Question

Difficulty: MediumConfigure Virtual Machine Scale Sets (VMSS)

An administrator is configuring an Azure Virtual Machine Scale Set (VMSS) named VMSS1 that uses Uniform orchestration. VMSS1 currently has five instances running a web application, and the upgrade policy is set to Manual. You need to configure VMSS1 to meet the following requirements:
- When a new version of the OS image is published, the instances must be updated automatically in batches to prevent application downtime.
- VMSS1 must scale out by two instances when the average CPU utilization exceeds 80%80\% for 10 minutes.
- VMSS1 must scale in by one instance when CPU utilization drops, but you must prevent rapid scaling loops (thrashing).

Which two configuration settings should you implement? (Select two.)

  1. Change the upgrade policy of VMSS1 to Rolling.Answer
  2. B
    Change the upgrade policy of VMSS1 to Automatic.
  3. Configure a scale-in rule with an average CPU threshold of 30%30\% and a cooldown period of 10 minutes.Answer
  4. D
    Configure a scale-in rule with an average CPU threshold of 75%75\% and a cooldown period of 1 minute.

Answer

To configure the scale set correctly, you must change the upgrade policy to Rolling to ensure updates occur in batches without causing application downtime, and configure a scale-in rule with a low CPU threshold (such as 30%30\%) and an adequate cooldown period (such as 10 minutes) to prevent instance thrashing.
The Rolling upgrade policy updates instances in batches, which prevents downtime by keeping a subset of instances online. To prevent instance thrashing, the scale-in metric threshold must be set significantly lower than the scale-out threshold, and the cooldown period must be long enough to let the system stabilize after scaling actions.

Step-by-Step Solution

1
Select the appropriate upgrade policy for batch updates.
The Rolling upgrade policy is chosen, as it updates instances in sequential batches rather than all at once.
This maintains application availability during image upgrades.
2
Determine the scale-in CPU threshold and cooldown window to prevent thrashing.
A threshold of 30%30\% CPU utilization with a 10-minute cooldown is selected.
This leaves a wide margin between scale-out (80%80\%) and scale-in (30%30\%), giving the new instances time to boot and handle load before another scaling decision is made, preventing rapid scale-up/scale-down cycles.

Key Concept

Configuring Virtual Machine Scale Set upgrade policies and preventing autoscale thrashing.
Rate this question