分布式对象共享耦合信息的一种通用容错模型  

Fault Tolerance Model for Distributed Objects to Share Coupling Information

作　　者：刘晖[1] 沈钧毅[1] 彭勤科[1] 燕彩蓉[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049

出　　处：《西安交通大学学报》2002年第12期1266-1270,共5页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目(60175015);西安交通大学IBM博通软件开发中心资助项目.

摘　　要：提出了一种分布式对象共享耦合信息的通用容错模型,它具有集中计算、多播发送、分散并行纠错的特点,通过对象间含有捎带功能的对偶心跳检测,可在规定间隔内监测出对象意外死亡、信息丢失等异常.分散并行纠错只需检查本地局部信息是否被正确耦合及传送,是一种降低纠错率以减少时间、空间开销的折衷.按照泊松分布计算的分散并行纠错率说明,对于8节点集群,当出错率为10-3时,纠错率为94 5596%,空间及时间开销仅为原来的22 22%.若采用可靠多播,分散并行纠错率恒为100%.A general, fault tolerance model for distributed objects is presented to share coupling information whose computation depends on all valid local information dynamically produced by distributed objects.Its main characteristics include centralized computation, multicast and decentralizedparallel error checking (DPEC). By means of dual heartbeats that carry certain piggyback tasks ,the crash of objects and the loss of information can be detected and corrected in the given period. DPEC only needs to check whether the local information is coupled and transmitted in a right way, i.e., only parts of the whole data are checked in parallel. It is a tradeoff between correctness probability and spatial or temporal spending. Correctness probability of DPEC is computed with Poisson process and illustrated. When the fault probability equals 10-3, the correctness probability of 8 objects is 94559 6% while the spatial and temporal spending is only 2222% of the traditional value. If reliable multicast is used, the correctness probability of DPEC keeps constant and is equal to 100%.

关 键 词：通用容错模型 集群 分布式对象 耦合信息 信息共享 纠错率 网络技术 

分 类 号：TP393[自动化与计算机技术—计算机应用技术] TP302.[自动化与计算机技术—计算机科学与技术]