Q: 1
An administrator is experiencing storage performance issues on a Windows Server 2019 VM with the
following configuration:
vCPU: 1
VRAM: 8 GB
vSCSI: VirtIO SCSI Controller
vDisk: 2 (100 GB, 250 GB)
vNIC: VirtIO Fast Ethernet
The AHV cluster is healthy, and other Windows VMs are performing well.
Which configuration change should be reviewed to enhance VM performance?
Options
Discussion
Why not A if the VM had super high disk I/O instead of just generic lag?
Its D. With just 1 vCPU, the VM is probably hitting a CPU bottleneck and that slows down disk ops too. Adding more vCPUs lets the system handle I/O better. I think this is the main thing to fix here, unless there was way heavier disk activity than described. Anyone see a reason to pick A instead?
Sick of these Nutanix questions trying to trick you with controller stuff, it's D tbh.
D would be my pick here. With just 1 vCPU, the VM can easily become CPU-bound when handling storage I/O, especially on Windows. Adding vSCSI controllers (A) really only helps in high throughput setups, not in basic performance fixes like this. Pretty sure more vCPUs (D) is what the exam wants, unless I missed something-agree?
A or D? Exam practice questions and the official guide both talk about adding vSCSI controllers for bigger workloads. I thought more vSCSI could help with disk performance, so not sure here. Anybody else using lab setups see a real difference?
Pretty sure it's D here. Official guide and practice exams both say to check vCPU first for generic storage lag like this, unless the VM is super heavy on disk I/O. Anyone see a reason to pick something else?
It’s D, not A. Single vCPU is almost always the limiting factor unless it’s a crazy I/O use case.
D fits here. Only one vCPU limits how many disk and network requests the VM can handle, so bumping that up is best for general storage slowness. Saw similar advice in both the official guide and some practice questions. Pretty confident, but let me know if you disagree.
Not convinced A is useful here, since adding another vSCSI controller really benefits high disk I/O VMs, and this VM only has two disks with average size. D looks more valid to me because a single vCPU often can't keep up with Windows storage requests. People might pick C but CVM resources affect the whole cluster, not just one VM. Anyone disagree?
D imo. With just 1 vCPU, Windows tends to bottleneck on I/O no matter what storage controller you pick. Option A (adding another vSCSI) works for very large throughput demands but not basic cases like this. Not sure if anyone else noticed, but B is a common trap since Balance-TCP doesn't touch storage problems. Open to other views but pretty sure it's D.
Be respectful. No spam.
