I have a 16 VCPU server running 6.5. For ease of explanation, I have 16 VMs with 1 VCPU assigned to each. When monitoring the system as a whole, I see random spikes where just 1 or 2 VCPUs will spike to 100%. All of the rest are stable at the time when this happens. Should this be a concern. And while we are on the subject... Is 1 VCPU assigned specifically to 1 VM, or do all VCPUs share the load for the system as a whole. I.E., if VM A is idle and it's VCPU is at 0% usage, and VM B's VCPU is maxed out, will VM A's VCPU pitch in and help?
↧