支持多代理的云存储数据完整性审计方法

An Audit Method of Data Integrity for Supporting Multiple Proxies in Cloud Computing

由于云存储服务面临许多损坏数据的风险,检验数据完整性便成为一个亟需解决的基本问题。数据持有性验证(provable data possession,PDP)是检验云存储数据完整性的重要方法。然而,在传统的PDP模型中,单审计代理易造成单点故障并且易形成性能瓶颈。为此,提出了一种支持多代理的数据完整性审计方法(multi-proxies PDP,MP-PDP)。该方法采用循环链表管理多代理节点,使用审计队列存储文件的审计任务,实现了审计任务分发、节点监控、失效节点切换和动态增加代理等功能,并且利用备份节点消除了系统的单点故障。实验结果表明,MP-PDP有效减少了文件的审计执行时间,并且能够快速增删审计代理。

Since cloud storage service faces many security risks that can damage data,checking data integrity has become increasingly urgent. Provable Data Possession（ PDP） is an important method for verifying data integrity in cloud computing. But the single proxy in the traditional PDP models easily becomes the single point of failure and catches the performance bottleneck. So we propose an improved PDP,called MP-PDP by us,for supporting mutiple proxies in cloud computing. It adopts a circular linked list and uses audit queues to store the audit tasks. It achieves such functions as assigning audit tasks,monitoring nodes,switching failed node and dynamically adding proxy and uses the backup node for eliminating the single point of failure. The experimental results indicate that MP-PDP can efficiently reduce the audit time for files and quickly add or delete the audit proxy.