Why Default CPU request in namespace limitrange takes precedence over the VMs configured vCPU?

Issue

• In a namespace configured with limitrange, all the virt-launcher pods of the virtual machines got the same resources.requests.cpu although they have different vCPUs configured in the VM.
• In a namespace, all the virt-launcher pod's resources.requests.cpu is set to Default Request configured in the limits and is not configured as per cpuAllocationRatio.