beta-autopilot-private-cluster - v39.0 - Drift on cluster_autoscaling.default_compute_class_enabled

[clevandowski]

TL;DR

I enabled manually via gcloud cli the ability to define a default computeClass on my gke autopilot cluster:

gcloud container clusters update <mycluster> --location=<mylocation> --enable-default-compute-class

Since I upgraded the module beta-autopilot-private-cluster to v39.0, terraform detects a change on cluster_autoscaling.default_compute_class_enabled, and the module is not able to manage this case. So I can't reconcile the terraform config with the real cluster, and I need to have a default computeClass.

Expected behavior

Enable management of cluster_autoscaling.default_compute_class_enabled in the module beta-autopilot-private-cluster (and also in all autopilot modules)

Observed behavior

$ terraform plan
...
Terraform used the selected providers to generate the following execution plan. Resource actions are indicated with the following symbols:
  ~ update in-place

Terraform will perform the following actions:

  # module.gke.google_container_cluster.primary will be updated in-place
  ~ resource "google_container_cluster" "primary" {
        id                                       = "projects/************/locations/europe-west9/clusters/********"
        name                                     = "********"
        # (37 unchanged attributes hidden)

      ~ cluster_autoscaling {
          - default_compute_class_enabled = true -> null
            # (3 unchanged attributes hidden)

            # (5 unchanged blocks hidden)
        }

        # (43 unchanged blocks hidden)
    }

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

Terraform Configuration

module "gke" {
  source  = "terraform-google-modules/kubernetes-engine/google//modules/beta-autopilot-private-cluster"
  version = "~> 39.0"

  project_id        = var.project_id
  name              = var.cluster_name
  regional          = true
  region            = var.region
  network           = google_compute_network.main.name
  subnetwork        = google_compute_subnetwork.main.name
  ip_range_pods     = google_compute_subnetwork.main.secondary_ip_range[0].range_name
  ip_range_services = google_compute_subnetwork.main.secondary_ip_range[1].range_name
  release_channel                 = "REGULAR"
  enable_private_endpoint         = false
  enable_private_nodes            = true
  enable_vertical_pod_autoscaling = true
  network_tags                    = [var.cluster_name]
  node_pools_cgroup_mode          = "CGROUP_MODE_V2"
  deletion_protection             = false
  enable_l4_ilb_subsetting        = true
  stateful_ha                     = false
  gke_backup_agent_config         = false
  ray_operator_config = {
    enabled            = true
    logging_enabled    = true
    monitoring_enabled = true
  }

  create_service_account = true
  grant_registry_access = true
}

Terraform Version

Terraform v1.13.2
on linux_amd64
+ provider registry.terraform.io/hashicorp/aws v6.5.0
+ provider registry.terraform.io/hashicorp/google v6.49.3
+ provider registry.terraform.io/hashicorp/google-beta v6.45.0
+ provider registry.terraform.io/hashicorp/kubernetes v2.38.0
+ provider registry.terraform.io/hashicorp/random v3.7.2

Terraform Provider Versions

Providers required by configuration:
.
├── provider[registry.terraform.io/hashicorp/google] ~> 6.49.3
├── provider[registry.terraform.io/hashicorp/aws]
├── provider[registry.terraform.io/hashicorp/kubernetes]
└── module.gke
    ├── provider[registry.terraform.io/hashicorp/google] >= 6.42.0, < 8.0.0
    ├── provider[registry.terraform.io/hashicorp/google-beta] >= 6.42.0, < 8.0.0
    ├── provider[registry.terraform.io/hashicorp/kubernetes] ~> 2.10
    └── provider[registry.terraform.io/hashicorp/random] >= 2.1.0

Providers required by state:

    provider[registry.terraform.io/hashicorp/aws]

    provider[registry.terraform.io/hashicorp/google]

    provider[registry.terraform.io/hashicorp/google-beta]

    provider[registry.terraform.io/hashicorp/random]

Additional information

No response

[DrFaust92]

I think it requires this as this is an upstream issue causing the diff

[clevandowski]

Yes it's exactly what I need