Terraform-provider-google: Error 400 creating google_container_cluster on v3.29.0

[Eshanel]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request.
• Please do not leave _+1_ or _me too_ comments, they generate extra noise for issue followers and do not help prioritize the request.
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.
• If an issue is assigned to the modular-magician user, it is either in the process of being autogenerated, or is planned to be autogenerated soon. If an issue is assigned to a user, that user is claiming responsibility for the issue. If an issue is assigned to hashibot, a community member has claimed the issue already.

Terraform Version

Terraform v0.12.25 - Terraform Cloud

• provider.google v3.29.0

Affected Resource(s)

• google_container_cluster

Terraform Configuration Files

# Copy-paste your Terraform configurations here.
#
# For large Terraform configs, please use a service like Dropbox and share a link to the ZIP file.
# For security, you can also encrypt the files using our GPG public key:
#    https://www.hashicorp.com/security
#
# If reproducing the bug involves modifying the config file (e.g., apply a config,
# change a value, apply the config again, see the bug), then please include both:
# * the version of the config before the change, and
# * the version of the config after the change.
resource "google_container_cluster" "cluster" {
  provider    = google-beta
  name        = var.CLUSTER_NAME
  location    = var.CLUSTER_ZONE
  description = var.CLUSTER_DESCRIPTION

  // https://www.terraform.io/docs/providers/google/r/container_cluster.html
  // We can't create a cluster with no node pool defined, but we want to only use
  // separately managed node pools. So we create the smallest possible default
  // node pool and immediately delete it.
  remove_default_node_pool = true
  initial_node_count       = 1

  enable_binary_authorization = false
  enable_kubernetes_alpha     = false
  enable_legacy_abac          = false
  enable_shielded_nodes       = true
  enable_intranode_visibility = true
  default_max_pods_per_node   = 110
  logging_service             = "logging.googleapis.com/kubernetes"
  monitoring_service          = "monitoring.googleapis.com/kubernetes"
  network                     = google_compute_network.cluster_vpc.self_link
  subnetwork                  = google_compute_subnetwork.cluster_vpc_subnetwork.self_link
  resource_labels             = var.GOOGLE_LABELS

  addons_config {
    horizontal_pod_autoscaling {
      disabled = false
    }
    http_load_balancing {
      disabled = false
    }
    network_policy_config {
      disabled = false
    }
  }
  cluster_autoscaling {
    enabled = false
  }
  database_encryption {
    state    = "ENCRYPTED"
    key_name = google_kms_crypto_key.cluster_kms_key_etcd.self_link
  }
  ip_allocation_policy {
    cluster_ipv4_cidr_block  = ""
    services_ipv4_cidr_block = ""
  }
  maintenance_policy {
    daily_maintenance_window {
      start_time = "11:00"
    }
  }
  master_auth {
    client_certificate_config {
      issue_client_certificate = false
    }
  }
  master_authorized_networks_config {
    dynamic cidr_blocks {
      for_each = var.CLUSTER_MASTER_AUTHORIZED_IPS
      content {
        display_name = cidr_blocks.key
        cidr_block   = cidr_blocks.value
      }
    }
  }
  network_policy {
    enabled = true
  }
  private_cluster_config {
    enable_private_nodes    = true
    enable_private_endpoint = false
    master_ipv4_cidr_block  = "172.16.0.0/28"
  }
  release_channel {
    channel = "REGULAR"
  }
}

Debug Output

Panic Output

Expected Behavior

The cluster must be created.

Actual Behavior

The plan was successful, but the apply got an error from the Google APIs

Error: googleapi: Error 400: DefaultMaxPodsConstraint can only be used if IpAllocationPolicy.UseIpAliases is true., badRequest

  on .terraform/modules/k8s/main.tf line 81, in resource "google_container_cluster" "cluster":
  81: resource "google_container_cluster" "cluster" {

Steps to Reproduce

1. terraform apply

Important Factoids

The plan is run from Terraform Cloud.
Also, we are running this config every morning, with success until today.

References

[jahernandezmartinez13]

With the provider.google v3.29 was failing for me as well.
I had to downgrade to terraform-provider-google_v3.14.0_x5 to make it work again.

