ibm_pi_instance is not able to complete with pi_health_status "WARNING"

[Beilmann]

Documentation describes, that by setting "health state" to "warning" by creating a new PowerVS instance, the creation process would complete successfully after the instance becomes state "warning". This would reduce deployment time of large instances from 1 hour to 15-20 minutes.

At the moment it is not working.

Documentation:
https://registry.terraform.io/providers/IBM-Cloud/ibm/latest/docs/resources/pi_instance

Terraform CLI and Terraform IBM Provider Version

Terraform v1.0.11
Provider 1.36.0

Affected Resource(s)

ibm_pi_instance

Terraform Configuration Files

resource "ibm_pi_instance" "snat-host" {
    pi_memory             = "4"
    pi_processors         = "1"
    pi_instance_name      = "${var.prefix}-SNAT"
    pi_proc_type          = "shared"
    pi_image_id           = var.pi_image_id 
    pi_network_ids        = var.pi_network_ids
    pi_key_pair_name      = var.pi_key_pair_name
    pi_sys_type           = "s922"
    pi_health_status     = "WARNING"
    pi_cloud_instance_id  = var.pi_cloud_instance_id
    pi_volume_ids         = []
    pi_storage_type       = "tier1"

    timeouts {
     create = "60m"
     update = "30m"
     delete = "30m"
    }

    lifecycle {
      ignore_changes = all
      prevent_destroy = true
    }
}

Expected Behavior

Instance creation completes after the remote instance becomes health state 'warning'

Actual Behavior

Instance creation is running for 2 hours and ends with an error.

Additional

Discussed internally. Probably change to:

if *pvm.Status == helpers.PIInstanceAvailable && (pvm.Health.Status == allowableStatus || pvm.Health.Status == "OK") {

in isPIInstanceRefreshFunc would help.