基于数据容灾系统的服务漂移实现  

Implementation of Service Migration Based on Data-Level Disaster Recovery System

作　　者：向坤[1] 刘晓洁[1] 赵奎[1] 李峰[1] 

机构地区：[1]四川大学计算机学院网络安全实验室,四川成都610065

出　　处：《计算机技术与发展》2010年第7期152-155,共4页Computer Technology and Development

基　　金：国家高技术产业示范工程(I2006J1780);信息产业部电子发展基金(I2007J292)

摘　　要：传统的数据容灾技术只能提供数据的实时备份与恢复,无法保证服务的连续性。服务容灾技术可以在服务级别上保证业务连续性和数据一致性,但系统设计往往过于复杂。一种基于传统数据容灾系统的服务漂移方案可以有效地解决该问题。该方案在客户服务器出现故障后,使用另一台拥有相同服务配置的镜像服务器,通过iSCSI协议和存储服务器上的逻辑卷镜像建立链接,并基于此虚拟逻辑卷继续提供服务。待客户服务器故障修复后,再切换回本服务器提供服务。这种方案可以使传统的数据容灾系统实现服务级容灾,进一步提高企业关键业务的容灾能力。Traditional data-level disaster recovery technology can provide data-level real-time backup and restoration,while still have limitations in providing service continuity.The service-level disaster recovery technology can provide service continuity and data consistency in service-level,but often requires a complicated system design.An implementation of service migration based on the traditional remote disaster recovery system can solve all these problems.When the client server suffered from disasters and stopped providing services,another image server with the same software configuration use the iSCSI protocol to initiate a connection to the storage server and established an iSCSI based remote logic volume.With the virtual logic volume,the image server can keep providing services and ensure the data consistency.When the client server restored from the malfunction,the system can switch from the migrating state to the original state,and the client server can keep providing services.With this implementation the traditional remote disaster recovery system can provide service-level data consistency and service continuity,which will further improve the companies＇ disaster recovery capabilities of key business.

关 键 词：数据容灾 服务容灾 服务漂移 服务连续性 数据一致性 ISCSI 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]