基于可信云计算的非集中式元数据存储结构优化  

Decentralized Metadata Storage Structure Optimization Based on Trusted Cloud Computing

作　　者：杨阔 李海涛 张雪梅 YANG Kuo;LI Hai-tao;ZHANG Xue-mei(Baoshen Railway Group Co.,Ltd.,Locomotive Branch,Shenmu,Shaanxi 719300,China;Anhui Anwei Technology Co.,Ltd.,Hefei,Anhui 230000,China)

机构地区：[1]包神铁路集团机务分公司,陕西神木719300 [2]安徽安为科技有限公司,安徽合肥230000

出　　处：《计算技术与自动化》2023年第1期183-187,共5页Computing Technology and Automation

摘　　要：在非集中式元数据的存储调用过程中,在硬件结构复杂、负载量大的情况下,参数设置优化步骤复杂,导致带宽达不到存储应用的期望,为此设计了一种基于可信云计算的非集中式元数据存储结构优化方法。设计非集中式元数据存储结构总体框架,计算最优元数据存储结构,通过能量检测对存储分区进行筛选,建立基于可信云计算的存储度量模型,设计可信度量存储报告机制,引入行列混合存储,设计内部结构分布图,实现结构优化。测试结果显示:在不同优化方法下,设计的优化方法所得到的存储结构下并行读写的写入带宽不会受到服务器数量变化的影响,高负载读写下的聚集带宽也较优。In the storage and calling process of decentralized metadata,when the hardware structure is complex and the load is large,the optimization steps of parameter setting are complex,resulting in the bandwidth not meeting the expectations of storage applications.Therefore,a decentralized metadata storage structure optimization method based on trusted cloud computing is designed.This paper Designs the overall framework of decentralized metadata storage structure,calculates the optimal metadata storage structure,screens the storage partition through energy detection,establishs the storage measurement model based on trusted cloud computing,designs the trusted measurement storage reporting mechanism,introduces row column hybrid storage,designs the internal structure distribution map,and realizes the structure optimization.The test results show that under different optimization methods,the write bandwidth of parallel reading and writing under the storage structure obtained by the designed optimization method will not be affected by the change of the number of servers,and the aggregation bandwidth under high load reading and writing is also better.

关 键 词：可信云计算 存储结构 存储框架 并行读写 行列混合存储结构 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]