unableto create empty volume

[is2admin]

Description

im unable to create an empty volume - unlike in the dcd i have to name

'image_name', 'licence_type', or 'image_alias' must be set

i just want to create additional empty drives with hosts easy in dcd but with terraform that seems not possible

Expected behavior

create empty volume

Environment

Terraform version:

Terraform v1.3.1

Provider version:

version = "6.3.2"

OS:

ubuntu

Configuration Files

How to Reproduce

Error and Debug Output

[cristiGuranIonos]

Terraform supports creating attached volumes. This is why it also requires server_id. Maybe you can try using ansible for this meanwhile? https://github.com/ionos-cloud/module-ansible/blob/master/docs/api/compute-engine/volume.md

[lachnerd]

I know that and i am able to create a volume.
server_id is no problem but i must name an image, but why ?
Doesn't make any sense to add another volume with ubuntu on it - i want a volume thats without os that i can use for app storage/logs etc.
With DCD this is an easy task - but it seems impossible with terraform.
Thx for the tip with ansible module but this is not an option.

[cristiGuranIonos]

You can try to set licence_type to "Unknown", see if that helps.

[jbuchhammer]

Usually licence_type is required. But if if an image_name or image_alias is provided the licence_type is already configured at the image and will be used instead. In that case the licence_type can be omitted.
Besides "UNKNOWN" you can also use "OTHER" as value.

[lachnerd]

ok i tested adding a volume with

resource "ionoscloud_volume" "example" {
  datacenter_id           = var.datacenter_id
  server_id               = ionoscloud_server.host.id
  name                    = "Volume Example"
  availability_zone       = var.availability_zone
  size                    = "100"
  disk_type         = var.disk_type
  licence_type            = "OTHER"   
  ssh_key_path = [var.public_key_file]
  image_password        = random_string.root_password.result
}

Does not work (same with 'licence_type = "UNKNOWN"')

'''
│ Error: an error occured while creating a volume: 500 Internal Server Error {
│ "httpStatus" : 500,
│ "messages" : [ {
│ "errorCode" : "301",
│ "message" : "Oops! Something went very wrong. Please contact the administrator"
│ } ]
│ }
'''

[gabriel-savu]

Hello, we will take a look and then will inform you when we'll have a solution.

[cristiGuranIonos]

There was a backend maintenance and that might have cauised the 500 err response. Would it be possible to try again and let us know if it reproduces? Thanks!

[gabriel-savu]

Hello, I managed to reproduce the behavior for the volume. We opened a issue too because it's an api error, but if you want to create a empty volume, you can remove the ssh_key_path for the moment.
For me worked like this:

resource "ionoscloud_volume" "example" {
  datacenter_id           = ionoscloud_datacenter.example.id
  server_id               = ionoscloud_server.example.id
  name                    = "Volume Example"
  availability_zone       = "AUTO"
  size                    = "100"
  disk_type               = "HDD"
  licence_type            = "OTHER"   
}

We'll lelt you know when the pronlem with ssh_key_path will be solved and thank you for let us know.

[lachnerd]

│ Error: an error occured while creating a volume: 422 Unprocessable Entity { │ "httpStatus" : 422, │ "messages" : [ { │ "errorCode" : "100", │ "message" : "[(root).properties.imagePassword] Attribute 'image' or 'imageAlias' must be set for password" │ } ] │ }

nope unfortunately also not working...

[gabriel-savu]

hello, this is the error for the plan that I provided above? I am asking, because with that example I was not able to reproduce this error, but looking to the error, I think you have the user_data field set in terraform plan, which force you to provide either image or image_alias field.

[lachnerd]

hello, this is the error for the plan that I provided above?

yap, correct

[gabriel-savu]

but you have the user_data field set in terraform plan? Normally this field will trigger this error. If you want to use user_data field, you need to provide either a public image or image_alias. if not it will return an the api error from above

[lachnerd]

negative, no "user_data" field set

  datacenter_id           = var.datacenter_id
  server_id               = ionoscloud_server.host.id
  name                    = "Volume Example"
  availability_zone       = var.availability_zone
  size                    = "33"
  disk_type         = var.disk_type
  licence_type            = "UNKNOWN"   
  #ssh_key_path = [var.public_key_file]
  #image_password        = random_string.root_password.result
}```

[gabriel-savu]

can you please attach the server too to a message here be more easy to reproduce?

[lachnerd]

server ressource

resource "ionoscloud_server" "host" {
    name                  = var.name
    datacenter_id         = var.datacenter_id
    cores                 = var.cpu_core
    ram                   = var.ram * 1024
    availability_zone     = var.availability_zone
    cpu_family            = var.cpu_family
    image_name            = var.image_name
    image_password        = random_string.root_password.result
    #Will be renamed to ssk_keys in the future, to allow users to set both the ssh key path or directly the ssh key
    ssh_key_path = [var.public_key_file]
    type                  = var.type
    volume {
        name              = "system"
        size              = var.disk_size
        disk_type         = var.disk_type
    }
    nic {
        lan               = var.datacenter_lan_id
        name              = "${var.name}-lan"
        dhcp              = true
        firewall_active   = var.firewall_active
        #default 'inbound'
        #firewall_type     = "BIDIRECTIONAL"
        #ips               = [ ionoscloud_ipblock.example.ips[0], ionoscloud_ipblock.example.ips[1] ]
        ips               = var.ip_adresses
  }

  lifecycle {
    #nic makes trouble on reapply otherwise
    ignore_changes = [availability_zone, nic]
  }
}

[gabriel-savu]

Thank you, I'll try to reproduce and come with some updates

[gabriel-savu]

I could't managed to reproduce the error with the tf plan that you provided, so here is a working example:

resource "ionoscloud_server" "host" {
    name                  = "Server Example"
    datacenter_id         = ionoscloud_datacenter.example.id
    cores                 = 1
    ram                   = 1024
    availability_zone     = "ZONE_1"
    cpu_family            = "AMD_OPTERON"
    image_name            = "ubuntu-22.04-server-cloudimg-amd64-20220902"
    image_password        = "aPassword"
    #Will be renamed to ssk_keys in the future, to allow users to set both the ssh key path or directly the ssh key
    ssh_key_path          = ["path/to/public/rsa/key"]    
    type                  = "ENTERPRISE"
    
    volume {
        name              = "system"
        size              = "100"
        disk_type         = "HDD"
    }
    nic {
         lan               = ionoscloud_lan.example.id
        name              = "system nic"
        dhcp              = true
        firewall_type     = "BIDIRECTIONAL"
        #default 'inbound'
        #firewall_type     = "BIDIRECTIONAL"
        ips               = [ ionoscloud_ipblock.example.ips[0], ionoscloud_ipblock.example.ips[1] ]
        #ips               = var.ip_adresses
  }

  lifecycle {
    #nic makes trouble on reapply otherwise
    ignore_changes = [availability_zone, nic]
  }
}

maybe there is a variable value which trigger this error, so if you want, you can provide the values for vars, except sensitive data, of course to try to use them and see if I can get 422 error.

[gabriel-savu]

Hello, again, we found a case for the error, if you provide a licence_type for the volume resource, but the server does not have a image_name, then a 500 error will be displayed. We opened a ticket for this for another team and we will reply wen we'll have some updates.