GCP : Not able to attach instance_group to the target pool as backend while creating loadbalancer

[ghost]

This issue was originally opened by @b-rajesh as hashicorp/terraform#25689. It was migrated here as a result of the provider split. The original body of the issue is below.

---

Terraform Version

0.12
...

Terraform Configuration Files

resource "google_compute_instance_template" "nginx-plus-gwy-template" {
  name        = "nginx-plus-gwy-template"
  tags = ["nginx-plus-api-gwy"]

  labels = {
    environment = "dev"
  }
  machine_type         = var.machine_type
  can_ip_forward       = false

  scheduling {
    automatic_restart   = true
    on_host_maintenance = "MIGRATE"
  }

  // Create a new boot disk from an image
  disk {
    source_image = "nginxplus-r22"
    auto_delete  = true
    boot         = true
  }
  network_interface {
    network = google_compute_network.vpc.name
    subnetwork = google_compute_subnetwork.subnet.name
    access_config {
    }
  }
  service_account {
    scopes = [
      "https://www.googleapis.com/auth/compute",
    ]
  }
  lifecycle {
    create_before_destroy = true
  }
}

resource "google_compute_instance_group_manager" "nginx-plus-gwy-group-manager" {
  name               = "nginx-plus-gwy-instance-group-manager"
  base_instance_name = "nginx-plus-api-gwy"
  zone               = var.zones
  target_size        = "3"
  version {
    instance_template = google_compute_instance_template.nginx-plus-gwy-template.id
  }
}


resource "google_compute_forwarding_rule" "gce-ext-lb-forwarding-rule" {
  depends_on            = [google_compute_subnetwork.subnet]
  name                  = "gce-external-lb"
  target                = google_compute_target_pool.default.self_link
  load_balancing_scheme = "EXTERNAL"
  network_tier           = "STANDARD"
  ip_address            = var.ip_address
  ip_protocol           = var.protocol
  port_range            = var.port_range
}

resource "google_compute_target_pool" "default" {
  name             = "ext-loadbalancer-nginx-plus"
  //instances        = google_compute_instance_group_manager.nginx-plus-gwy-group-manager.self_link
  //session_affinity = var.session_affinity
  health_checks = google_compute_http_health_check.default.*.name
}

resource "google_compute_http_health_check" "default" {
  name    = "nginx-plus-health-check"

  request_path        = var.health_check_path
  port                = var.health_check_port
  check_interval_sec  = var.health_check_interval
  healthy_threshold   = var.health_check_healthy_threshold
  unhealthy_threshold = var.health_check_unhealthy_threshold
  timeout_sec         = var.health_check_timeout
}

...

Debug Output

This above config creates external TCP load balancer successfully , however I am not able to assign the instance group as Backend through Terraform as there is no option I can see with google_compute_target_pool to achieve the same.

I have attached screenshot from the gcp console as I could achieve the same through the ui.

Expected behaviour

Is this a bug or terraform is still yet to get this feature added to it ? I need some guidance , please help....

[edwardmedia]

@b-rajesh could you post the debug log?

[b-rajesh]

@edwardmedia
As usual its my bad :( as I missed to refer the target pool in the google_compute_instance_group_manager . Now it works fine and load balancer can see those instances as backend.

On other note - my humble opinion too - it would have been much better if I am able to reference target_pool referencing the instance_group_manager . It would be better to do this way to understand the flow
LB --> TargetPool --> InstanceGroup
however it look like this now
LB --> TargetPool <-- InstanceGroup

Since TargetPool already have a way to reference to Instances directly then why can't we refer the instance group?

May be I am missing the big picture ? Any pointer to understand this would be much appreciated...

[edwardmedia]

@b-rajesh I am glad your issue has been resolved. To answer your question, could you review below doc to see if that makes sense? I am closing this issue now. Please feel free to reopen it if you still need to continue the discussion. Thank you

https://cloud.google.com/compute/docs/instance-groups/adding-an-instance-group-to-a-load-balancer#adding_a_managed_instance_group_to_a_target_pool

[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 👉 hashibot-feedback@hashicorp.com. Thanks!