An administrator observes the following symptoms for a virtual machine: -CPU usage is consistently above 90% -CPU ready value is consistently above 20%. -Application performance is impacted. Which…

Adding more vCPUs increases scheduling complexity because the hypervisor must find enough simultaneous free physical CPU threads to co-schedule all vCPUs together, which typically raises CPU ready time rather than reducing it under an already contended host.

Reducing vCPUs does not address the root cause of host-level CPU contention and may degrade application throughput if the workload legitimately relies on multiple vCPUs for parallel processing.

Without VMware Tools, guest OS idle loops perform a CPU spin-wait rather than issuing a halt instruction, causing idle VMs to consume physical CPU cycles they are not actively using; installing VMware Tools on every VM enables proper CPU halting so unused cycles are returned to the host scheduler, reducing contention and lowering the target VM's CPU ready time.

CPU shares determine a VM's relative priority in the ESXi CPU scheduler; increasing shares for this VM means it is scheduled ahead of lower-priority VMs during contention periods, directly reducing the time it spends waiting in the run queue and lowering the CPU ready metric.