terraform_labels can't be updated on google_compute_instance_template

[khtelus]

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 v1.5.2
on linux_amd64

Provider version 5.3.0

Affected Resource(s)

• google_compute_instance_template

Terraform Configuration Files

 resource "google_compute_instance_template" "default" {
   project              = var.project_id
   name_prefix          = "my-host-template-"
   machine_type         = var.instance_template.machine_type
   region               = var.region
   can_ip_forward       = false
   tags                 = var.tags

   disk {
     boot         = true
     source_image = data.google_compute_image.default.self_link
     disk_type    = "pd-balanced"
     disk_size_gb = data.google_compute_image.default.disk_size_gb
   }

   network_interface {
     subnetwork         = var.subnet
     subnetwork_project = var.vpc_host_project == null ? var.project_id : var.vpc_host_project
   }

   service_account {
     email  = var.instance_template.service_account_email
     scopes = ["cloud-platform"]
   }

   metadata = {
     enable-oslogin  = "TRUE"
     enable-osconfig = "TRUE"
   }

   lifecycle {
     create_before_destroy = true
   }

   labels = var.labels
 }

Debug Output

Panic Output

Expected Behavior

With create_before_destroy set to true as per the configuration a new template should be created on update (in this case to terraform_labels) and then the deposed object deleted.

Actual Behavior

Errors out with Error: doesn't support update

Step #1 - "terraform-apply":   # module.my_hosts.google_compute_instance_template.default will be updated in-place
Step #1 - "terraform-apply":   ~ resource "google_compute_instance_template" "default" {
Step #1 - "terraform-apply":         id                   = "projects/my-project/global/instanceTemplates/my-host-template-20230320160805321300000001"
Step #1 - "terraform-apply":         name                 = "my-host-template-20230320160805321300000001"
Step #1 - "terraform-apply":         tags                 = [
Step #1 - "terraform-apply":             "tag-a",
Step #1 - "terraform-apply":             "tag-b",
Step #1 - "terraform-apply":         ]
Step #1 - "terraform-apply":       ~ terraform_labels     = {
Step #1 - "terraform-apply":           + "applicationgroup"   = "appgroup"
Step #1 - "terraform-apply":           + "contact"            = "admins_at_domain_dot_com"
Step #1 - "terraform-apply":           + "costcenter"         = "1122334455"
Step #1 - "terraform-apply":           + "dataclassification" = "tbd"
Step #1 - "terraform-apply":           + "env"                = "np"
Step #1 - "terraform-apply":         }
Step #1 - "terraform-apply":         # (12 unchanged attributes hidden)
Step #1 - "terraform-apply": 
Step #1 - "terraform-apply":         # (4 unchanged blocks hidden)
Step #1 - "terraform-apply":     }
Step #1 - "terraform-apply": 
Step #1 - "terraform-apply": Plan: 0 to add, 3 to change, 0 to destroy.
Step #1 - "terraform-apply": module.my_hosts.google_compute_instance_template.default: Modifying... [id=projects/my-project/global/instanceTemplates/my-host-template-20230320160805321300000001]
Step #1 - "terraform-apply": ╷
Step #1 - "terraform-apply": │ Error: doesn't support update
Step #1 - "terraform-apply": │ 
Step #1 - "terraform-apply": │   with module.my_hosts.google_compute_instance_template.default,
Step #1 - "terraform-apply": │   on .terraform/modules/my_hosts/main.tf line 70, in resource "google_compute_instance_template" "default":
Step #1 - "terraform-apply": │   70: resource "google_compute_instance_template" "default" {
Step #1 - "terraform-apply": │ 
Step #1 - "terraform-apply": ╵

Steps to Reproduce

1. terraform apply

Important Factoids

References

[zli82016]

Sorry for the issue. @khtelus, can you please confirm if are you trying to upgrade the provider from 4.x to 5.3.0? And also can you please provide the value of var.labels? Thanks.

[khtelus]

Thanks @zli82016 the provider (and my remote state) has already been upgraded unfortunately. Had I known, would have pinned to last v4....

Value of var.labels in this case is exactly the same as terraform_labels in example above. So var.labels would be:

labels = {
  "env"                = "np",
  "contact"            = "admins_at_domain_dot_com",
  "costcenter"         = "1122334455",
  "applicationgroup"   = "appgroup",
  "dataclassification" = "tbd"
}

Thanks!

[zli82016]

The workaround could be to make some other change- such as adding a new label or modifying the existing one to make the update success.

[khtelus]

The workaround could be to make some other change- such as adding a new label or modifying the existing one to make the update success.

This does indeed appear to be a workaround @zli82016 - if I add a new "test" tag for example running a plan step changes the expected operation on the instance group to a replace vs. an updated in-place although would need to test this out.

[khtelus]

@zli82016 I think one of the keys points here is - I don't want to change my template. Changing the template will of course in turn force re-creation of the VM instances based on this template. So adding another change (such as another label or a tag for example) to workaround this change is undesirable in my case...

[zli82016]

The PR 9394 made the fix to the resource google_compute_instance_template by adding an empty update method to make the provider upgrade succeed without resource recreation.

[saurabh21316]

@zli82016 @khtelus

I know this issue is closed but I would appreciate if someone can please help.
terraform plan:

module.ComputeInstance.google_compute_instance.mon_abc_c will be updated in-place

~ resource "google_compute_instance" "mon_abc_c" {
id = "projects/abc/zones/northamerica-northeast1-c/instances/mon-abc-c"
name = "mon-abc-c"
tags = [
"allow-healthchecks",
"mon-abc",
]
~ terraform_labels = {
+ "allow-healthchecks" = (known after apply)
+ "mon-abc" = (known after apply)
}
# (18 unchanged attributes hidden)

    # (5 unchanged blocks hidden)
}

when run terraform apply, it doesn't throw any error and running terraform plan again it shows the same output

[zli82016]

Hello, @saurabh21316 , can you please provide more details how to reproduce this issue? Which provider version does it happen? Does it happen when upgrading the provider from 4.x to 5.x? Thanks.

[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.