Multi-CPU Fault Tolerance (SMP-FT)

The number of applications running on a server, require multi vCPU configurations, has grown exponentially. In a recent VMware customer data study, it was found that over 68 percent of the virtual machines have more than one vCPU. This means that current fault tolerance capabilities of one vCPU, used to maximize availability, can not be used on 68 percent of workloads. The cost of unplanned downtime for this class of applications could be devastating. For example, in a typical retail environment, unplanned downtime for a mission-critical application, such as a POS system, could result in losses of up to $1,000 per minute per store, during the peak shopping season.

vSphere 6.0 further innovates on technologies that deliver availability for unplanned downtime. vSphere first introduced software-based fault tolerance in 2009, and now takes it to another level by delivering true continuous availability for larger virtual machines running more than one vCPU. With...