Journal of Contemporary Issues in Business and Government

Abstract

Cloud computing is one of the leading computing paradigms that offers services like Infrastructure as a Service called IaaS, Platform as a Service called PaaS, Software as a Service called SaaS to users on a pay per use model. The massive data centers that help cloud offer all the above stated services are virtualized. Virtualization enables easy management of resources. However, the massive physical servers in the data centers tend to consume enormous energy, leading to high environmental impact. So, energy conservation with optimum usage and management is one of the prominent areas of research in cloud. The major techniques to manage energy is to identify unused physical resources and put them to low power state or sleep state. But, the usage of resources depends heavily on the user requirements in an elastic environment like cloud. Hence machine learning techniques can be used to predict the usage patterns thereby identifying the physical resources required to fulfill the user demand. This paper aims to survey the avenues wherein machine learning can be applied to help energy management in a cloud data center