云环境中DAG调度算法的设计与实现..doc

大连理工大学本科毕业设计（论文） 云环境中DAG调度算法的设计与实现 Design and Implementation of DAG Scheduling Algorithms in Cloud Environment 学 院（系）： 计算机科学与技术学院科学与技术 xxx 学 号： xxxxxxxx 指 导 教 师： xxx 评 阅 教 师： 完 成 日 期： 大连理工大学 Dalian University of Technology 摘 要 近年来随着网格、云计算工作流等异构分布式计算技术的发展， 本文围绕多DAG共享异构分布式资源调度的若干问题展开了研究，这些问题包括：具有多优先级的多DAG调度和具有期限约束的多DAG调度吞吐量最大化、费用优化以及费用优化的公平性等。对这些问题的解决将有利于提高网格、云计算工作流等异构分布式计算系统的资源利用率、合理处理多个DAG应用之间的调度关系和有效降低用户DAG应用的费用，因此有着重要的理论意义和应用价值。 关于DAG共享异构分布式资源调度的研究关于DAG调度算法的研究。Java语言DAG静态调度算法HEFT、CPOP和LBP，还实现了侧重公平性的E-Fairness算法，最后实现了混合调度算法MMHS。在实现这些算法的基础上，还测试这些算法的相关性能，如调度时间和公平性等。同时实现了DAG调度仿真器，在仿真器基础上，可以方便地进行各种算法的研究，而且方便做算法性能的实验测试。 关键词：多DAG调度;多优先级;公平性;仿真器 Design and Implementation of DAG Scheduling Algorithms in Cloud Environment Abstract In recent years, with the grid, cloud computing workflows and other heterogeneous distributed computing technology, scheduling of multiple DAGs sharing on heterogeneous distributed resources is becoming a hot topic of concern. At present, despite about multiple DAGs sharing on Heterogeneous Distributed Resource Scheduler has made some progress, but there are still many problems to be further studied and resolved. This paper focuses on a number of issues more DAG sharing on Heterogeneous Distributed Resource Scheduling. These issues include: a multi-priority DAG scheduling and deadline constraints have multiple DAGs scheduling to maximize throughput, cost optimization and cost optimization of the fair and so on. Solving these problems will help improve grid, cloud computing resource utilization, workflow and other heterogeneous distributed computing systems, rational treatment of multiple DAGs scheduling relationship between applications and reduce user DAG application fee, so there are theoretical significance and application value. Research on the DAG share in Heterogeneous Distributed Resource Scheduling is research on D