Nutanix performance troubleshooting guidelines emphasize the importance of adequate vCPU resources when diagnosing single-VM performance issues. The documentation states that "under-provisioned CPU is a primary cause of perceived storage latency and throughput limitations, as storage I/O processing and VirtIO- SCSI operations depend heavily on VM compute resources." Running a Windows Server VM with only one vCPU significantly constrains its ability to queue and process disk I/O, handle interrupts, and run kernel-level services. This is especially true when multiple disks are attached. Adding a second controller does not solve underlying CPU bottlenecks. Adjusting br0 TCP settings is not applicable to storage. Increasing CVM resources is unnecessary because the problem is isolated to a single VM while the cluster and other VMs perform normally.
Therefore, increasing vCPU allocation is the required corrective action.