I have enabled dynamic memory allocation on my VMs recently, and as the title suggests, I'm very confused with the "memory demand" value shown in Hyper-V Manager.
For example, I have an Ubuntu 18.04.1LTS guest created using quick create (so it must be officially supported, right?). In the guest OS, I can see the memory usage being 1.76G + 134M swap. If I'm not mistaken, this means ~1.8G memory demand, correct? Even if I include the 10% memory buffer I've set in the VM settings, that's ~2G. However, Hyper-V shows the guest OS has a memory demand of more than 4096M, thus it's assigning all 4G of dynamic memory (max) to the guest OS.
So this is very confusing to me, where does this "memory demand" come from? Can someone else confirm the same situation or is it just me? Can I tell Hyper-V / guest OS to release the memory they are not actively using?
Does this unit have a soul?
Thread source link: https://social.technet.microsoft.com/Forums/windowsserver/en-US/5df55291-d147-4472-a454-a8d9f2ca15f8/hyperv-showing-much-higher-memory-demand-than-displayed-in-guest-os?forum=win10itprovirt