RESEARCH ON HIGH AVAILABILITY SERVER POOL MODEL BASED ON N+1 REDUNDANCY STRUCTURE
(0) A typical high availability server cluster is composed of dual load balancer and server pool. It can avoid single failure point of the load balancer and hide some server failure in server pool. But it brings the cost of more loads for each normal sever to tolerate fault of the typical server pool. The workload of the server with failure is distributed to other servers in the server pool for continuing service, thus aggravate the burden of the server pool, and increase the probability of server failure. In order to solve the problem, high availability server pool system based on N+1 redundancy server is designed and implemented in this paper. Multi-timer in a single process, quick takeover with redundancy server and diversiform check mechanism for server pool in two ways are discussed. The experimental results have shown that the model cost a little system resource and few time in a period of one time takeover, and satisfy the necessity of the server pool on high availability.
High Availability Server Pool Service Takeover Heartbeat Mechanism
XIAOGUANG DING LEI SHI ZHIMIN GU
School of Computer Science and Technology, Beijing Institute of Technology, Beijing 100081, China School of Information Engineering, Zhengzhou University, Zhengzhou 450052, China
国际会议
开封
英文
450-456
2006-10-15（万方平台首次上网日期，不代表论文的发表时间）