In migrating three on‑premise servers to virtual machines, which minimum CPU resources on the second host would best support the new VMs?

In migrating three on‑premise servers to virtual machines, which minimum CPU resources on the second host would best support the new VMs?

• A. 1,600 MHz for the application server and 2,300 MHz for each web server
• B. 3,300 MHz for the application server and 4,600 MHz for each web server
• C. 3,300 MHz for the application server and 5,250 MHz for each web server
• D. 5,250 MHz for the application server and 3,100 MHz for each web server

Answer: (A)

When migrating to virtual machines, you size the CPU resources for each VM to meet its workload while ensuring the host can supply the total demand without bottlenecks. For three VMs—a application server and two web servers—you sum the minimum CPU needs of all VMs to determine the host’s required capacity. The allocation that assigns 1,600 MHz to the application server and 2,300 MHz to each web server provides a total of 6,200 MHz across the host. This is the smallest total that still meets each VM’s needs, making efficient use of the host’s CPU resources. The other options require more total CPU for the VMs, which would waste capacity or reduce headroom for spikes. In practice, you’d also account for hypervisor overhead and potential overcommit, but as a minimum-size solution, this distribution is the most appropriate.

When migrating to virtual machines, you size the CPU resources for each VM to meet its workload while ensuring the host can supply the total demand without bottlenecks. For three VMs—a application server and two web servers—you sum the minimum CPU needs of all VMs to determine the host’s required capacity.

The allocation that assigns 1,600 MHz to the application server and 2,300 MHz to each web server provides a total of 6,200 MHz across the host. This is the smallest total that still meets each VM’s needs, making efficient use of the host’s CPU resources. The other options require more total CPU for the VMs, which would waste capacity or reduce headroom for spikes. In practice, you’d also account for hypervisor overhead and potential overcommit, but as a minimum-size solution, this distribution is the most appropriate.