基于决策树模型的分布式存储数据纠删码修复  

Research on a New Algorithm of Erasure Correcting Code for Distributed Storage Data Based on Decision Tree Model

作　　者：沈洪敏 周功建[1] SHEN Hong-min;ZHOU Gong-jian(Tan Kah Kee College,Xiamen University,Zhangzhou Fujian 363105,China)

机构地区：[1]厦门大学嘉庚学院,福建漳州363105

出　　处：《计算机仿真》2022年第6期473-477,共5页Computer Simulation

摘　　要：当今大数时代背景下,海量大数据的存储备份时刻冲击着当前先进的数据存储与纠删技术。分布式数据存储系统作为经典数据容错技术,在进行数据保障的过程中采用容错技术、多副本存储备份技术以及误码数据纠删等方式来保证数据存储的可靠性。纠删码技术以其数据存储过程中资源消耗低、可靠性高等优点在数据纠删存储领域得到了广泛应用,但是传统纠删技术依然存在数据修复速度低、修复率低等缺点。因此,结合数据决策模型提出基于决策树模型的分布式数据纠删码修复算法。算法首先建立决策树模型,然后将决策树与纠删码技术相结合建立纠删决策树模型。最后,给出了对应的数据仿真,同时实验对比可以看出提出的决策树纠删数据模型在修复速度、数据修复率、容错性等方面具有很好的有效性。In the times of the big data era, the storage and backup of massive big data always impact the current advanced data storage and erasure technology. As a classical data fault-tolerant technology, distributed data storage system adopts fault-tolerant technology, multi-copy storage and backup technology, and error data correction and deletion to ensure the reliability of data storage. Erasure code technology has been widely used in the field of data erasure storage because of its low resource consumption and high reliability in the process of data storage. However, the traditional erasure technology still has the disadvantages of low data repair speed and low repair rate. Therefore, combined with the data decision model, this paper proposes a distributed data erasure code repair algorithm based on the decision tree model. The algorithm first establishes the decision tree model, and then combines the decision tree and erasure code technology to establish the erasure decision tree model. Finally, the corresponding data simulation experiments are given. At the same time, the experimental comparison shows that the decision tree correction data model proposed in this paper has good effectiveness in repair speed and data repair rate.

关 键 词：决策树 分布式系统 数据存储 纠删码 大数据 

分 类 号：TP333[自动化与计算机技术—计算机系统结构] TN911.22[自动化与计算机技术—计算机科学与技术]