We currently deploy virtual machines (vesphere) with an variable amount of disks using dynamic disk block. In linux we have to format/mount etc. them manually later. Now I want to automate that too with a script executed from terraform on the operating system.

But how can I identify the disks? Lets assume I attach 3 additional disks to a VM. I could hope and assume that disk 1 is /dev/sdb, disk 2 is /dev/sdc and so on. But thats not really how I want to apply that. Is there a way to put a label on the disk that is actually seen in the operating system (rhel linux) so I can be 100% sure that I work with the disk I want to?

  dynamic "disk" {
for_each = var.data_disks
content {
  label             = "disk${disk.value["lun"]}"
  size              = disk.value["disk_size_gb"]
  filesystem        = disk.value["filesystem"]
  mountpoint        = disk.value["mountpoint"]
  mountopts         = disk.value["mountopts"]
  unit_number       = disk.value["lun"]
  storage_policy_id = var.vsphere_storage_policy_id
  eagerly_scrub     = var.eagerly_scrub
  thin_provisioned  = var.thin_provisioned
}

}

filesystem, mountpoint and mountopts are things I add over a json file and plan to use for the bash-script. The "label" is not available in linux.

Thanks in advance