云计算关键技术之分布式计算认知研究  

Research on the cognition of distributed computing——One of the key technology of cloud computing

作　　者：龚强[1,2] 

机构地区：[1]中国地震局工程力学研究所,哈尔滨150080 [2]黑龙江省有色金属地质勘查局,哈尔滨150090

出　　处：《信息技术》2014年第8期1-3,9,共4页Information Technology

基　　金：地震工程远程协同试验系统的软件研究与开发(0618001)

摘　　要：面向云计算应用领域的并行计算和分布式计算,较之以往有新的特点和实现原则。并行计算可以分为时间并行和空间并行,空间并行因其使用多个处理器并发执行计算,更具实用价值。分布式计算的工作模式是,中心服务器将一个大型计算任务分割成若干个或更多个任务单元,利用网络分派给多台计算机分别去完成,任务完成后再返回到中心服务器。分布式计算具有容错能力更强、维护更灵活、性价比更高的比较优势。CAP和BASE理论是当今互联网领域颇受关注且十分流行的分布式NoSQL的理论基础。Parallel computing and distributed computing for cloua computing appncanons nave newcharacteristics and implementation principles than in the past. Parallel computing can be divided intotime in parallel and space in parallel, space in parallel has more practical value for its use of multipleprocessors to concurrently execute computing. The work pattern of distributed computing is that the centreserver divides a large computing task into several or more task units, by the network assigns to multiplecomputers to complete respectively, and returns to the centre server after completing the mission.Distributed computing has comparative advantages including stronger fault tolerant ability, more flexiblemaintenance and higher cost performance. CAP and BASE theory which are widely concerned and verypopular in modern internet are the theoretical basis of distributed NoSQL.

关 键 词：云计算 并行计算 分布式计算 比较优势 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]