基于文件过滤驱动的连续数据捕获技术及其容灾应用研究  

Research on the Techniques of Continuous Data Capturing Based on File System Filtering Driver and its Application on Disaster Tolerance

作　　者：王丽娜[1,2] 熊琦[2] 王德军[2,3] 何征[1,2] 潘腾飞[1,2] 

机构地区：[1]武汉大学计算机学院,湖北武汉430079 [2]空天信息安全与可信计算教育部重点实验室,湖北武汉430079 [3]中南民族大学计算机学院,湖北武汉430074

出　　处：《小型微型计算机系统》2011年第3期477-482,共6页Journal of Chinese Computer Systems

基　　金：教育部科技重点项目(108087)资助;国家"八六三"高技术研究发展计划项目(2009AA01Z442)资助;国家自然科学基金项目(60970114)资助;武汉市科技攻关项目(200810321130)资助

摘　　要：根据数据容灾技术的发展趋势,结合连续数据保护的具体目标和现有数据捕获技术的优缺点,提出一种基于文件过滤驱动的连续数据捕获方法,描述了连续数据捕获的实现流程,并详细阐述了在连续数据捕获中所涉及的IRP截获技术、内核数据传输技术,最后使用该方法构建了一个面向业务连续性的安全容灾备份软件并给出了相关测试结果.与连续快照和API原语劫持相比,该方法具有数据捕获效率高、无数据遗漏、代价较低、对上层应用影响小的特点.According to the developing trend of disaster tolerant technology and combing the detailed purpose of continuous data protecting, a file filtering driver based continuous data capturing technology is proposed first. Then the detailed working process is described, and related critical technology including IRP intercepting, captured data transferring in kernel model and continuous data caching technology is proposed. Finally, a continuous data protecting system is constructed based on this technology, and related system testing metrics are presented. Compared with continuous snapshot and na？ ve API robbing, the technology based on file filtering driver can work with high efficiency, no data lose and low cost, and will make little influence on the performance of upper alaolication.

关 键 词：连续数据捕获 WINDOWS驱动模型 内核数据传输 文件过滤驱动 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]