error when destroying and recreating google_compute_per_instance_config/google_compute_region_per_instance_config

[adrian-gierakowski]

When a google_compute_per_instance_config is created, both a vm and the instance config linked to it are created.

However when google_compute_per_instance_config is destroyed, the is vm remains untouched

This leads to errors when a deleted google_compute_per_instance_config is subsequently recreated, since a vm with given name already exists as part of the mig.

google_compute_per_instance_config should (on destruction( either delete the vm instance it initially creates (via instanceGroupManagers/deleteInstances), or provide a flag to toggle this behaviour.

I'm happy to submit a PR.

Note: in the example below google_compute_region_instance_group_manager.target_size is set to null for the reason provided in the comment.

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 --version

Terraform v0.13.6
+ provider registry.terraform.io/hashicorp/google v3.66.1

Your version of Terraform is out of date! The latest version

Affected Resource(s)

• google_compute_region_per_instance_config
• google_compute_per_instance_config

Terraform Configuration Files

provider "google" {
  project    = "my-test-project-ar90s"
  region     = "us-central1"
  # Version set to prevent automatic major version upgrades.
  version    = "~> 3.66.1"
}

variable "size" {
  description = "Number of instances to create"
}

resource "google_compute_network" "vpc_network" {
  name = "vpc-network"
}

data "google_compute_image" "my_image" {
  family  = "debian-9"
  project = "debian-cloud"
}

resource "google_compute_instance_template" "igm-basic" {
  name           = "my-template"
  machine_type   = "e2-small"
  can_ip_forward = false
  tags           = ["foo", "bar"]

  disk {
    source_image = data.google_compute_image.my_image.self_link
    auto_delete  = true
    boot         = true
  }

  network_interface {
    network = google_compute_network.vpc_network.name
  }

  service_account {
    scopes = ["userinfo-email", "compute-ro", "storage-ro"]
  }
}

resource "google_compute_region_instance_group_manager" "rigm" {
  description = "Terraform test instance group manager"
  name        = "my-rigm"

  version {
    name              = "prod"
    instance_template = google_compute_instance_template.igm-basic.self_link
  }

  base_instance_name = "rigm"
  region             = google_compute_instance_template.igm-basic.region
  # Needs to be set to null, and not var.size, since otherwise "random" instance
  # will be deleted when target_size decreases, while we want metadata.index of
  # all existing instances to be contiguous (with values between: 0 and 
  # var.size - 1)
  target_size        = null

  update_policy {
    instance_redistribution_type = "NONE"
    type                         = "OPPORTUNISTIC"
    minimal_action               = "REPLACE"
    max_surge_fixed              = 0
    max_surge_percent            = null
    max_unavailable_fixed        = 3
    max_unavailable_percent      = null
    min_ready_sec                = 30
  }
}

resource "google_compute_region_per_instance_config" "configs" {
  count = var.size

  region = google_compute_region_instance_group_manager.rigm.region
  region_instance_group_manager = google_compute_region_instance_group_manager.rigm.name

  name = "my-rigm-instance-${count.index}"

  preserved_state {
    metadata = {
      index = count.index
    }
  }

  remove_instance_state_on_destroy = true
}

Debug Output

Panic Output

Expected Behavior

Actual Behavior

Steps to Reproduce

after running:

terraform init
terraform apply -var size=2
terraform apply -var size=1
terraform apply -var size=2

I get the following error:

...
google_compute_region_per_instance_config.configs[1]: Creating...
google_compute_region_per_instance_config.configs[1]: Still creating... [10s elapsed]

Error: Error waiting to create RegionPerInstanceConfig: Error waiting for Creating RegionPerInstanceConfig: Resource projects/my-test-project-ar90s/zones/us-central1-f/instances/my-rigm-instance-1 is already a member of projects/my-test-project-ar90s/regions/us-central1/instanceGroupManagers/my-rigm.


  on main.tf line 71, in resource "google_compute_region_per_instance_config" "configs":
  71: resource "google_compute_region_per_instance_config" "configs" {

Important Factoids

References

• #0000

[adrian-gierakowski]

note that this can be worked around by calling gcloud compute instance-groups managed delete-instances ... from within local-exec provisioner executed on destroy of google_compute_region_per_instance_config, but the solution should really be incorporated natively into the resource itself

[github-actions]

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 have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.