Using dynamic voltage frequency scaling and service‐level agreement to reduce energy consumption in cloud datacenters based on distance

Abstract

Energy consumption in cloud is an important topic. Previous works have reduced energy consumption, but have not consider and optimal selection of servers in a datacenter. In our paper, we try to elicit optimal selection of servers in a datacenter. We will modify the architecture of the system and the proposed algorithm using service‐level agreement (SLA) for datacenter servers in this selection, we consider the distance between the servers and the job and if there is a violation of SLA, the migration of virtual machines (VMs) occurs. On the other hand, unnecessary migration of VMs increases energy consumption. So we provide an algorithm for migrating VMs, and because we need task scheduling to perform tasks in the cloud, we will also provide algorithms for tasks migrating. Energy consumption is important for us, we use dynamic voltage frequency scaling to reduce energy consumption. On the other hand, the selection of service providers is done in the SLA management, and these service providers are located on different datacenters, we consider the parameter of the distance to the service consumer in selecting the appropriate datacenter.