The Reduce energy consumption in cloud computing data centres by optimizing virtual machines

Main Article Content

Marziyeh Bahrami, AblofazTarghi Haghighat, Majid Gholipour


In this paper, strategies for deciding on the need to migrate and find suitable hosts as the destination of migration are presented. The proposed algorithm, using time series prediction method and double smooth development technique (DES), predicts the processor efficiency in the future and also proposes the optimal relationship for the dynamic low threshold. The algorithm identifies and categorizes under-Loadandover-load hosts by comparing current and predicted CPU productivity with dynamic high and low thresholds, and based on this classification, migration takes place from hosts that qualify for migration. This article identifies a type of host as a troublesome host that most likely disrupts the predictive and decision-making process. In the face of these types of hosts, the algorithm implements policies to modify or put them to sleep. To find suitable hosts as an immigration destination, all over-load, over-load, low-load, and under-Loadhosts are removed from the list of suitable destinations. Improvements of 86.2%, 28.4% and 87.2% on average and in the criteria of the number of migrations of virtual machines, energy consumption and the rate of SLA violations, respectively, are the achievements of this article.

Article Details