期刊文献+

有控制信息的三阶段负载可任意划分应用的调度

Control Information-Aware Divisible Load Scheduling with Three Phases
下载PDF
导出
摘要 文章研究需要控制信息的负载可任意划分应用的调度问题.控制信息是除了要处理的负载之外为了完成计算所必须的辅助信息,它的规模往往是由应用的性质和要处理的负载量所共同决定的.文章通过形式化地定义控制信息而扩展了DLS模型,在此基础上分析了LIFO和FIFO调度策略,分别得到了它们的解析形式的调度方案.基于这些解析解,分析了控制信息对两种调度策略的影响,并在此基础上对它们的调度性能进行了严格的比较.分析结果表明,与没有控制信息的情况不同,在这个新的模型下FIFO策略总是优于LIFO策略. Divisible load scheduling with control information is considered in this paper. Control information is the necessary factor for the computation, except for the real load itself. This paper introduces the overhead of control information into the divisible load scheduling model. Two efficient-proved schemes. LIFO and FIFO are revised under the extended model. Closed-form solutions for the process time of these schemes are derived. Based on these expressions, the effect of control information on DLS scheduling strategies are analyzed. With the introduced control information, the application's response time doesn't reduce continuously, which is a concave function of processor number. The necessary and sufficient condition for the best processor number is derived. The performances of these schemes are rigorously compared, and it is proved that FIFO is always the better not as without control information.
出处 《计算机学报》 EI CSCD 北大核心 2007年第4期638-647,共10页 Chinese Journal of Computers
基金 国家自然科学基金(60475011)资助
关键词 负载可任意划分应用 负载可任意划分应用的调度 数据并行 负载划分 调度 divisible load application divisible load scheduling data parallel load partition scheduling
  • 相关文献

参考文献11

  • 1Bharawaj V,Ghose D,Robertazzi T G.Divisible load theory:A new paradigm for load scheduling in distributed systems.Cluster Computing,2003,6(1):7-17
  • 2Robertazzi T G.Ten reasons to use divisible load theory.Computer,2003,36(5):63-68
  • 3Beaumont O,Casanova H,Legrand A,Robert Y,Yang Y.Scheduling divisible loads on star and tree networks:Results and open problems.IEEE Transactions on Parallel and Distributed Systems,2005,16(3):207-218
  • 4Drozdowski M,Wolniewicz P.Experiments with scheduling divisible tasks in clusters of workstations//Proceedings of the Euro-Par 2000 Parallel Processing.London,2000:311-319
  • 5Barlas G.Collection-aware optimum sequencing of operations and closed-form solutions for the distribution of a divisible load on arbitrary processor trees.IEEE Transactions on Parallel and Distributed Systems,1998,9(5):429-441
  • 6Siegel H J,Armstrong J B,Watson D W.Mapping computer-vision-related tasks onto reconfigurable parallelprocessing systems.Computer,1992,25(2):54-63
  • 7Li X,Bharadwaj V,Ko C.Distributed image processing on a network of workstations.International Journal of Computers and Applications,2003,25(2):1-10
  • 8Altilar D,Paker Y.Optimal scheduling algorithms for communication constrained parallel processing//Proceedings of the Euro-Par 2002 Parallel Processing.Paderborn,2002:197-206
  • 9Ko K,Robertazzi T G.Equal allocation scheduling for data intensive applications.IEEE Transactions on Aerospace and Electronic Systems,2004,40(2):695-704
  • 10Chan S K,Bharadwaj V,Ghose D.Large matrix-vector products on distributed bus networks with communication delays using the divisible load paradigm:Performance and Simulation.Mathematics and Computers in Simulation,2001,58(1):71-92

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部