分布式文件系统数据读取方法研究与探讨  

Research and Discussion on Data Reading Method of Distributed File System

作　　者：陈佳[1,2] 张冰[2] 朱志宇[2] CHEN Jia;ZHANG Bing;ZHU Zhiyu(School of Internet of Things and Artificial Intelligence,Wuxi Vocational College of Science and Technology,Wuxi,China,214028;School of Electronics&Information,Jiangsu University of Science and Technology,Zhenjiang,China,212003)

机构地区：[1]无锡科技职业学院物联网与人工智能学院,江苏无锡214028 [2]江苏科技大学电子信息学院,江苏镇江212003

出　　处：《福建电脑》2024年第12期63-65,共3页Journal of Fujian Computer

摘　　要：分布式文件系统中,客户端频繁访问数据节点会导致网络开销增大。为减少客户端对地址节点的访问次数,提升数据读取效率,本文提出一种基于D-S证据理论的数据读取方法。利用D-S证据理论进行不确定信息预测,来获得下一次需访问的数据节点地址。实验结果为:初期预测准确率较低,但随着读取次数增加,预测准确率显著提升,读取时间大幅减少,最终接近本地数据读取速度。In distributed file systems,frequent client access to data nodes can lead to increased network overhead.To reduce the number of client visits to address nodes and improve data reading efficiency,this paper proposes a data reading method based on D-S evidence theory.By utilizing the D-S evidence theory for uncertain information prediction,the address of the next data node to be accessed can be obtained.The experimental results show that the initial prediction accuracy is low,but as the number of reads increases,the prediction accuracy significantly improves,the reading time decreases significantly,and ultimately approaches the local data reading speed.

关 键 词：D-S证据理论 分布式文件系统 数据读取 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]