论文部分内容阅读
MapReduce是Google提出的一种并行计算框架模型,主要目的是用于大规模数据的并行计算。MapReduce是当今云计算的核心技术之一,它的开源实现Hadoop已经成功地被应用到不同的项目中。但是,当用户想使用这种并行计算框架时,却需要深入了解Hadoop的配置方式、编程API、运行方式等。本文提出增强的云化并行计算框架系统,旨在给用户提供一个并行计算集群,将并行计算能力作为服务提供给用户,同时简化用户的配置、编程、打包、上传并运行的这一系列操作。增强的并行计算框架系统还提出一套标记性语言用于协助用户编写MapReduce程序,快速建立MapReduce工作任务。
MapReduce is a parallel computing framework proposed by Google model, the main purpose is for large-scale data parallel computing. MapReduce is one of the core technologies of cloud computing today. Its open source implementation Hadoop has been successfully applied to different projects. However, when users want to use this parallel computing framework, they need to know more about how Hadoop is configured, the programming APIs, how they are run, and more. This paper proposes an enhanced cloud computing framework for parallel computing, designed to provide users with a parallel computing cluster that provides parallel computing capabilities as a service to users while simplifying the user’s configuration, programming, packaging, uploading, and running operations. Enhanced parallel computing framework system also proposed a set of mark-up language used to assist users to write MapReduce program, the rapid establishment of MapReduce tasks.