The most operationally efficient and rapid method to meet the requirements is to use a Managed Instance Group (MIG) based on an instance template. An instance template defines the configuration of the virtual machines, ensuring consistency. The MIG uses this template to manage a group of identical VMs. By configuring the MIG's autoscaler with a policy based on average CPU utilization, Google Cloud's managed service will automatically add or remove VMs to maintain the desired performance level. This native solution directly addresses the need to use VMs, scale on CPU, and do so efficiently without custom tooling.

A. Google Kubernetes Engine is a container orchestration service. It scales container pods, not virtual machines directly as required by the policy.

C. This option describes schedule-based autoscaling, which scales based on time, not on the required real-time CPU usage metric.

D. Building and maintaining a custom scaling solution with third-party tools is significantly less operationally efficient and more complex than using the native, fully managed autoscaler.

1. Google Cloud Documentation

"Autoscaling groups of instances": "Autoscaling helps your applications gracefully handle increases in traffic and reduces costs when the need for resources is lower. You define the autoscaling policy

and the autoscaler performs automatic scaling based on the measured load." The document further specifies CPU utilization as a primary scaling metric. (Source: Google Cloud

Compute Engine Documentation

"Autoscaling groups of instances").

2. Google Cloud Documentation

"Scaling based on CPU utilization": "You can configure the autoscaler to scale a managed instance group (MIG) based on the average CPU utilization of the instances in the group... The autoscaler adds or removes virtual machine (VM) instances from the MIG to maintain the target CPU level." This directly supports option B. (Source: Google Cloud

Compute Engine Documentation

"Autoscaling policies"

Section: "Scaling based on CPU utilization").

3. Google Cloud Documentation

"Instance templates": "An instance template is a resource that you can use to create VM instances and managed instance groups (MIGs). Instance templates define the machine type

boot disk image...

and other instance properties." This confirms that an instance template is the correct prerequisite for a MIG. (Source: Google Cloud

Compute Engine Documentation

"Instance templates").

4. Google Cloud Documentation

"Google Kubernetes Engine (GKE) overview": "GKE is a managed

production-ready environment for deploying containerized applications." This highlights that GKE's primary abstraction is the container

not the underlying VM

making option A incorrect for the stated requirement. (Source: Google Cloud

GKE Documentation

"Concepts"

Section: "Overview").