Relax the requirement for `unit_number=0` requirements on `r/virtual_machine_disk_subresource` with SATA or SCSI

[masert]

Terraform Version

TF v0.13.6

vSphere Provider Version

V1.24.3

Affected Resource(s)

Terraform Configuration Files

module "dv77" {
  source      = "./mod_vm_rhel_v5_import"
  mod_vm_name = "fr0-vs2a64-dv77"
...
  mod_size_disk0 = 120
  mod_un_disk0   = 1
...
}

variables.tf 

variable "mod_un_disk0" {
  # unit_number
  type    =  string
  default = "default"
}
locals {
...
  mod_un_disk0               =  var.mod_un_disk0 == "default" ? 0 : var.mod_un_disk0
...
}


resource "vsphere_virtual_machine" "mod_vm_v5" {
...
  lifecycle {
    ignore_changes = [
       annotation,
       custom_attributes,
       enable_logging,
       cpu_hot_add_enabled,
       disk.0.unit_number,
    ]
  }

  name                   = var.mod_vm_name
  num_cpus               = var.mod_num_cpu
  num_cores_per_socket   = var.mod_num_cores_per_socket
  ...

  cdrom {
    client_device = true
  }

  disk {
    label            = "disk0"
    size             = local.mod_size_disk0
    eagerly_scrub    = local.mod_eagerly_scrub_disk0
    thin_provisioned = local.mod_thin_provisioned_disk0
    disk_sharing     = "sharingNone"
    unit_number      = local.mod_un_disk0
    io_reservation   = "0"
    io_share_level   = "normal"
  }

#  disk {
#    label            = "disk1"
#    size             = var.mod
...
}

Debug Output

...
2021-01-27T12:45:23.618+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] Datastore with ID "datastore-585" found
2021-01-27T12:45:23.655+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] DiskDiffOperation: Beginning disk diff customization
2021/01/27 12:45:23 [ERROR] eval: *terraform.EvalDiff, err: at least one disk must have a unit_number of 0 for SATA or SCSI or 1 for IDE
2021/01/27 12:45:23 [ERROR] eval: *terraform.EvalSequence, err: at least one disk must have a unit_number of 0 for SATA or SCSI or 1 for IDE
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] DiskDiffOperation: Beginning collective diff validation (indexes aligned to new config)
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskLabelOrName: label: "disk0" name: ""
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskLabelOrName: Using defined label value "disk0"
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskPathOrName: Using defined path value "fr0-xxxxxxx-dv77/fr0-xxxxxxxxxx-dv77-000001.vmdk"
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] disk.0 (key 2001 at scsi:0:1) (fr0-xxxxxxxxxx-dv77/xxxxxxxxxxx-dv77-000001.vmdk): Beginning diff validation
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskLabelOrName: label: "disk0" name: ""
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskLabelOrName: Using defined label value "disk0"
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] diskPathOrName: Using defined path value "fr0-xxxxxxxx-dv77/fr0-xxxxxxxxxx-dv77-000001.vmdk"
2021-01-27T12:45:23.658+0100 [DEBUG] plugin.terraform-provider-vsphere_v1.24.3_x4: 2021/01/27 12:45:23 [DEBUG] disk.0 (key 2001 at scsi:0:1) (fr0-xxxxxxxxx-dv77/fr0-xxxxxxxx-dv77-000001.vmdk): Diff validation complete
2021-01-27T12:45:23.660+0100 [DEBUG] plugin: plugin process exited: path=.terraform/plugins/yyyyyyy/xxxxxxxx/template/2.1.2/linux_amd64/terraform-provider-template_v2.1.2_x4 pid=28827
2021-01-27T12:45:23.660+0100 [DEBUG] plugin: plugin exited
2021-01-27T12:45:23.690+0100 [DEBUG] plugin: plugin process exited: path=.terraform/plugins/yyyyyyyy/xxxxxx/null/2.1.2/linux_amd64/terraform-provider-null_v2.1.2_x4 pid=28802
2021-01-27T12:45:23.691+0100 [DEBUG] plugin: plugin exited
2021-01-27T12:45:23.691+0100 [WARN] plugin.stdio: received EOF, stopping recv loop: err="rpc error: code = Unavailable desc = transport is closing"
2021-01-27T12:45:23.691+0100 [WARN] plugin.stdio: received EOF, stopping recv loop: err="rpc error: code = Unavailable desc = transport is closing"
2021-01-27T12:45:23.693+0100 [DEBUG] plugin: plugin process exited: path=.terraform/plugins/yyyyyyyy/xxxxxx/vsphere/1.24.3/linux_amd64/terraform-provider-vsphere_v1.24.3_x4 pid=28839
2021-01-27T12:45:23.693+0100 [DEBUG] plugin: plugin exited
...

