An enhanced approach for profit scheduling with maximum utilization of resources in cloud computing environment

Abstract

With the increasing demand to effectively handle the huge volume of data in real time in the current scenario, there has been an increased need in research for centralized cost effective management systems. Daily fluctuations in the market value and dynamic behavior of input load conditions cause the management and control systems to hang up during peak demand seasons or consequently crash. Hence, a gradual migration of almost all business and information services sector have been observed in recent times to clouds which offer an optimal solution in various aspects for a cost effective implementation. Cloud computing basically means the delivery of services from a pool of resources effectively to the user based on demand. They share data and offer services to their customers on a global scale based on demand and in a very transparent manner. Several service models as well as deployment types are offered by cloud to the service providers to increase the quality of service rendered to the customers. Virtual machines are the heart of any cloud computing service as they are responsible for translation of physical machine into multiple virtual systems for multiple users providing a wide range of services. The primary objective of the thesis lies in developing an effective scheduling strategy for the virtual machines which are scalable, computationally quick and energy savings optimized. The thesis proposes a number of algorithms based on profit based scheduling, cluster partitioning for reduction newlineof time and greedy algorithms for energy savings. Apart from formulating simple heuristic algorithms for virtual machine scheduling in different cloud models for SaaS services, the thesis also investigates the performance of the algorithms for scalability, energy efficient and optimal resource allocation. The programming has been implemented on C/C++ in a Cloud Sim toolkit and the validation of results demonstrates the feasibility, compatibility and efficiency in improvement of QoS parameters. newline newline newline