A: vSphere and vCenter Server can monitor many performance counters on hosts and virtual machines (VMs). Its counter list is so big, in fact, that figuring out which are most important can be difficult. If you're worried about memory utilization in your vCenter environment, make sure you're keeping an eye on these five (technically, six):
mem.active.average: This is the average amount of memory actively being used, based on recently touched memory pages.
mem.consumed.average: With three different focuses, this can be the amount of memory consumed by a VM for guest memory, amount used on a host, or amount of host machine memory used by all powered on VMs in a cluster.
mem.swapin.average and mem.swapout.average: These two metrics refer to the quantity of data read into or out of machine memory from the swap file since the VM was powered on.
mem.swapped.average: This metric refers to the average amount of guest physical memory that's been swapped out to its swap file by the VMkernel. This counter represents VMkernel swapping, not guest swapping.
mem.vmmemctl.average: This metric refers to the amount of memory that has been allocated by the VM memory control driver (also known as the balloon driver) and can be related to a single VM or an entire host.