Thanks for your reply. I'd tend to agree with you that the more important/relevant metric is %CPU Ready for the VM as a whole. I can't think of a time (until a discussion I had yesterday) I heard anyone refer to %CPU Ready on a per-vCPU basis. There are several sources out there (thanks Google!) that imply or outright state that %CPU Ready should be broken down to a per-vCPU metric. However the VMware KBase article (2002181) for calculating %CPU Ready from the summation values says nothing about doing so. I know that technically each vCPU has a max. of 100% (of any CPU-related metric) and therefore a 4 vCPU machine would have the potential of 400% CPU utilization or 400% CPU ready...but who actually uses metrics that way? I don't know anyone (not saying this is necessarily a good thing) that says "the machine is using 300% CPU"...they say "it's using 75% CPU". Both the performance tab within the client and vROPS tend to line up with the per-VM metrics...you can pull the per-vCPU numbers but who does that?
Is there anyone on the per-vCPU "side" of the argument that can state their case? If that's the "right" way I'm open to it but I'd like to understand the utility of breaking the number down that far. If I say "keep the CPU ready on that VM below 5%" does anyone out there interpret that as 5% times the number of vCPUs in the VM?