基于云计算的软件资源服务平台研究

发布时间：2018-03-27 05:13

本文选题：云计算　切入点：软件资源管理　出处：《东华大学》2012年硕士论文

【摘要】：信息科技的迅猛发展使得软件资源越来越丰富,作为软件资源之一的构件也以惊人的数量增长,为了更好的使用构件资源,我国目前已在7个城市建立了不同的构件库,但各库的管理和检索差异使资源共享存在难度。传统的软件资源管理系统主要基于集中式的管理方式,出现“资源积累多,可以利用的资源少”的难题,许多软件资源无法得到有效利用,这既浪费开发资源,又占用大量的存储资源。因此,研究可以统一管理软件资源的服务平台是急需解决的问题。云计算的出现,为分布式构件库的资源共享和综合管理及检索提供了新的思路和方法。云计算数据处理是将分布在不同地方的多台服务器资源集中起来,由一个大型的数据处理中心进行管理。本文对云计算应用于软件资源管理进行了研究,针对平台上资源管理的特点,对软件资源,主要是构件资源的管理方面做了研究。设计在实现框架方面选择开源云计算技术实现框架Hadoop,它提供一个分布式并行编程的框架,并可以简化分布式程序的开发。论文的主要研究内容和创新点如下：首先,论文介绍了云计算的特点和实现云计算的相关技术,对传统的软件资源相关理论及其管理系统进行了分析,指出了传统软件资源管理系统在软件资源共享使用时存在的不足,提出了基于云计算的软件资源服务平台框架,并对平台进行了设计和研究。其次,根据软件资源服务平台的特点以及需求,对实现云计算的技术框架,选择基于实现MapReduce并行编程模型的Hadoop云计算系统开源框架,可以降低实现成本,并使得软件资源管理和复用效率大幅提高。第三,结合软件资源分布存储以及平台的分布式处理的特点,对选择使用的刻面检索方法所采用的刻面进行了设计,并采用XML对刻面元信息进行了描述,同时对资源信息存入HBase中进行了表设计。为了加快分布式存储资源的检索效率,软件资源搜索引擎的离线爬取进行了设计和说明
[Abstract]:The rapid development of information technology makes the software resources more and more abundant. As one of the software resources, the component has also increased by an astonishing amount. In order to make better use of the component resources, China has established different component libraries in seven cities at present. However, the differences in the management and retrieval of each database make it difficult to share resources. The traditional software resource management system is mainly based on the centralized management mode. Many software resources can not be used effectively, which not only wastes the development resources, but also takes up a lot of storage resources. Therefore, it is urgent to solve the problem of researching the service platform that can manage the software resources in a unified way. It provides new ideas and methods for resource sharing, comprehensive management and retrieval of distributed component libraries. Cloud computing data processing is a collection of server resources distributed in different places, which is managed by a large data processing center. In this paper, the application of cloud computing in software resource management is studied. In view of the characteristics of resource management on the platform, this paper studies the management of software resources, mainly component resources, and designs an open source cloud computing implementation framework, Hadoop, which provides a framework for distributed parallel programming. And can simplify the development of distributed programs. The main contents and innovations of the thesis are as follows:. First of all, the paper introduces the characteristics of cloud computing and related technologies to implement cloud computing, and analyzes the traditional software resource related theory and its management system. The shortcomings of traditional software resource management system in software resource sharing are pointed out. The framework of software resource service platform based on cloud computing is put forward, and the platform is designed and studied. Secondly, according to the characteristics and requirements of software resource service platform, the open source framework of Hadoop cloud computing system based on implementing MapReduce parallel programming model can reduce the cost of implementing cloud computing. The efficiency of software resource management and reuse is greatly improved. Thirdly, according to the characteristics of distributed storage of software resources and distributed processing of the platform, the facet used in the selected facet retrieval method is designed, and the facet meta-information is described by XML. In order to speed up the retrieval efficiency of distributed storage resources, the off-line crawling of software resources search engine is designed and explained.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP311.52

【引证文献】