Energy Efficiency and Establishing Service Level Agreement using Fuzzification of Virtual Machine Selection Policies for Migrating in Cloud Computing

Cloud computing has grown exponentially as an essential technology for the development of Internet services these days. The spread of this technology has led to the creation of large-scale data centers and complex practical applications with a wide range of communication needs. There has been an increased demand for cloud services in recent decades which led to higher energy consumption and Service Level Agreement Violation (SLAV). Through utilizing the Service Level Agreement Violation (SLAV) parameter, it is possible to calculate the SLA of a data center. Via reducing the number of migrations of Virtual Machine (VM) aggregation, we can establish the SLA and reduce the SLAV accordingly. In this study, we have proposed a suitable method per consolidation for choosing an appropriate VM to migrate from an overload host to a destination host using fuzzy logic. For doing so, a variety of VM selection policies for migration, such as Minimum Migration Time (MMT), Maximum Correlation (MC), migration control, and Maximum Power Reduction (MPR) have been fuzzified; hence different parameters in selecting a VM for migration have been considered. Using CloudSim simulator was indicated that the proposed approach had made improvements in the case of the number of migrations, total SLAV, SLA time per active host, and energy consumption. Using the migration policies fuzzification, we reduced SLAV and energy consumption.