期刊名称：IAENG International Journal of Computer Science
印刷版ISSN：1819-656X
电子版ISSN：1819-9224
出版年度：2021
卷号：48
期号：3
语种：English
出版社：IAENG - International Association of Engineers
摘要：Cloud computing is a promising technology for providing efficient virtualized compute and storage resources to users on a pay-per-usage model. Large-scale geographically distributed data centers have been established to support the increasing demand for cloud services. Execution of dataintensive workloads is a challenging problem especially when objectives such as load balancing and energy reduction are essential to reduce cloud providers operational cost while providing high quality-of-service to users. Meantime, the high rates of data transfers result in network congestion that increases the makespan of workloads. This paper presents a novel Energy-Efficient and Bandwidth-Aware workload allocation method to run data-intensive applications on geodistributed cloud DCs. We formulated the workload allocation problem as a multi-objective optimization problem to minimize the workload makespan, data centers energy consumption, and communication network congestion overhead. We designed a meta-heuristic genetic algorithm to find a near-optimal workload allocation. Extensive simulations using synthetic and real traces showed a 32% average reduction of workload makespan and 35% average reduction in network traffic compared to benchmark allocation methods.
关键词：Green Computing;Energy Efficiency;GeoDistributed Data Centers;Task Scheduling;Deadline;Dataintensive
