QUEUING AND SCHEDULING IN CLOUD COMPUTING

Abstract

 Cloud computing is a prevalent computing model that facilitates on-demand services. The services are offered on a pay-per-use model. The primary objective of cloud service providers is to optimize resource utilization by minimizing execution time and costs while maximizing profits. Efficient scheduling algorithms remain a primary concern in cloud computing. This study is divided into 2 sections. In first section, A scheduling method is presented to enhance the scheduling process by using certain helpful properties of queueing theory. According to experimental findings, our approach makes better use of the global scheduler and cuts waiting times. With the use of queuing models, the performance of cloud systems is enhanced in this area. Additionally, the suggested model was tested experimentally in a number of models that outperform other models in terms of usage and reaction times. Finally, a scheduling method is developed that determines the minimum and maximum waiting times for each work in the waiting queues. The idea of grouping the tasks based on burst time is presented in the second part of the suggested technique. Traditional scheduling techniques, such First Come First Serve, Shortest Job First, EASY, Combinational Backfill, and Improved Backfill Using Balance Spiral Method, partition the workload. The suggested approach resolves this issue while concurrently reducing hunger. Additionally, this study focuses on several current scheduling algorithms and problems associated with them in cloud computing. The suggested MQS technique prioritizes dynamic work selection in order to solve the best cloud scheduling issue, and as a result, it makes efficient use of the available empty capacity.