[Eshanel]

I found a successful workaround with the beta argument networking_mode set to VPC_NATIVE

I wonder if the problem is the default blank cluster_ipv4_cidr_block and services_ipv4_cidr_block in ip_allocation_policy block which doesn't inferred the VPC_NATIVE mode for the cluster ?

FYI: My configuration now looks like this:

resource "google_container_cluster" "cluster" {
  provider    = google-beta
  name        = var.CLUSTER_NAME
  location    = var.CLUSTER_ZONE
  description = var.CLUSTER_DESCRIPTION

  // https://www.terraform.io/docs/providers/google/r/container_cluster.html
  // We can't create a cluster with no node pool defined, but we want to only use
  // separately managed node pools. So we create the smallest possible default
  // node pool and immediately delete it.
  remove_default_node_pool = true
  initial_node_count       = 1

  enable_binary_authorization = false
  enable_kubernetes_alpha     = false
  enable_legacy_abac          = false
  enable_shielded_nodes       = true
  enable_intranode_visibility = true
  default_max_pods_per_node   = 110
  logging_service             = "logging.googleapis.com/kubernetes"
  monitoring_service          = "monitoring.googleapis.com/kubernetes"
  networking_mode             = "VPC_NATIVE" // Added to avoid cluster creation error
  network                     = google_compute_network.cluster_vpc.self_link
  subnetwork                  = google_compute_subnetwork.cluster_vpc_subnetwork.self_link
  resource_labels             = var.GOOGLE_LABELS

  addons_config {
    horizontal_pod_autoscaling {
      disabled = false
    }
    http_load_balancing {
      disabled = false
    }
    network_policy_config {
      disabled = false
    }
  }
  cluster_autoscaling {
    enabled = false
  }
  database_encryption {
    state    = "ENCRYPTED"
    key_name = google_kms_crypto_key.cluster_kms_key_etcd.self_link
  }
  ip_allocation_policy {
    cluster_ipv4_cidr_block  = ""
    services_ipv4_cidr_block = ""
  }
  maintenance_policy {
    daily_maintenance_window {
      start_time = "11:00"
    }
  }
  master_auth {
    client_certificate_config {
      issue_client_certificate = false
    }
  }
  master_authorized_networks_config {
    dynamic cidr_blocks {
      for_each = var.CLUSTER_MASTER_AUTHORIZED_IPS
      content {
        display_name = cidr_blocks.key
        cidr_block   = cidr_blocks.value
      }
    }
  }
  network_policy {
    enabled = true
  }
  private_cluster_config {
    enable_private_nodes    = true
    enable_private_endpoint = false
    master_ipv4_cidr_block  = "172.16.0.0/28"
  }
  release_channel {
    channel = "REGULAR"
  }
}

[milentzvetkov]

Confirmed that the workaround is working just now, I spend a good 2-3 hours of wondering why just 2 days ago everything was finе....
Thanks @Eshanel you owe you one for this.
Latest working provider 3.28 as I can see in my cloudbuild history...

[M3t0r]

As @jahernandezmartinez13 pointed out, just using a previous version works. But since it took me a while to find the syntax for specifying versions, I'm posting what helped me here:

provider "google" {
  version = "~> 3.18, != 3.29.0"
  project = local.gcp_project
  region  = local.region
}

provider "google-beta" {
  version = "~> 3.18, != 3.29.0"
  project = local.gcp_project
  region  = local.region
}

[farmdawgnation]

Hi folks,

I'm just here to comment that I ran into a related issue with a different error message. While attempting to create a private cluster, I was receiving this error message:

Alias IP addresses are required for private cluster, please make sure you enable alias IPs when creating a cluster.

My ip_allocation_policy is using cluster_secondary_range_name and services_secondary_range_name to call out specific ranges in the network to pull addresses from. It appears as though I had to specify networking_mode = "VPC_NATIVE" explicitly to get things working.

[megan07]

Sorry for the mix-up with this! Yes, networking_mode = "VPC_NATIVE" is what we want to use going forward. It was the default when ip_allocation_policy was set before, however, we made this change in the most recent release. Unfortunately, we missed this backward-compatability in implementation. I have created a PR to address this change. Thanks for your patience!

[farmdawgnation]

No worries, thanks for the quick reply

[hashibot[bot]]

I'm going to lock this issue because it has been closed for _30 days_ ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 [email protected]. Thanks!