Proactive Workload Prediction and Resource Management in Hybrid Cloud using Machine Learning Techniques

Abstract

newlineCloud Computing (CC) paradigm has improved information and communication in newlinerecent years and provided a backbone to modern infrastructure. CC enhances the newlineservices of organizations such as Government, industries, and academia with a payas- newlineyou-go model. More than 60% application workload is migrated to CC. The applications newlinehosted on CC heavily use resources and generate more traffic, specifically newlineduring specific events. The management of resources is one of the issues in CC. newlineTo achieve better quality in service provisioning and avoid Service Level Agreement newline(SLA) violation, the elasticity of resources is a major requirement in CC. The hybrid newlinecloud model excels in resource requirements with private and public cloud services to newlinedeploy elasticity applications. The resource monitoring and prediction improve the newlineresource management policy with elasticity. For elasticity, a traditional adaptive policy newlineimplements threshold-based auto-scaling approaches that are adaptive and simple newlineto follow. However, such a static threshold policy may not be effective during a newlinehigh-dynamic and unpredictable workload. An efficient auto-scaling technique that newlinepredicts the system load is essential. Balancing the dynamism of load through the best newlineauto-scale policy is still a challenging issue. This research work addresses resource newlineprediction mechanisms to handle workload demands in CC through ML techniques. newlineThis work explores how these techniques can be adapted to resource management newlineproblems to increase resource availability and reduce SLA violations of Cloud data newlinecenters while simultaneously satisfying application QoS requirements. The data center newlineparameters such as CPU utilization and users requests are analyzed and suggest newlinean algorithm using Machine learning and Queuing theory concepts that pro-actively newlineindicate an appropriate number of future computing resources for short-term resource newlinedemand. The experiment shows that the suggested model enhances the elasticity of resources newlinewith performance metrics. The suggested approach i