I understand CPU ready times and CPU contention; what I can't seem to find the answer to is how I should evaluate CPU ready times on a multi-CPU system. If I am trying to make sure my VM is under the 5% CPU Ready time threshold, should I be looking at the CPU ready times for the VM itself, or for each vCPU? I have a system right now that has high CPU ready times on the VM itself (which I understand the reason for), but the individual vCPUs are well below 5% CPU Ready time. So which should I be looking at?
↧