Load Optimized M/M/M Queueing for Parallel Job Scheduling in Multiple Cloud Centers

Abstract

The cloud computing environment enables cloud users to execute their applications in remote data centers. Many of these applications are considered to be highly complex in nature, requiring parallel processing capabilities. Parallel job scheduling techniques, mainly focus on improving throughput or the information processed by the cloud center in a given interval of time and reducing average task waiting time. For a data center that deals with parallel jobs, it is required to design an optimal scheduler resulting in minimal utilization of memory rate for scheduling each job. In this paper, we present M/M/M Queuing System and Load Optimized (QS-LO) model for parallel job scheduling in multiple cloud centers. The QS-LO model is designed as two algorithms, namely parallel job scheduling and load balancing with aiming at improving the throughput rate and reducing the average task waiting time in multiple cloud centers. We perform experimental analysis using benchmarks and synthetic datasets to measure the performance of the proposed algorithm. The experimental results are compared with the existing parallel job schedulers in terms of the job assigned, throughput memory utilization rate and scheduler time interval. The experimental results show that the QS-LO model is able to improve the throughput rate and also reduce the average task waiting time when compared to the state-of-the-art works.

Modernization of Legacy Batch Processes: An Upgrade to Prolonged Batch Runs with Advanced Computing and Parallel Processing Techniques

“…Migrating batch processes to cloud-based platforms can provide the scalability needed to handle varying loads efficiently. Cloud services like AWS, Google Cloud, and Microsoft Azure offer auto-scaling capabilities that adjust resources based on real-time demand [3].…”