Research On Data Query Optimization On Magic Platform

As a commercial business modeling system platform, Magic platform uses zero-coding, parametric, graphical model customizing pattern, getting rid of heavy coding work, and it can guarantee the quality of software effectively, reduce software development cost and improve software development efficiency. It is a kind of flexible software production platforms. The magic platform is mainly used to develop e-government system. With the improving of government office automation and information level, e-government system has become an important part of government departments, the service type it provides is more and more, and the system itself becomes more enormous and complex. As the government's external service window, external service quality of e-government directly affect the government's credibility. The operating platform is the support part of magic platform, and it has significant impact on the performance of e-government system. It is of great importance to study the query performance of business implementation in magic platform.On the basis of the e-government system operated on magic platform and the characteristics of its business, this paper research on how to improve the performance of business operation on the platform, and mainly complete the following tasks:Firstly, we analyze the characteristics of business in business system running on the magic platform. The characteristics are the law of business operation, implementation time, the amount of generation and so on. According to research on analysis, classification and characteristics of business, we can make specific optimization better aiming at the characteristics of business running on the platform.Secondly, through analyzing the business process to determine the direction of optimization to the business system, mainly including decomposition and optimization of business process, hot data prefetching optimization, data query optimization and database connection pooling optimization. Ant colony algorithm has been applied on solving data prefetching, and has achieved some effect. This paper starts from the business characteristics running on the e-government system to improve the ant colony algorithm specifically, expect to greatly improve the data prefetching performance of the business system and the hit rate during prefetching process. In terms of handling query optimization, we determine the query optimization algorithm fit for this system by comparing the performance status in the system of different algorithms, such as the immune algorithm, genetic algorithm and immune genetic algorithm. By contrast, the immune genetic algorithm can bring about obvious performance improvement. On this basis, we integrate the characteristics of data query on the platform, by selecting the inputs of immune genetic algorithm to further improve the data query performance.Finally, by means of the experiment, we compare data query prior to the improvement with improved by query performance from different points of view and different depths, and expect it can fully illustrate that bringing in pre-fetching algorithm, data query optimization algorithm and database connection pooling number automatically calculating strategy can produce superior performance improvement for the magic platform.