摘要
Linux virtual server是典型的服务器机群结构,通过双前端负载分配器和后端服务器池提供高可用服务。当后端服务器池发生故障时,故障服务器的工作量将通过负载分配算法重新分配到其他可用的服务器上,结果导致服务器池负担变重,进而可用服务器的故障概率增加。该文设计并实现了高可用N+1冗余结构模型系统,改进了典型的服务器池的上述缺陷。实验表明模型有着极低的资源占用率和非常短的故障切换时间。
Linux virtual server is a typical server cluster, which provides high availability through dual load balancer and server pool. When server pool goes wrong, the workload of failure server will be redistrubuted to other available servers by using load distribute algorithm, as a result that server pool has heavy burden, and failure probability of available server increases greatly. High availability N＋I redundancy structure model of server pool is designed and implemented, in order to solve the problems. Experimental results present that the model costs a few system resources and little time in a period of one time takeover.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第15期112-114,117,共4页
Computer Engineering
关键词
高可用性
服务器池
服务切换
心跳技术
high availability
server pool
service takeover
heart beat
