通信学报 ›› 2018, Vol. 39 ›› Issue (10): 97-117.doi: 10.11959/j.issn.1000-436x.2018213

• 学术论文 • 上一篇    下一篇

storm平台下工作节点的内存电压调控节能策略

蒲勇霖1(),于炯1,2,鲁亮2,卞琛2,廖彬3,李梓杨1,4   

  1. 1 新疆大学软件学院,新疆 乌鲁木齐 830008
    2 新疆大学信息科学与工程学院,新疆 乌鲁木齐 830046
    3 新疆财经大学统计与信息学院,新疆 乌鲁木齐 830012
    4 新疆润物网络有限公司,新疆 乌鲁木齐 830002
  • 修回日期:2017-12-25 出版日期:2018-10-01 发布日期:2018-11-23
  • 作者简介:蒲勇霖(1991?),男,山东淄博人,新疆大学博士生,主要研究方向为内存计算、流式计算、绿色计算等。|于炯(1964?),男,新疆乌鲁木齐人,博士,新疆大学教授、博士生导师,主要研究方向为并行计算、分布式系统、绿色计算等。|鲁亮(1990?),男,新疆乌鲁木齐人,新疆大学博士生,主要研究方向为分布式系统、内存计算、绿色计算。|卞琛(1981?),男,江苏南京人,博士,新疆大学副教授,主要研究方向为分布式系统、内存计算、绿色计算等。|廖彬(1986?),男,新疆乌鲁木齐人,博士,新疆财经大学副教授、硕士生导师,主要研究方向为分布式系统、数据库理论与技术、绿色计算等。|李梓杨(1993?),男,新疆乌鲁木齐人,新疆大学硕士生,主要研究方向为流式计算、内存计算等。
  • 基金资助:
    国家自然科学基金资助项目(61262088);国家自然科学基金资助项目(61462079);国家自然科学基金资助项目(61562086);国家自然科学基金资助项目(61363083);国家自然科学基金资助项目(61562078);国家科技部科技支撑项目(2015BAH02F01);新疆维吾尔自治区研究生科研创新项目(XJGRI2016028)

Energy-efficient strategy for work node by DRAM voltage regulation in storm

Yonglin PU1(),Jiong YU1,2,Liang LU2,Chen BIAN2,Bin LIAO3,Ziyang LI1,4   

  1. 1 School of Software,Xinjiang University,Urumqi 830008,China
    2 School of Information Science and Engineering,Xinjiang University,Urumqi 830046,China
    3 School of Statistics and Information,Xinjiang University of Finance and Economics,Urumqi 830012,China
    4 Xinjiang Runwu Network Limited Company,Urumqi 830002,China
  • Revised:2017-12-25 Online:2018-10-01 Published:2018-11-23
  • Supported by:
    The National Natural Science Foundation of China(61262088);The National Natural Science Foundation of China(61462079);The National Natural Science Foundation of China(61562086);The National Natural Science Foundation of China(61363083);The National Natural Science Foundation of China(61562078);The Science and Technology Support Projects of Ministry of National Science and Technology(2015BAH02F01);The Research Innovation Project of Graduate Student in Xinjiang Uygur Autonomous Region(XJGRI2016028)

摘要:

针对传统大数据流式计算平台节能策略并未考虑数据处理及传输的实时性问题,首先根据数据流处理的特点与storm集群的结构,建立有向无环图、实例并行度、任务资源分配与关键路径模型。其次结合拓扑执行关键路径与系统性能的分析,提出一种 storm 平台下工作节点的内存电压调控节能策略(WNDVR-storm,energy-efficient strategy for work node by dram voltage regulation in storm),该策略针对是否有工作节点位于拓扑执行的非关键路径上设计了 2 种节能算法。最后根据系统数据处理及传输的制约条件确定工作节点 CPU 使用率与数据传输量的阈值,并对选定的工作节点内存电压做出动态调整。实验结果表明,该策略能有效降低能耗,且制约条件越小节能效率越高。

关键词: 大数据, 流式计算, storm, 关键路径, 内存电压, 能耗

Abstract:

Focused on the problem that traditional energy-efficient strategies never consider about the real time of data processing and transmission,models of directed acyclic graph,parallelism of instance,resource allocation for task and critical path were set up based on the features of data stream processing and the structure of storm cluster.Meanwhile,the WNDVR-storm (energy-efficient strategy for work node by dram voltage regulation in storm) was proposed according to the analysis of critical path and system performance,which included two energy-efficient algorithms aiming at whether there were any work nodes executing on the non-critical path of a topology.Finally,the appropriate threshold values fit for the CPU utilization of work node and the volume of transmitted data were determined based on the data processing and transmission constraints to dynamically regulate the DRAM voltage of the system.The experimental result shows that the strategy can reduce energy consumption effectively.Moreover,the fewer constraints are,the higher energy efficiency is.

Key words: big data, stream computing, storm, critical path, DRAM voltage, energy consumption

中图分类号: 

No Suggested Reading articles found!