Hadoop生态体系安全框架综述  

An Overview of Hadoop Security Framework

作　　者：陈玺[1,2] 马修军[1,2] 吕欣[3] 

机构地区：[1]北京大学信息科学技术学院,北京100871 [2]机器感知与智能教育部重点实验室(北京大学),北京100871 [3]国家信息中心博士后科研工作站,北京100045

出　　处：《信息安全研究》2016年第8期684-698,共15页Journal of Information Security Research

摘　　要：Hadoop项目已经成为最流行的开源云计算和大数据分析框架,同时其安全机制也受到越来越多的重视,从设计原则、系统架构、主要威胁、安全机制、设计挑战等方面对Hadoop的安全框架进行综述,其中包括Hadoop开源项目的安全机制及企业级安全解决方案.学术界对Hadoop安全机制的研究也很活跃,提出了可信平台、加密演算法、混合加密算法、三重数据加密算法、并行加密等多种方案.Hadoop生态系统涉及很多组件,不同组件有不同的安全策略,当前的研究热点是在提升细粒度、高度模块化、可扩展等目标的同时,兼顾考虑性能、开销、易用性等问题.Hadoop has become the most popular analysis framework which is used to open source of cloud computing and big data,at the same time,its security mechanism has also been treated as a critical item.This paper first gives an overview on Hadoop design principle,architecture,major threat,security mechanism,and design challenge which include the security solutions of enterprise and the security mechanism of Hadoop.The research on security mechanism of Hadoop are active,which present trusted platform,encryption algorithm,mixed encryption algorithm,TDEA and parallel encrypt algorithm and so on.Since Hadoop ecosystem involves many components,and different components have different security mechanisms,the current research focus is to enhance fine-grained,highly modular,expandability and other aspects,while taking the performance,cost,usability and other issues into account.

关 键 词：HADOOP 大数据 安全机制 安全组件 细粒度 高度模块化 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]