基于“军卫一号”数据库的DataGuard数据容灾方案设计与实现  

Design and implementation of Data Guard data disaster recovery scheme based on No.1 Military Medical Project Database

作　　者：王兴强[1] 侯玉玲[2] 刘长兴[1] 刘国伟[1] 

机构地区：[1]济南军区总医院信息科,济南250031 [2]山东省军区济南第三干休所,济南250014

出　　处：《医疗卫生装备》2016年第5期52-54,共3页Chinese Medical Equipment Journal

基　　金：国家自然科学基金(81502896);山东省自然科学基金(BS2015YY022)

摘　　要：目的 :对Oracle数据库的Data Guard机制进行了研究,用于"军卫一号"数据库的异地容灾。方法 :采用Oracle数据库的Data Guard机制,建立"军卫一号"数据库的主库和备库。主库设置为强制记录日志模式,备库增添4个联机日志文件并启动到MOUNT状态。在"军卫一号"系统中进行操作,主库将事务记录在数据库联机日志文件中,实时或定时传送给备库,备库通过增添的4个联机日志文件接收,并读取数据库日志文件中主库提交的事务,依据数据重做机制,在备库中重新执行。结果:"军卫一号"数据库主库出现故障,引起异常宕机无法恢复,备库切换成主库模式,宕机重启,"军卫一号"系统可正常运行,数据零丢失。结论:该方案实现"军卫一号"数据库实时同步备份,数据零丢失,操作简单,数据库容灾恢复时间短,可行性强,恢复后的数据库经严密测试,系统运行平稳。Objective To study DataGuard mechanism of Oracle database for disaster recovery of ＂No.1 Military Medical Project＂ database. Methods The primary database and standby database of ＂No. 1 Military Medical Project＂ were established by DataGuard mechanism of Oracle database. The primary database was set to force logging mode, and the standby database added four online log files and was started to the mount state. In ＂No.1 Military Medical Project＂, transactions were recorded in online log files of the primary database, and transferred to the standby database at real time or periodically. The standby database received and read the transactions through the four added online log files, and then the transactions were re-executed according to data redo mechanism. Results The primary database exception downtime could not be restored for failure, and ＂No.1 Military Medical Project＂ could run normally with no data loss, after the standby database was switched into the primary mode and restarted.Conclusion This scheme is characteristic with real-time backup, zero data loss, simple operation and short disaster recovery, and the system runs smoothly with strong feasibility.

关 键 词：军卫一号 DATAGUARD 数据库 

分 类 号：R82[医药卫生—临床医学] TP311.13[自动化与计算机技术—计算机软件与理论] TP309[自动化与计算机技术—计算机科学与技术]