Memory overcommitment is relatively straightforward since the introduction of large pages and other factors which affect Transparent Page Sharing (TPS). Memory overcommitment is built on the age-old vSphere principal that due to VMs not requiring all the amount of memory they are configured with, as well as leveraging TPS to share common memory blocks, we can overcommit memory on a host or cluster relatively safely.
We have already touched on this with the introduction of large pages with 64-bit operating systems and hardware-assisted memory virtualization; virtual machines now generally consume all the memory they are configured with. Also, combine this with a recent change to combat a rare security edge case inter-virtual ...
