PROJECT TITLE :

A Multi-objective Optimization Scheme for Job Scheduling in Sustainable Cloud Data Centers

ABSTRACT:

Globally, there has been a rapid increase in the green city revolution for a number of years due to an exponential increase in the demand for an eco-friendly environment. This demand has been the driving force behind the rapid increase. As a direct consequence of this, the load shifting of significant energy consumers from conventional power grids to renewable energy sources (RES) has become an unavoidable necessity. In this regard, cloud data centers, also known as DCs, have emerged as significant consumers of energy that are wholly dependent on power grids to fuel the day-to-day operations of their businesses. Despite this, their overall energy consumption has skyrocketed, which has led to a sizeable increase in the rate of the global carbon footprint. The most effective method for overcoming these challenges is to make strategic use of renewable energy sources, which come with a host of well-established benefits, including lower operational costs and lower carbon emissions. In light of the information presented above, the work that is being proposed has as its ultimate objective the design of an all-encompassing workload classification, job scheduling, and virtual machine placement architecture for cloud data centers that are powered by renewable energy sources (RES) and power grids. In order to accomplish this, a multi-objective optimization strategy that functions in two stages has been proposed. During the first phase of the project, a random forest-based wrapper scheme called Boruta is utilized for the purpose of selecting relevant feature sets for the incoming workload. After this, a classification of the workload using an approach based on locality-sensitive hashing and support vector machines is carried out. During phase II, a multi-objective optimization problem for job scheduling and VM placement is formulated. This problem takes into consideration a variety of parameters, including service level agreement (SLA), energy cost, carbon footprint rate (CFR), and availability of renewable energy sources. The problem was ultimately solved by employing an improved heuristic approach that was founded on a greedy strategy. Our experimental evaluations show an average improvement of approximately 31 percent in energy utilization, 28 percent in energy cost, and 36 percent in CFR when compared with the existing schemes; however, there is a slight degradation in SLA assurance (about 2 percent).