A Research And Application Of Distributed E-commerce System Based On LBS

With the rapid development of mobile Internet,e-commerce of location-based service has been welcomed by more and more people,but the e-commerce platform background data volume is huge,the traditional centralized database can not solve the data storage.In order to solve the massive data storage of e-commerce platform,the distributed database storage scheme based on middleware is designed and implemented in this thesis.As the e-commerce system business is complex,the consistency requirements of the database are higher,so the use of relational-based distributed database storage is the most appropriate choice.Based on the distributed database designed in this thesis,this thesis also realizes the location-based service of e-commerce supported by the distributed database system.The main work of the thesis is as follows:(1)Studying the distributed database storage technology,and analying the advantages and disadvantages of each storage scheme.Firstly,the shortcomings of various fragmentation algorithms are analyzed in this thesis based on the distributed database fragmentation algorithm.Then we proposed an improved jump consistent hash algorithm-weight jump consistent hash algorithm by combining jump consistent hash algorithm proposed by Google and the CRUSH algorithm in the distributed file system.The algorithm applied to the data slice.The algorithm is designed to achieve the uniform distribution of data,taking into account the node weight as the focus,and fully meeting the requirements of the performance,balance and monotonicity of the fragmentation algorithm.(2)What is pointing out in thesis is that the high cost of fragmented connection under the ordinary connection operation after studing and analyzing the fragmented connection scheme in the distributed database.In order to reduce the transmission cost of fragmented connection in distributed database cluster environment,a fragmentation connection program based on Bloom filter is proposed.In this thesis,the program is described and analyzed in detail,and the feasibility of the scheme is proved by mathematical calculation and function model.(3)The distributed database storage scheme is realized by combining fragmentation algorithm and the middleware Mycat.Completing a synchronous backup of the database by using the open source tool Otter This thesis finally designed and realized the distributed e-commerce system based on geography service based on the requirements analysis,architecture design and implementation of the key modules of the system.At the end of this thesis,the performance,balance and monotonicity of the proposed algorithm are verified by experiments.It is concluded that the fragmentation algorithm is better in performance,and the data distribution is also uniform.At the same time,the monotonicity is better under the condition of equal weight of nodes,and the change of the weight which near the first node will cause the migration of large scale data.