Research And Realization To The Semantic Caching Management Of Large Transaction Processing

With the popularization of network computing, the scale of the certain business system' data up to TB level, e'en PB level. Storing and processing of tera data becomes the pivotal issue to impact natural running of system. The statistic analysis of massive information database is a time-consuming and resource-consuming task, as well as the essential factor to influence system performance. Therefore, to optimize the process of statistic analysis query becomes the major issue to enhance massive information database performance.This article takes L arge T ransaction Process System (LTPS for short) as the background and research the Semantic Caching in distributed environment to solve the issue on optimizing statistic query. B ased o n t his t heory, an optimizing s olution o n M id-tier Semantic C aching i s projected to emphasize on design and make part of caching management possible. The main tasks of this article include:1. Based on LTPS, according to characteristic of statistic analyses to tera data ,systematic structure in Mid-tier Semantic Caching are brought forward. That framework can support with caching share maximum and raise whole performance of the system availably.2. Customize cache, Temporary cache, Dummy cache, 3 different types of cache are defined is projected in this article. Establish cache by system with user-defined cache together, since support customers design cache by themselves also can estabilish and maintenance cache automatic according to analysis of query request of customer.3. Based on the classic strategies, in terms of the feature of LTPS, cache weak coherency policy based on period of validity is projected, based on analyses for query request, ascertain period of maintenance.4. This article points out WLFU (Weighted Least Frequently Used) cache replacement strategy according to management mechanism of cache queue. That strategy since respected the will of the customer and also will carry on the modification and replacement of caching at the same time.5. Designed and carried out a Semantic Caching of Large-scale Transaction Processing and check before acceptance through a topic the third square test, test the conclusion is excellent.