virtual machine scale set using terraform without load balancer? - azure

Question

I'm trying to deploy a Virtual Machine Scale Set Extension via Terraform but there are few issues here, The requirement was to implement without loadbalancer attached
resource "azurerm_virtual_machine_scale_set" "example" {
name = "mytestscaleset-1"
location = azurerm_resource_group.example.location
resource_group_name = azurerm_resource_group.example.name
# automatic rolling upgrade
# automatic_os_upgrade = true
upgrade_policy_mode = "Rolling"
rolling_upgrade_policy {
max_batch_instance_percent = 20
max_unhealthy_instance_percent = 20
max_unhealthy_upgraded_instance_percent = 5
pause_time_between_batches = "PT0S"
}
sku {
name = "Standard_F2"
tier = "Standard"
capacity = 2
}
storage_profile_image_reference {
publisher = "Canonical"
offer = "UbuntuServer"
sku = "16.04-LTS"
version = "latest"
}
storage_profile_os_disk {
name = ""
caching = "ReadWrite"
create_option = "FromImage"
managed_disk_type = "Standard_LRS"
}
storage_profile_data_disk {
lun = 0
caching = "ReadWrite"
create_option = "Empty"
disk_size_gb = 10
}
os_profile {
computer_name_prefix = "testvm"
admin_username = "myadmin"
}
os_profile_linux_config {
disable_password_authentication = true
ssh_keys {
path = "/home/myadmin/.ssh/authorized_keys"
key_data = file("C:/Users/User/Downloads/VmSS key/azkey")
}
}
network_profile {
name = "terraformnetworkprofile"
primary = true
ip_configuration {
name = "TestIPConfiguration"
primary = true
subnet_id = azurerm_subnet.example.id
public_ip_address_configuration {
name = "Avx192"
idle_timeout = 30
domain_name_label = "vjst23"
}
}
}
tags = {
environment = "staging"
}
}
Once deployed its giving error for health probe
│ Error: compute.VirtualMachineScaleSetsClient#CreateOrUpdate: Failure sending request: StatusCode=0 -- Original Error: Code="BadRequest" Message="Rolling Upgrade mode is not supported for this Virtual Machine Scale Set because a health probe or health extension was not provided."
│
│ with azurerm_virtual_machine_scale_set.example,
│ on Se.tf line 81, in resource "azurerm_virtual_machine_scale_set" "example":
│ 81: resource "azurerm_virtual_machine_scale_set" "example" {
How to provide health probe directly if there is no load balancer attached to the deployment?

Answer 1

As you are deploying without load balancer . So , you need to do the below changes in your code:
Change the upgrade_policy_mode = "Rolling" to upgrade_policy_mode = "Manual"/"Automatic"
Remove the below block :
rolling_upgrade_policy {
max_batch_instance_percent = 20
max_unhealthy_instance_percent = 20
max_unhealthy_upgraded_instance_percent = 5
pause_time_between_batches = "PT0S"
}