
  tags = [ "development" ]

  node_pool {
    name       = var.doks_primary_node_pool_name
    size       = var.doks_primary_node_pool_size
    max_nodes  = var.doks_primary_node_pool_max
    min_nodes  = var.doks_primary_node_pool_min
    node_count = var.doks_primary_node_pool_count
  }

}

resource "digitalocean_kubernetes_node_pool" "doks_secondary_node_pool" {
  cluster_id = digitalocean_kubernetes_cluster.doks_hosted_cluster.id

  name       = var.doks_secondary_node_pool_name
  size       = var.doks_secondary_node_pool_size
  node_count = var.doks_secondary_node_pool_count

}

Expected behavior

% terraform -v       
Terraform v1.2.5
on darwin_amd64
+ provider registry.terraform.io/digitalocean/digitalocean v2.21.0

The node pool tagged as terraform:default-node-pool should be imported when executing terraform import digitalocean_kubernetes_cluster.

Debug Output I'm concerned about the level of detail in the debug output; I have the gist and can email the link if necessary.

Panic Output None

Additional context

Changing the terraform:default-node-pool tag to the secondary node pool results in the "correct" node pool being imported along with the cluster.

Important Factoids None

References None