基于Hadoop的云计算数据主动销毁研究  

Active Data Deletion of Cloud Computing based on Hadoop

作　　者：杨榆[1,2] 雷敏[1,2] 周椿入 吕瀚森 何佳羲 

机构地区：[1]北京邮电大学信息安全中心,北京100876 [2]灾备技术国家工程实验室,北京100876

出　　处：《成都信息工程学院学报》2015年第3期230-233,共4页Journal of Chengdu University of Information Technology

摘　　要：云安全是当前云计算研究的重点方向之一,云数据在云存储服务器端删除后的残留问题是云安全主要问题之一。针对云服务器端数据删除的残留问题,提出一种主动销毁机制(Active Data Deletion,ADD),该机制结合数据覆写算法以改进原有数据删除机制。经过仿真验证,ADD机制能有效销毁原始数据。同时针对不同的安全需求采用了不同复杂度的覆写算法,并测试了不同算法的性能。The security of Cloud Computing becomes the current research focus of this field. The Cloud data residue is one of the most serious problems of cloud security. Combined with data overwrite algorithms,this paper improves the mechanism of deleting data in HDFS and put forward a new mechanism,namely the Active Data Deletion（ ADD） mechanism,which overwrites the original data before deleting it. Tested on the full distributed cluster of Hadoop,it showed that optimized HDFS based on ADD can destruct the block effectively. Moreover,the mechanism can balance the safety and performance requirement by using different overwrite algorithm according to different safety requirement.

关 键 词：云计算 数据残留 主动销毁 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]