分布式环境下云存储的数据存储策略研究

发布时间：2018-01-03 22:34

本文关键词：分布式环境下云存储的数据存储策略研究　出处：《武汉理工大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着信息技术的发展以及互联网的普及,全球数据总量日益膨胀,已然进入以数据为中心的时代。特别是在Web2.0时代,新的网络服务模式促使用户的上网行为从信息消费者向信息生产者转变,大量的社交类、论坛型网站、电子商务网站等将产生爆炸式规模的数据。这些数据以非结构化数据类型为主。同时,随着产业信息化进程的推进以及企业业务的发展,全世界每天又将有大量的数据被创造或被复制,例如电信账单数据、金融交易数据以及卫星探测数据等等。一方面,全球数据量是如此庞大并持续快速增长,而另一方面,大量廉价的存储设备空间并未被有效利用。云存储的目的之一正是将这些闲置的存储空间虚拟化统一以适应大量数据的存储。海量数据的存储越来越成为许多数据密集型企业面临的问题和挑战,包括结构化信息,半结构化信息和非结构化信息的存储。目前海量数据的存储也变得越来越普及,但是对存储的要求也变得越来越高。存储的智能化便是其中一个较为实际的要求。尽管目前对智能化存储还没有一个统一的定义,智能化存储与许多方向也有着关联,它体现在存储的每一个细节上(例如,存储网络的智能化、语义检索的智能化、存储服务器的智能化等等),以达到卓越的可用性、高可行性和管理的智能化并对应用系统和用户透明。随着语义网技术的发展,语义存储系统的提出为存储智能化提供了可能。本文介绍了分布式环境及其相关的分布式存储技术,分析了当前分布式存储系统所面临的问题,然后介绍了语义网相关的技术包括本体的相关理论、资源描述框架等,并通过本体建模工具Protege为分布式存储系统资源分类领域建立本体模型,在该本体模型上对存储资源对象进行相应的资源分类和规则推理以实现语义路由,以兼容目前的各类分布式存储文件系统又能便于扩展,克服了目前系统适合单一场景、存储设备利用不充分的缺陷。
[Abstract]:With the development and popularization of information technology and the Internet, the total global data expansion, has entered into a data centric era. Especially in the Web2.0 era, a new network service model to the user's online behavior change information from consumers to producers of information, a large number of social forums, websites, e-commerce websites will the explosion type scale data. These data based on non structured data types. At the same time, along with the advancement of information industry and the development of businesses, all over the world every day will have a large amount of data to be created or copied, such as telecom billing data, financial transaction data, satellite data and so on. On the one hand, the global data the amount is so large and sustained rapid growth, on the other hand, a large number of cheap storage space has not been effectively used. One purpose of cloud storage is These idle storage spaces are virtualized to adapt to the storage of a large number of data.
More and more become the massive data storage data intensive enterprise faces many problems and challenges, including structured information, semi-structured and unstructured information storage. The storage of massive data has become more and more popular, but the storage requirements have become increasingly high. Intelligent storage is one the practical requirements of intelligent storage. Although there is not a uniform definition, intelligent storage and many directions also have relevance, it is reflected in every detail of the storage (e.g., intelligent storage, intelligent network, semantic retrieval intelligent storage server, etc.) to to achieve superior usability, high feasibility and intelligent management and application system and users. With the development of semantic web, the semantic storage system provides the possibility for intelligent storage.
This paper introduces a distributed environment and the distributed storage technology, analyzes the current problems of distributed storage system, and then introduces the related theories including the ontology technology related to the semantic web, resource description framework, and established ontology model for classification of resources in the field of distributed storage system through the ontology modeling tool Protege, in the ontology model the storage resource object resources are classified and the corresponding inference rules to achieve semantic routing, which is compatible with all kinds of distributed storage file system at present and can facilitate the expansion, overcome the system for a single scene, defects do not make full use of the storage device.

【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP333;TP391.1

【参考文献】