An Investigation on Scheduling Policies for Cloud-based Software Systems

Background: The rapid diffusion of cloud computing technology has been a focus of interest for enterprises due to its higher scalability and availability and greater elasticity. Nevertheless the limited scheduling mechanisms for running applications in the cloud have been a major challenge. Aim: This project introduces an effective scheduling algorithm, which attempts to maximize cloud resources utilization, improve the computation ratio, and reduce makespan, overhead and delay in a cloud-based software system. Method: (1) Analyze different scheduling algorithms which can be adopted in cloud-based systems and simulate these algorithms using CloudSim. (2) Evaluate these algorithms performance and determine both of their advantage and disadvantage. (3) Propose an improved scheduling algorithm or policy and verify the proposed algorithm in CloudSim as well as extend CloudSim. (4) Study MapReduce framework and grasp its operating principles. (5) Analyze MapReduce scheduling algorithms. (6) Propose and optimize an efficient scheduling algorithm on MapReduce. Conclusion: Proposed scheduling policies should effectively in improving the number of completed tasks, reduce costs, and promote the development of scheduling environment. In the premise of processing accuracy, improved MapReduce scheduling policy can improve the utilization rate of resources, reduce workloads of nodes and optimize management of resources.