Can't delete an instance group that's in use by a backend service

[danawillow]

Terraform Version

Terraform v0.11.7
provider.google (unversioned)

Affected Resource(s)

• google_compute_instance_group
• google_compute_backend_service
  (probably also google_compute_region_backend_service)

Terraform Configuration Files

resource "google_compute_instance" "ig_instance" {
  name           = "b-112171976"
  machine_type   = "n1-standard-1"
  zone           = "us-central1-c"

  boot_disk {
    initialize_params {
      image = "debian-cloud/debian-9"
    }
  }

  network_interface {
    network = "default"
  }
}

resource "google_compute_instance_group" "instance_group" {
  name      = "b-112171976"
  zone      = "us-central1-c"
  instances = ["${google_compute_instance.ig_instance.self_link}"]

  named_port {
    name = "https"
    port = "8443"
  }
}

resource "google_compute_backend_service" "default_backend" {
  name      = "b-112171976"
  port_name = "https"
  protocol  = "HTTPS"

  backend {
    group = "${google_compute_instance_group.instance_group.self_link}"
  }

  health_checks = [
    "${google_compute_https_health_check.healthcheck.self_link}",
  ]
}

resource "google_compute_https_health_check" "healthcheck" {
  name         = "b-112171976"
  request_path = "/health_check"
}

Debug Output

https://gist.github.com/danawillow/6962cbf60fdd78691d99fd610e5dbaf5

Panic Output

N/A

Expected Behavior

The instance group was deleted

Actual Behavior

google_compute_instance_group.instance_group: Error deleting InstanceGroup: googleapi: Error 400: The instance_group resource 'projects/graphite-test-danahoffman-tf/zones/us-central1-c/instanceGroups/b-112171976' is already being used by 'projects/graphite-test-danahoffman-tf/global/backendServices/b-112171976', resourceInUseByAnotherResource

Steps to Reproduce

1. terraform apply
2. Change the name of the instance group
3. terraform apply

Important Factoids

N/A

References

N/A

I think we probably want to handle this in a similar way to how we detach disks from instances before deleting them.

[Chupaka]

I thought, renaming instance group should be done via

  lifecycle {
    create_before_destroy = true
  }

That's what I actually did a day ago 🤣

[ghost]

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!