Byzantine Fault Tolerance with Window Mechanism for Replicated Services
Byzantine fault is the fault that can make the components behave arbitrary and may cause disastrous results.With the increasing malicious attacks and software errors,Byzantine fault tolerance has begun to draw more attention it deserves.Previous Byzantine fault tolerant algorithms have strong assumption that all the replicas is synchronous and do not support replicated calling services,which make them not practical and not suit for new computing model such as SOA.This paper proposes a new Byzantine fault tolerant algorithm based on well-known Byzantine fault tolerant algorithm CLBFT (Castro Liskov Byzantine Fault Tolerance) for replicated services in the calling endpoint.The algorithm works in asynchronous environments and support replicated calling services.To make the algorithm more practical,we incorporates important optimization-Window mechanism,which can make the replica batch process the message that reduce the response time much more than previous algorithms.Besides non-faulty process of the algorithm,we provide the faulty handling process to make the algorithm more robust.
State machine replication replica Byzantine fault tolerance service
Liu Chen Wei Zhou
School of Electrical and Information Engineering Wuhan Institute of Technology Wuhan, China School of Computer Central China Normal University Wuhan, China
国际会议
秦皇岛
英文
1255-1258
2015-09-18（万方平台首次上网日期，不代表论文的发表时间）