%0 Journal Article
%T E级超级计算机故障预测的数据采集方法
%A 胡维
%A 蒋艳凰
%A 刘光明
%A 董文睿
%A 崔新武&lt
%A /br&gt
%A HU Wei
%A JIANG Yanhuang
%A LIU Guangming
%A DONG Wenrui
%A CUI Xinwu
%J 国防科技大学学报
%D 2016
%R 10.11887/j.cn.201601016
%X 面向未来E级超级计算机，提出用于故障预测的数据采集框架，能够全面采集与计算结点故障相关的状态数据。采用自适应多层分组数据汇集方法，有效解决随着系统规模增长数据汇集过程开销过大的问题。在TH-1A超级计算机上的实现和测试表明，该数据采集框架具有开销小、扩展性好的优点，能够满足未来大规模系统故障预测数据采集的需求。</br>Aimed at an exascale supercomputer, an FPDC (failure prediction data collection framework) was introduced to fully collect the data related to the state of compute nodes’ health. An adaptive multi-layer data aggregation method was presented for data aggregation with less overhead. Extensive experiments, by implementing FPDC on TH-1A，indicate that the FPDC has the advantage of high efficiency and good scalability.
%K 超级计算机 故障预测 数据采集方法 数据汇集&lt
%K /br&gt
%K supercomputer failure prediction data collection method data aggregation
%U http://journal.nudt.edu.cn/gfkjdxxb/ch/reader/view_abstract.aspx?file_no=201601016&flag=1