基于云计算平台的海量图片存储系统设计与实现

发布时间：2018-01-20 14:59

本文关键词： 云计算海量图片存储 NoSQL数据库缓存热备负载均衡　出处：《北京邮电大学》2012年硕士论文　论文类型：学位论文

【摘要】：全球数据信息量快速增长,海量数据时代正在来临。类似Facebook、淘宝等社交类和在线商城类互联网企业,随着用户量和业务的发展,企业存储系统中的图片数量呈指数级增长,如何有效的存储和管理这些海量数据成为企业面临的巨大挑战。在海量数据信息时代,传统的存储架构存在扩展性较差的缺点,在用户数量及服务性能需求提高时,就只能依靠添加高端存储设备来暂时的解决问题,这样下去,存储环境会变得越来越复杂,不但没有从根本上解决海量数据存储带来的挑战,反而提高了企业管理和运营的成本。本文以个人和企业的图片数据快速增长为背景,介绍了个人和国内外企业面对的海量数据存储挑战。在此基础上设计和实现了基于云计算平台的海量图片存储系统。该系统需要解决的若干关键问题包括：架构的高可扩展性和高可靠性、图片元数据存储和管理、图片文件存储算法和图片缓存设计等。本系统由五个主要模块构成,即：图片元数据模块、图片存储模块、图片缓存模块、负载均衡模块和Web Services模块。最后对系统的上传、更新、查询和删除模块进行了功能测试同时对系统的上传图片和读取图片模块进行了性能测试。
[Abstract]:With the rapid growth of global data information, the age of massive data is coming. With the development of users and business, Internet enterprises such as Facebook, Taobao and other social and online shopping malls are developing. The number of images in the enterprise storage system increases exponentially. How to effectively store and manage these massive data has become a huge challenge for enterprises. In the age of mass data information. The traditional storage architecture has the disadvantage of poor scalability. When the number of users and service performance requirements increase, it can only rely on the addition of high-end storage devices to solve the problem temporarily. Storage environment will become more and more complex, not only does not fundamentally solve the huge data storage challenges, but also increases the cost of enterprise management and operation. This paper is based on the rapid growth of personal and corporate picture data. This paper introduces the challenges of mass data storage for individuals and enterprises at home and abroad, and then designs and implements a mass image storage system based on cloud computing platform. Some key problems that need to be solved in this system include:. High scalability and reliability of the architecture. The system is composed of five main modules: picture metadata module, picture storage module and picture cache module. Load balancing module and Web Services module. Finally, upload and update the system. The function of query and delete module is tested and the performance of upload and read pictures module is tested.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333;TP391.41

【引证文献】