2 . The virtual machine is restarted or the host is restarted.
3 . To restart, the virtual machine needs 1GB of memory, but it has only 512 MB
available. In this worst-case scenario, the Hyper-V host has no spare memory, and
no memory can be reclaimed from other virtual machines running on the host.This one and only scenario is where the new Smart Paging feature is utilized:
The virtual machine is being restarted (also caused by host restart).
There is no available physical memory.
No memory can be reclaimed from other virtual machines running on the host.At this time, a Smart Paging file will be created for the virtual machine in the location
specified in the configuration of the virtual machine and will be used by the virtual
machine as memory to complete startup. As soon as possible, that memory mapped to
the Smart Paging file will be ballooned out, and the Smart Paging file will be no longer
used and deleted. The target time to stop using the Smart Paging file is as soon as
possible and no longer than 10 minutes. The Smart Paging feature is used only to
provide reliable restart of virtual machines, and it is not used to provide overcommit
after boot.
WHY PAGE SHARING IS NOT USED BY HYPER-V
When virtualization technologies are used, it’s common to run many operating
system instances (often similar versions) on one physical piece of hardware. On
my main server, I have 18 virtual machines all running Windows Server 2012 R2.
The operating system version is the same, which means that a large part of their
memory content will have the same data as other virtual machines running the
same guest operating system.
The idea of page sharing is storing only duplicate pages of memory from all of the
virtual machines once stored in physical RAM, basically single-instance storage
for virtual machine memory. One way this can work is that a process in the
hypervisor looks at every page of memory for every virtual machine, creates a
hash value for each page of memory, and then compares the hash values. If a
duplicate hash is found, a bit-by-bit comparison of the memory pages is
performed to make sure that the memory pages really are identical, and then the
content is stored only once in memory and the duplicate virtual machine page
addresses point to the singly stored page. We are now sharing the page. This
seems like a great idea, but for many reasons it doesn’t work well with newer
operating systems, Windows Server 2008 and later.
First, page sharing works best on empty pages. However, as you saw in the
previous section with Windows Vista and above, memory is rarely left empty and
is used to cache as much as possible.
Second, memory pages are getting bigger—much bigger. In the past, memory