Paper Framework for Power Aware Scheduling in Cloud Environment

Rapid growth of the demand for computational power by scientific, business and web-applications has led to the creation of large-scale data centers consuming enormous amounts of electrical power. Despite the improvements in energy efficiency of the hardware, overall energy consumption continues to grow due to increasing requirements for computing resources. The main challenge faced by cloud data centers is this power consumption by the servers. There is no proper scheduling approach for cloud providers to provide an optimal scheduling of the virtual machines (VM) to keep the power consumption minimum and to keep the Quality of Service (QoS). The proposed power aware scheduler (PASE) is based on the macro-modeling of the servers by a-priori estimation of power consumption. The a-priori estimation gives a power metric which effectively represents the relative power consumption of the servers. The ranking of the servers is based on this power metric. Here scheduling is done based on this model and there is measure for power-performance trade-off to maintain the QoS. The power aware scheduling is done by effectively scheduling the VMs by their live migration in the cloud servers based on the power ranking of the servers and also on a measure based on its CPU load.