一种远程容灾系统的设计与实现

摘要

设计和实现一种远程容灾系统,系统采用本地冗余和异地冗余相结合,服务和存储相分离的结构,采用本地控制中心作为事件处理中心和数据转发的缓存.采用同步和异步相结合的数据备份方式进行数据的备份,保证了数据的完整性和可靠性;采用分级部署失效检测的方式对系统进行检测,并利用动态的失效检测算法有效地降低了误判率和加快了服务恢复的速度.测试表明该系统的数据恢复速度和服务切换速度有明显的提高.实际应用中系统运行稳定,性能良好,达到了高级别容灾系统要求.%This paper designs and implements a remote disaster tolerant system, the system uses a structure which combines local tolerant system with remote tolerant system, separates services and storage, it uses local control center as event processing center and data forwarding cache. In addition, it schemes with synchronization and asynchronism for data backup, ensures data integrity and reliability, it uses staged deployment failure detection to detect system, and uses dynamic failure detection algorithm to reduce the erroneous decision and accelerate the speed d service restoration. Through tests, the system' s data recovery and service switching have been significantly improved. In practical application, the system runs stable, performs well and reaches a high level of disaster tolerant system requirements.