任丰玲, 于凯, 陈威, 于炯. 云环境下基于关键路径划分集群的调度算法[J]. 微电子学与计算机, 2018, 35(3): 42-46.
引用本文: 任丰玲, 于凯, 陈威, 于炯. 云环境下基于关键路径划分集群的调度算法[J]. 微电子学与计算机, 2018, 35(3): 42-46.
REN Feng-Ling, YU Kai, CHEN Wei, YU Jiong. Partitioning Cluster Based on the Critical Path of Multiple DAGs Scheduling[J]. Microelectronics & Computer, 2018, 35(3): 42-46.
Citation: REN Feng-Ling, YU Kai, CHEN Wei, YU Jiong. Partitioning Cluster Based on the Critical Path of Multiple DAGs Scheduling[J]. Microelectronics & Computer, 2018, 35(3): 42-46.

云环境下基于关键路径划分集群的调度算法

Partitioning Cluster Based on the Critical Path of Multiple DAGs Scheduling

  • 摘要: 通过对云环境下多DAG工作流调度算法进行研究, 提出基于关键路径划分集群的多DAG工作流调度算法.该算法充分考虑了DAG工作流的截止期限, 设计了新的综合评价指标, 基于关键路径划分集群, 在保证DAG工作流在截止期限内完成的前提下, 提高多个DAG工作流之间的公平程度, 缩短DAG工作流的延迟执行时间, 降低DAG工作流的执行时间跨度.实验表明, 算法在公平性、延迟执行时间、执行时间跨度这三方面都有所改善.

     

    Abstract: Through the analysis of multiple DAGs workflow scheduling on the cloud, we propose multiple DAGs workflow scheduling based on critical path partitioning cluster algorithm. it takes into account the deadline of DAG workflow, designed a new comprehensive evaluation index, Partitioning Cluster Based on the Critical Path, ensuring DAGs completed before the deadline, improving the fairness between multiple DAGs, shortening the delay of execution time, reducing execution time span. Experiments show that this algorithm can improved in the aspect of fairness, time delay and the execution time span.

     

/

返回文章
返回