Research On Web Services Transmission Solution For Vast Data

Web Service is a common service-oriented architecture, but SOA system has a bottleneck of low efficiency in transmitting large amounts of data. How to improve transmission efficiency from database to client, should be considered seriously on system architecture design phrase. Improve transmission efficiency can enhance user experience, and respond client requests quickly when get large amounts of data.This dissertation summarize the state of Web Service transmission solution research, include typical and optimization strategy both at home and abroad. For example: compress data, reduce multiple calls, optimize XML parser, simplify XML tags, and cache mechanism. Then, based on Web Service transfer principle, the data transmission efficiency is analyzed. Combining with the characteristics of different data, propose a solution that different compression methods are used for different data to improve the efficiency of data transmission. The experiment results show that the selective compress solution can improve data transmission efficiency in Web Service. But for some higher compressed data, this method is unsatisfactory. In order to solve the problem, this dissertation brings up a solution with multi-threading. The new solution based on modular classification and multi-threading transfer principle, in order to enhance user experience satisfaction, propose for the transmission of large amounts of data, how to classify modulus according to the data characteristics, and retrieve data through multi-threading respectively, show the data in batches to the user step by step. The experimental results show that modular classification and multi-threading transmission on network can respond client requests quickly.On the other hand, with critical applications for the huge amount of data are more and more, the traditional centralized database has been overwhelmed, so widely used in distributed database has already begun. Based on distributed architecture and SOA, propose a query solution that modular classification combined with Web Service. This solution use multi-threading to obtain their respective data from corresponding distributed database. The experimental results show that it can improve transmission efficiency, and increase the scalability of system architecture.