Load Distribution and System Shutdown for Power-Efficient Computing in Heterogeneous Server Farms

Modern server farms consume large amounts of energy by computing devices and cooling equipment. During the course of the day the load of these servers usually varies significantly. Especially, in times when they are under-utilized and work in an power-inefficient range, load distribution should be optimized to minimize total power consumption. This work presents an approach to generate load distribution schemes for multiple, heterogeneous physical servers in a power-efficient manner. Furthermore, we developed an algorithm that identifies inefficient servers that can be powered off to further reduce power consumption significantly. Based on this new switching pattern, the workload distribution is recalculated.