Soru

Zorluk: OrtaConfigure Virtual Machine Scale Sets (VMSS)

You are configuring an Azure Virtual Machine Scale Set (VMSS) named `vmss-sales` that uses Uniform orchestration, has an initial instance count of 1010, and uses an Automatic upgrade policy. The scale set is deployed in a subnet that has only 55 available private IP addresses. You plan to configure an autoscale rule to scale out the VMSS by 44 instances. You need to ensure that the scale-out operation succeeds without failing due to a lack of available IP addresses. Which configuration change should you apply to the scale set?

  1. Set the `overprovision` property of the scale set to `false`.Cevap
  2. B
    Set the upgrade policy of the scale set to Manual.
  3. C
    Set the orchestration mode of the scale set to Flexible.
  4. D
    Configure a scale-in policy of OldestVM.

Cevap

Set the `overprovision` property of the scale set to `false`.
The correct option is to set the `overprovision` property of the scale set to `false`. By default, Virtual Machine Scale Sets have overprovisioning enabled, which means Azure provisions more VMs than requested to speed up deployment times. Once the requested number of instances are healthy, the extra instances are deleted. In a subnet with constrained IP availability (only 55 IPs available), overprovisioning 44 instances could cause the deployment to fail because the total number of provisioned VMs during the operation might exceed 55. Disabling overprovisioning ensures exactly 44 instances are provisioned, consuming only 44 IP addresses.

Adım Adım Çözüm

1
Analyze the subnet IP capacity and the scale-out requirement.
The subnet has 55 available private IP addresses, and the scale-out event requires 44 new instances.
Before configuring scaling, you must verify if the subnet has enough capacity for the target instance count.
2
Identify the cause of potential IP exhaustion during scale-out.
By default, Azure VMSS has overprovisioning enabled (`overprovision` = `true`), which spins up more instances than requested (e.g., 55 or 66 instances for a 44-instance request) to speed up deployment times.
Overprovisioning consumes extra IP addresses temporarily, which will fail if only 55 IPs are available and the overprovisioned count exceeds 55.
3
Disable overprovisioning in the VMSS configuration.
Setting `overprovision` to `false` ensures that Azure only attempts to provision exactly 44 instances, utilizing exactly 44 IP addresses.
This guarantees that the scale-out operation stays within the subnet's limit of 55 available IP addresses.

Anahtar Kavram

Disabling VMSS overprovisioning to prevent IP exhaustion in constrained subnets during scale-out operations.
Bu soruyu puanla