Energy-Aware and Cost Effective VM Allocation Strategy Based on Best Fit MultiValued Binpaking Algorithm
Gaurav Gupta, Gopal Singh Khichar, Rakesh Rathi, Ravindar Singh · 2018
Cloud computing is growing by providing features like rapid elasticity, measured services, on demand self-service, broad network access and pay-as-you-go concepts. Cloud providers are building large data centers to achieve the regularly growing demand of cloud users. Virtualization technology is the core concept behind cloud computing which provides computing services in the form of virtual machines. A large number of cloud services are requested by a large number of cloud users dynamically in the cloud computing. So, an efficient policy must be used for providing all resources to requesting users to fulfill their needs. Therefore, one of the central challenges in cloud computing is optimal resource allocation. Cloud providers use various virtual machine placement (VMP) policies to allocate the virtual machine to the available minimum number of physical servers. In this paper, Best Fit Multi valued Bin Packing approach is proposed for optimum resources allocation in which different virtual machines are allocated to minimum number of hosts by considering three parameters: RAM, CPU processing (MIPS) and network bandwidth. This approach also uses these parameters of the available physical servers to maximize the resource utilization for reducing energy consumption. The simulations of cloud environment for this approach are achieved by using CloudSim simulator. Our results show that maximum utilization of resources is achieved by choosing least number of hosts.