Abstract
Cloud computing is aimed at providing computing resources and services to the Cloud users on demand and on payper-use model with ease of flexibility, availability, reliability and elasticity. Increase usage of Cloud has lead to a concern of energy consumption by numerous Cloud data centers across globe. Dynamic virtual machine (VM) consolidation is best solutions to scale down energy consumption and scale up resource utilization. Maximum real works compromise with reduce number of hosts in data center without considering future resource requirements that may result into increase in unnecessary VM migration and service level agreement (SLA) violation. To address the issue, we recommend predicting host utilization and propose a modified utilization prediction aware best fit decreasing (MUP-BFD) algorithm to reduce energy consumption, decrease number of VM migrations and diminish number of SLA violation. The propose mechanism is planned to be implemented in CloudSim and the results are yet to be compared with existing mechanisms.
Authors
Dimpal Patel