Blocks of type "secondary_boot_disks" are not expected here

[alvicsam]

TL;DR

The change introduced in #1946 breaks tf plan for a simple GKE cluster.

Expected behavior

Simple GKE cluster after tf plan should not require secondary_boot_disks

Observed behavior

➜  base git:(as-tf-mod) ✗ tf plan
╷
│ Error: Unsupported block type
│
│   on .terraform/modules/gke/cluster.tf line 613, in resource "google_container_node_pool" "pools":
│  613:     dynamic "secondary_boot_disks" {
│
│ Blocks of type "secondary_boot_disks" are not expected here.

Terraform Configuration

# GKE

data "google_client_config" "default" {}

provider "kubernetes" {
  host                   = "https://${module.gke.endpoint}"
  token                  = data.google_client_config.default.access_token
  cluster_ca_certificate = base64decode(module.gke.ca_certificate)
}

module "gke" {
  source = "github.com/terraform-google-kubernetes-engine/terraform-google-kubernetes-engine?ref=e573cedee0a82517410f9d607ce69be7189124ca"

  project_id                = var.project_id
  name                      = "org-name-gke"
  regional                  = true
  region                    = var.region
  zones                     = ["europe-west1-b"]
  network                   = "org-name"
  subnetwork                = "org-name-europe-west1"
  ip_range_pods             = "org-name-pods-europe-west1"
  ip_range_services         = "org-name-services-europe-west1"
  create_service_account    = true
  grant_registry_access     = true
  default_max_pods_per_node = 20
  remove_default_node_pool  = true
  deletion_protection       = false
  datapath_provider         = "ADVANCED_DATAPATH" # dataplane v2

  node_pools = [
    {
      name              = "application-pool"
      machine_type      = "e2-standard-2"
      node_count        = 2
      autoscaling       = false
      local_ssd_count   = 0
      disk_size_gb      = 50
      disk_type         = "pd-standard"
      image_type        = "COS_CONTAINERD"
      auto_repair       = true
      auto_upgrade      = true
      preemptible       = false
      max_pods_per_node = 70
    },
    {
      name         = "application-pool2"
      machine_type = "n2-standard-8"
      min_count    = 0
      max_count    = 50
      autoscaling  = true
      local_ssd_count   = 0
      disk_size_gb      = 500
      disk_type         = "pd-ssd"
      image_type        = "COS_CONTAINERD"
      auto_repair       = true
      auto_upgrade      = true
      preemptible       = true
      max_pods_per_node = 20
    },
    {
      name         = "application-pool-ssd"
      machine_type = "n2-standard-8"
      min_count    = 0
      max_count    = 50
      autoscaling  = true
      local_ssd_ephemeral_storage_count = 2
      disk_size_gb                      = 500
      disk_type                         = "pd-ssd"
      image_type                        = "COS_CONTAINERD"
      auto_repair                       = true
      auto_upgrade                      = true
      preemptible                       = true
      max_pods_per_node                 = 20
      min_cpu_platform                  = "Intel Ice Lake"
      threads_per_core                  = 1
      secondary_boot_disk               = ""
    },
  ]

}

Terraform Version

Terraform v1.7.4

Additional information

cc @yorik

[yorik]

@alvicsam I think you have to upgrade hashicorp/google and/or hashicorp/google-beta providers to version at least 5.29.0.

[alvicsam]

@yorik yep, that helped, thank you!