disk offline only to actually fix the problems discovered during an online scan.
The maximum possible offline time for a volume is now 8 seconds, no matter
how large the volume. With this change, we can expect to see larger NTFS
volumes as organizations adopt Windows Server 2012 and above.Also important to note about scalability is that only very large virtual machines can be
created with tens of virtual processors, but the non-uniform memory access (NUMA)
topology is passed to the virtual machine, enabling the most optimal levels of
performance. This scalability applies to both Windows guest operating systems and
Linux, as Figure 1.6 shows with a 64 vCPU Linux virtual machine. Also note in the
figure the awareness of the NUMA nodes. This was another investment area in
Windows Server 2012: making Linux a first-class guest operating system. Nearly every
feature of Hyper-V worked equally for Windows guests and Linux guests.
Figure 1.6 Linux virtual machine running on Windows Server 2012 Hyper-V with 64
vCPUs
MOBILITY AND AVAILABILITY
As virtual machines became more scalable, the workloads that could be virtualized
increased exponentially, which makes keeping the virtual machines available even
more important. Windows Server 2012 made great advancements to the mobility and
resiliency of virtual machines. Windows Server 2008 R2 had introduced Live
Migration as a means to move virtual machines between nodes in a cluster that had
shared storage. Windows Server 2012 took this to the next level by allowing multiple
concurrent Live Migrations, which it would autoscale based on available bandwidth
and would queue until they could be performed based on network bandwidth