2V0-21.23 Question 382
Single answerYou are managing a virtualized environment and need to configure a virtual machine to ensure minimal interference from other workloads on the host. The application running on this VM is highly latency-sensitive, and it requires predictable CPU response times. Which configuration change would be the most appropriate to meet this requirement?
- A
Enable CPU affinity and assign the VM to specific physical cores.
- B
Increase the memory reservation for the VM to match its configured memory size.
- C
Set the VM's latency sensitivity to 'High' and reserve 100% of its allocated CPU resources.
- D
Enable VMware per-VM EVC to ensure compatibility with future host migrations.
Show answer and explanation
Correct answer: C
Explanation
Configuring a VM with high latency sensitivity is crucial for applications that require predictable CPU response times and minimal interference from other workloads. Setting the latency sensitivity to 'High' ensures the VM receives dedicated physical resources, while reserving 100% of its CPU allocation prevents resource contention. This combination is the most appropriate choice for ensuring low-latency, predictable performance.
- A. Incorrect.
While enabling CPU affinity can provide some control over CPU allocation, it is not the best practice for latency-sensitive workloads because it bypasses the ESXi scheduler, which can lead to inefficiencies. Additionally, CPU affinity does not directly address latency sensitivity requirements.
- B. Incorrect.
Increasing the memory reservation can improve performance by ensuring memory is not swapped, but it does not address the specific need for predictable CPU response times in latency-sensitive applications.
- C. Correct.
Setting the VM's latency sensitivity to 'High' and reserving 100% of its allocated CPU resources ensures that the VM gets dedicated CPU cycles, which is critical for predictable and low-latency performance.
- D. Incorrect.
VMware per-VM EVC is used to ensure CPU compatibility for vMotion, but it does not impact latency sensitivity or CPU resource guarantees.