分布式计算机系统中进行容错处理是提高可靠性的主要途径。本文分析了分布式系统进行并发处理的容错设计,在基于系统软件容错处理的设计中,提出了对系统作业、任务的动态切换,并在IBM PC Omninet局网支持下的分布式处理模型中,进行了初...分布式计算机系统中进行容错处理是提高可靠性的主要途径。本文分析了分布式系统进行并发处理的容错设计,在基于系统软件容错处理的设计中,提出了对系统作业、任务的动态切换,并在IBM PC Omninet局网支持下的分布式处理模型中,进行了初步实现。展开更多
Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgrounda...Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgroundand developing status of MPI. Then on this basis it will study and analyze the functions and features of MPI ,summa-rize its insufficiencies and gives some suggestions for modification.展开更多
The bases of distributed parallel network data storage technology(DPNDS) are high speed network tech-nology, database technology, artificial intelligence and multimedia technology. It will become the foundation of mod...The bases of distributed parallel network data storage technology(DPNDS) are high speed network tech-nology, database technology, artificial intelligence and multimedia technology. It will become the foundation of mod-ern information society because of the excellent management efficiency, reliability, availability, expansibility and lowcost. In this paper, we discuss the key technology of DPNDS which includes data distributing, parallel data process-ing, data network storage, object-oriented data storage and then give some suggestions about the new research direc-tion.展开更多
The Common Object Request Broker Architecture (CORBA) is a main specification applied to the dis-tributed parallel program design at present. It includes many services, in which Event Service can provide the asyn-chro...The Common Object Request Broker Architecture (CORBA) is a main specification applied to the dis-tributed parallel program design at present. It includes many services, in which Event Service can provide the asyn-chronous communication model for the distributed parallel program design. This paper introduces the Event Service ofCORBA, then designs and implements a distributed parallel model based on Event Service. This model is made to im-plement overlap of the computing and communication , to reduce the unnecessary time for waiting and enhance theparallel effect. The paper also shows the practical application of the model with an example of the distributed parallelmatrix multiplication.展开更多
文摘Message passing Interface (MPI) is a kind of network distributed parallel computing environments whichhave been widely used on super parallel computers and networks. First,this paper describes the rssearch backgroundand developing status of MPI. Then on this basis it will study and analyze the functions and features of MPI ,summa-rize its insufficiencies and gives some suggestions for modification.
文摘The bases of distributed parallel network data storage technology(DPNDS) are high speed network tech-nology, database technology, artificial intelligence and multimedia technology. It will become the foundation of mod-ern information society because of the excellent management efficiency, reliability, availability, expansibility and lowcost. In this paper, we discuss the key technology of DPNDS which includes data distributing, parallel data process-ing, data network storage, object-oriented data storage and then give some suggestions about the new research direc-tion.
文摘The Common Object Request Broker Architecture (CORBA) is a main specification applied to the dis-tributed parallel program design at present. It includes many services, in which Event Service can provide the asyn-chronous communication model for the distributed parallel program design. This paper introduces the Event Service ofCORBA, then designs and implements a distributed parallel model based on Event Service. This model is made to im-plement overlap of the computing and communication , to reduce the unnecessary time for waiting and enhance theparallel effect. The paper also shows the practical application of the model with an example of the distributed parallelmatrix multiplication.
