Design And Implementation Of Storage Layer Of Content Management System

With the rapid development of the traditional Internet and the mobile communication networks, network bandwidth increase and the means of user access diversity, value-added services based on content become a focus in telecommunications business, in this case, multimedia content becomes a source of value-added. How to make the content at the right time with the correct form deliver to the right place and people become one of the core functions of many businesses.In many service platform they need to effective manage and integrate multimedia content, with multimedia content sustained rapid growth, a content management system to meet the needs of business platform that provide service relevant to content is implemented, as a unified infrastructure, integrate and reuse excellent content resources, not only can reduce overall product development costs, but also provide opportunities to build new business platform which provide content service, thereby winning advantage in the competition.Based on the above background, this thesis designs a content management system's underlying storage layer. To provide an effective solution for Content Management System (CMS) to store mass data, and let the system give fast responses to high concurrent requests, after studying the traditional storage management, distributed file systems and NoSQL, and investigating some frequent questions of applying relational database system in CMS, a design using the open software Hadoop as the underlying distributed storage infrastructures is proposed. In order to effectively organize and manage content and provide services related to content, the system adopts Java Content Repository to provide standard operation of content after analyzing some existing systems and considering the demand of the system. Above all, the CMS's underlying storage layer is designed and implemented, to meet the needs of fast reading and writing.It's found that most content management system using the database as the underlying storage, as well as writing SQL statements for data analysis through research and analysis. But the database has some shortcomings that cannot quickly read and write when facing mass data, reading often introducing additional IO, the scalability is poor, and relational table cannot flexible copy with various media under rich media environment. In examining the other storage systems such as distributed file system, NoSQL and so on, as well as their scalability, flexibility and efficiency, the system use hadoop as the underlying storage which implements a distributed storage and distributed computing framework. It has good scalability, the store caypacity can increase with the machine number, has built-in backup, and the distributed computing framework it provides can be used for media identification, retrieval and further processing. In order to be able to organize content well and the storage layer provides excellent service to the upper layer, this thesis examines the Java content repository, the specification use the tree structure is flexible, standardized content services in the specification are recommended to enterprise application to handle and management content, operation including in the specification is to define a set of standard content operating interface, the specification has been used by many major manufacturers such as IBM, ORACLE, etc.