信息网络安全 ›› 2015, Vol. 15 ›› Issue (8): 35-40.doi: 10.3969/j.issn.1671-1122.2015.08.006

• 技术研究 • 上一篇    下一篇

一种基于多核集群的海量数据加密存储策略研究

武艳娜(), 赵泽茂, 胡晶芳, 梁斌   

  1. 杭州电子科技大学通信工程学院,浙江杭州 310018
  • 收稿日期:2015-03-09 出版日期:2015-08-01 发布日期:2015-08-21
  • 作者简介:

    武艳娜(1987-),女,河南,硕士研究生,主要研究方向:信息安全与密码学;赵泽茂(1965-),男,四川,教授,博士,主要研究方向:信息安全与密码学;胡晶芳(1989-),女,浙江,硕士研究生,主要研究方向:信息安全;梁斌(1990-),男,江西,硕士研究生,主要研究方向:信息安全。

  • 基金资助:
    浙江省自然科学基金[Y1100818]

Massive Data Encryption Storage Strategy Based on Multi-core Cluster

WU Yan-na(), ZHAO Ze-mao, HU Jing-fang, LIANG Bin   

  1. School of Communication Engineering, Hangzhou Dianzi University, Hangzhou Zhejiang 310018, China
  • Received:2015-03-09 Online:2015-08-01 Published:2015-08-21

摘要:

随着互联网和云计算技术的快速发展,云计算中海量用户隐私和数据的安全问题已经越来越引起人们的关注,而加密是保护数据安全的首选方法。为了快速而有效地实现海量数据的安全存储,文章在研究多核处理器集群在海量数据加密存储中的应用的基础上,提出在多核Linux集群上使用混合OpenMP & MPI技术来处理海量数据的解决方案,为多核集群提供在集群节点内部以及节点之间的并行策略,实现海量数据并行处理的层次结构化。同时采用并行虚拟文件系统PVFS2作为多核集群的文件系统,利用其并行I/O的特点实际解决集群系统中海量数据并行存储的问题。文章在多核集群上开发了一个海量数据加密与存储系统摸型。实验结果表明,该系统模型不仅能够有效实现海量数据的存储,还能为海量数据的安全起到一定的保护作用。

关键词: 多核集群, 海量数据, 加密存储, OpenMP, MPI, PVFS2

Abstract:

With the rapid development of the Internet and the cloud computing technology, the security problem of massive user privacies and data has caused more and more attentions, and encryption is the first choice to protect the data. In order to achieve the secure storage of massive data rapidly and effectively, on the basis of the research on the application of multi-core processor cluster in the massive data encryption storage, this paper proposes a scheme of dealing with massive data by using OpenMP&MPI hybrid technology on the multi-core Linux cluster, providing the parallel strategy for the multi-core cluster inside the node or between the nodes of the cluster, which realizes the hierarchical structured parallel processing of the massive data. This paper uses the parallel virtual file system PVFS2 as multi-core cluster file system and makes use of the characteristic of the parallel I/O to actually solve the problem of parallel storage of the massive data in the cluster system. This paper proposes a model of massive data encryption storage system based on multi-core cluster. The experimental results show that the model not only can achieve the storage of massive data effectively, but also can protect massive data to some extent.

Key words: multi-core cluster, massive data, encryption storage, OpenMP, MPI, PVFS2

中图分类号: