Font Size: a A A

Research And Application Of Distributed Data Storage Based On MyCAT

Posted on:2017-01-22Degree:MasterType:Thesis
Country:ChinaCandidate:C WangFull Text:PDF
GTID:2308330503978267Subject:Computer technology
Abstract/Summary:PDF Full Text Request
As the flourish of the technology of Internet, cloud computing and Internet of things, the data continue to accumulate and the size of data is rising rapidly. The challenge of Big Data can be found in every field. Currently, Relational database is still the mainstream data storage method adopted by most enterprises. However, traditional centralized relational database are gradually revealed its flaws when storing and processing mass data, such as low read/write efficiency, poor scalability,which severely restrict the development of Internet applications in the age of big data. Therefore, how to solve the problem of mass data storage using relational databases, has became a hot pot in the academia and the industry.As one of the main solutions to large data storage and processing, The distributed relational database implement storage and process of big data by building a distribute cluster, it distributes the data in the original single database to multiple nodes by building a distributed cluster, and data storage and process are implemented by the collaborate of each node. However, due to the distributed storage, such operations often need to across multiple nodes. Thus it brings a lot of problems when processing data,In order to shield the complexity of data processing under distributed environments,distributed data bases always introduce middleware as proxy to integrate each node in the cluster for implementing di-stributed processing, thereby reduce the complexity of a distributed database.Therefore, this article analyzed the some of the mainstream distributed database storage schemes and existing problems of them.on this basis, proposed a set of MyCAT-based data storage design and implementation solutions by introducing MyCAT distributed database middleware, combined with a specific system storage expansion needs of a real estate transaction system.Detail work contents as follow:1) Analyzed some of the the current mainstream distributed database extension methods and compared and the strengths and weaknesses of several most well-known distributed database middleware.On this basis, this article puts emphases on technical characteristics and principles of distributed database middleware.2) Analyzed the requirements of storage expansion solution of system through combining with the project background and practical business of a real estate transaction system.3) According to above requirement analysis,detail designed a MyCAT-based distributed data storage for the estate transaction system,Put forward a design scheme with the functions of data fragmentation, cross-database queries, separated reading and writing and high availability based on MyCAT distributed distributed data storage solution.4) Implemented the MyCat-based distributed data storage methods based on the above design scheme and combined with practical dilation of data storage of the estate transaction system.The MyCAT based data storage solutions adopted MySQL as underlying physical storage instance, the intermediate layer is Mycat which integrated underlying instance,and provided easily use distributed database servicee to the application layer.It implementd distributed storage and minimize the influence on application at the same time. The feasibility of this storage solution was verified by successfully expanding the data storage capacity real estate transaction system which will be demonstrated in the last of this article.
Keywords/Search Tags:big data, distributed relational database system, middleware, mycat, mysql
PDF Full Text Request
Related items