Panic Output

Expected Behavior

Actual Behavior

Fail with in apply or plan
Error: at least one disk must have a unit_number of 0 for SATA or SCSI or 1 for IDE

on mod_vm_rhel_v5_import/main_v5_DEV_rhel_import.tf line 5, in resource "vsphere_virtual_machine" "mod_vm_v5":
5: resource "vsphere_virtual_machine" "mod_vm_v5" {

Steps to Reproduce

1. terraform import ---> with instance with one SCSI disk (unit_number is 1)

2. terraform apply

3. ./terraform_136 plan -target module.dv77
   Refreshing Terraform state in-memory prior to plan...
   ...
   module.dv77.module.env_var.data.terraform_remote_state.global_var: Refreshing state...
   ...
   module.dv77.vsphere_virtual_machine.mod_vm_v5: Refreshing state...

---

Warning: Resource targeting is in effect
...
Error: at least one disk must have a unit_number of 0 for SATA or SCSI or 1 for IDE

on mod_vm_rhel_v5_import/main_v5_DEV_rhel_import.tf line 5, in resource "vsphere_virtual_machine" "mod_vm_v5":
5: resource "vsphere_virtual_machine" "mod_vm_v5" {

Important Factoids

instance state after import :
./terraform_136 state show module.dv77.vsphere_virtual_machine.mod_vm_v5
...
cdrom {
client_device = true
device_address = "ide:1:0"
key = 3002
}

disk {
    attach           = false
    controller_type  = "scsi"
    datastore_id     = "datastore-585"
    device_address   = "scsi:0:1"
    disk_mode        = "persistent"
    disk_sharing     = "sharingNone"
    eagerly_scrub    = true
    io_limit         = -1
    io_reservation   = 0
    io_share_count   = 1000
    io_share_level   = "normal"
    keep_on_remove   = true
    key              = 2001
    label            = "disk0"
    path             = "fr0-xxxxxxx-dv77/fr0-xxxxxxxxx-dv77-000001.vmdk"
    size             = 120
    thin_provisioned = false
    unit_number      = 1
    uuid             = "aaaaaa-bbbbbbb-cccccccc-ddddddddd-fffffffff"
    write_through    = false
}

network_interface {
    adapter_type          = "vmxnet3"
    bandwidth_limit       = -1
    bandwidth_reservation = 0
    bandwidth_share_count = 50
    bandwidth_share_level = "normal"
    device_address        = "pci:0:7"

...

References

error coming from 'vsphere/internal/virtualdevice/virtual_machine_disk_subresource.go' line 656

• #0000

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 other comments that do not add relevant new information or questions, 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

[DanPhonovation]

+1 on this. It's a blocker for me. I have to do my modifications out-of-bounds of terraform as a result.

In my case, I also imported a virtual machine. It's a production machine that's been running for quite some time.
I also have it's dev equivalent exhibiting the same behavior.

Both have their only "Hard Disk 1" as SCSI 0:1 and terraform sees "unit_number=1".

If I don't specify in my resource a unit number, it will attempt a change-in-place and switch it to 0.
If I set it to a "1" to match, it complains with "Error: at least one disk must have a unit_number of 0 for SATA or SCSI or 1 for IDE"

I can see how it might be good practice, but if VMWare allows it - so should terraform?

P.S. I'm on Terraform v0.14.7

[github-actions]

Marking this issue as stale due to inactivity in the past 180 days. This helps us focus on the active issues. If this issue is reproducible with the latest version of the provider, please comment. If this issue receives no comments in the next 30 days it will automatically be closed.
If this issue was automatically closed and you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. Thank you!