高级检索
    肖 达, 舒继武, 陈 康, 郑纬民. 一个网络归档存储中实用的数据持有性检查方案[J]. 计算机研究与发展, 2009, 46(10): 1660-1668.
    引用本文: 肖 达, 舒继武, 陈 康, 郑纬民. 一个网络归档存储中实用的数据持有性检查方案[J]. 计算机研究与发展, 2009, 46(10): 1660-1668.
    Xiao Da, Shu Jiwu, Chen Kang, Zheng Weimin. A Practical Data Possession Checking Scheme for Networked Archival Storage[J]. Journal of Computer Research and Development, 2009, 46(10): 1660-1668.
    Citation: Xiao Da, Shu Jiwu, Chen Kang, Zheng Weimin. A Practical Data Possession Checking Scheme for Networked Archival Storage[J]. Journal of Computer Research and Development, 2009, 46(10): 1660-1668.

    一个网络归档存储中实用的数据持有性检查方案

    A Practical Data Possession Checking Scheme for Networked Archival Storage

    • 摘要: 在网络归档存储中,数据持有性检查(DPC)用来在实际文件访问发生之前实时地检测远端服务器是否仍然完好地持有文件.提出一个实用的DPC方案.在一个挑战-应答协议中,检查者要求服务器计算文件中若干随机指定的数据块的一个Hash值,并和对应的校验块一起返回,以此判断文件的持有性.通过这种随机抽样校验的方法,在保证足够的置信度的同时降低了持有性检查的计算和通信开销.同时提出一个基于校验块循环队列的挑战更新机制,从而允许动态增加检查者可发起的有效挑战的次数.分析表明检查者端的存储开销和检查者和服务器间的通信开销均为常数量级.测试结果表明一次置信度为99.4%的持有性检查的计算开销为1.8ms,和磁盘I/O开销相比可以忽略不计;通过避免使用公钥密码系统,将文件预处理的计算开销降低了3个数量级.

       

      Abstract: Data possession checking (DPC) is used in networked archival storage to check in real time if the remote server holds a file intact before the actual access to the file occurs. The authors present a practical DPC scheme. In a challenge-response protocol, the checker ascertains the possession of a file by asking the server to compute a hash value of some randomly appointed data blocks of the file and return it together with a corresponding verification block. With this random sampling verification method, the computational and communication overheads of possession checking are reduced while a sufficiently high confidence level is obtained. A challenge renewal mechanism based on verification block circular queue is also proposed to allow the dynamic increase of the number of effective challenges which can be issued by the checker. Analysis shows that the storage overhead on the checker side and the communications overhead between the checker and the server are constant. Experimental results show that the computational overhead of a check with a confidence level of 99.4% is 1.8ms, which is negligible compared with the cost of disk I/O; The computational overhead of file preprocessing is reduced by three orders of magnitude by avoiding using public-key cryptosystem.

       

    /

    返回文章
    返回