proxmox_virtual_environment_file set destination file name about terraform-provider-proxmox HOT 3 CLOSED

Hi @outworld66! 👋🏼

I might not have fully grasped your use case, but there's an option to specify the target file name in the PVE storage via the file_name attribute of proxmox_virtual_environment_file. Could this be what you're looking for? You can see an example here.

Additionally, you don't have to reference file IDs via proxmox_virtual_environment_file resource IDs. For example, this line:

user_data_file_id = "local:snippets/cloud-config.yaml"

from terraform-provider-proxmox.

@bpg Hi! Thanks for the answer.

In my case, source_raw is not used, so specifying file_name is problematic. I would like to keep separate files with templates, and not insert the content into source_raw.

Usage

user_data_file_id = "local:snippets/cloud-config.yaml"

would help me, but the proxmox node does not have such files. The indication "local:..." uses the path on the proxmox node, and not the local directory from where terraform is executed? I checked, it seems to be so.

How then to upload cloud-init files to the node? Provided that separate cloud-init files are used for each VM? If you put it all together into a module, it looks impossible.

workaround: use ssh provider to upload files to the proxmox node. but why if there is a proxmox api.

from terraform-provider-proxmox.

I found a more suitable option. I'm surprised I didn't notice it before :)
For snippet, each VM needs to indicate the original name:

resource proxmox_virtual_environment_file cloud_config_meta_data {
  content_type = "snippets"
  datastore_id = "local"
  node_name    = var.virtual_environment_node_name

  source_file {
    path = "./modules/proxmox_vm/cloud-init-configs/data-vm/data-vm-meta-data"
  }
}

then the file will not be overwritten and the original snippet will be used for each VM.

from terraform-provider-proxmox.