Apache Hadoop是一种广泛使用的分布式系统基础架构,它实现了一个分布式文件系统(HDFS),并假定系统中的数据节点是同构的。当云系统向上扩展时,数据节点很可能变得异构。而绝大多数的研究是为了提高Hadoop在异构环境下的性能,很少注意到数据安全的改进。文章提出的Sec HDFS数据分配方案通过秘密共享技术来提高异构Hadoop系统中数据存储的安全性,可在提高安全性的同时保持系统的性能。
Apache Hadoop is a widely used distributed system infrastructure, which implements a distributed file system (HDFS) and assumes that the data nodes in the system are homogeneous. When the cloud system expands, the data nodes are likely to become heterogeneous. The vast majority of the research is to improve the performance of Hadoop in heterogeneous environment, while little to improve the data security. In this paper, the SecHDFS data allocation scheme is proposed to improve the security of data storage in a heterogeneous Hadoop system by secret sharing technology, which can improve the security while maintaining the system performance.