Weighted Scheduling of Mixed Gang Jobs on Distributed Resources
Georgios L. Stavrinides, Helen D. Karatza · 2020
Over the recent years, due to the rapid growth of the Internet of Things (IoT) and its related applications, the efficient utilization of distributed computing resources is crucial. The selection of appropriate resource scheduling techniques for workloads with specific characteristics in order to achieve low response times, is still an open research challenge. In this paper, we focus on scheduling gang-type applications with the aim of achieving low response times, while at the same time considering fairness in job service. We examine complex gang jobs and we study the impact of their characteristics on the performance of various gang scheduling techniques. The simulation results show that the performance of the scheduling approaches is dependent on the characteristics of the workload.