期刊
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
卷 93, 期 -, 页码 278-289出版社
ELSEVIER
DOI: 10.1016/j.future.2018.10.046
关键词
Cloud computing; Workflow scheduling; Multi-objective optimization; Fuzzy dominance sort; HEFT
资金
- National Natural Science Foundation of China [61802185, 61272420, 61872185, 61502234]
- Natural Science Foundation of Jiangsu Province, China [BK20180470, BK20150785]
- Natural Science Foundation of Shanghai, China [16ZR1409000]
More and more enterprises and communities choose cloud computing platforms to deploy their commercial or scientific workflow applications along with the increasing popularity of pay-as-you-go cloud services. A major task of cloud service providers is to minimize the monetary cost and makespan of executing workflows in the Infrastructure as a Service (laaS) cloud. Most of the existing techniques for cost and makespan minimization are designed for traditional computing platforms which cannot be applied to the cloud computing platforms with unique service-based resource managing methods and pricing strategies. In this paper, we study the joint optimization of cost and makespan of scheduling workflows in laaS clouds, and propose a novel workflow scheduling scheme. In this scheme, a fuzzy dominance sort based heterogeneous earliest-finish-time (FDHEFT) algorithm is developed which closely integrates the fuzzy dominance sort mechanism with the list scheduling heuristic HEFT. Extensive experiments using the real-world and synthetic workflows demonstrate the efficacy of our scheme. Our scheme can achieve significantly better cost-makespan tradeoff fronts with remarkably higher Hypervolume and can run up to hundreds of times faster than the state-of-the-art algorithms. (C) 2018 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据