Priority based job scheduling technique that utilizes gaps to increase the efficiency of job distribution in cloud computing

Abstract

A growing number of services, accessible and usable by individuals and businesses on a pay-as-you-go basis, are being made available via cloud computing platforms. The business services paradigm in cloud computing encounters several quality of service (QoS) challenges, such as flow time, makespan time, reliability, and delay. To overcome these obstacles, we first designed a resource management framework for cloud computing systems. This framework elucidates the methodology of resource management in the context of cloud job scheduling. Then, we study the impact of a Virtual Machine’s (VM’s) physical resources on the consistency with which cloud services are executed. After that, we developed a priority-based fair scheduling (PBFS) algorithm to schedule jobs so that they have access to the required resources at optimal times. The algorithm has been devised utilizing three key characteristics, namely CPU time, arrival time, and job length. For optimal scheduling of cloud jobs, we also devised a backfilling technique called Earliest Gap Shortest Job First (EG-SJF), which prioritizes filling in schedule gaps in a specific order. The simulation was carried out with the help of the CloudSim framework. Finally, we compare our proposed PBFS algorithm to LJF, FCFS, and MAX–MIN and find that it achieves better results in terms of overall delay, makespan time, and flow time.