The Research On Optimization Of Parallel Data Query Based On Multi-processor

Caching is a classic technology to improve performance of the applications in computerscience. Well-organized cache data can reduce the overhead cost in the IO processing andimprove the performance. Most of DBMS only use the frequencies of data accessed and usethe traditional cache structures, like the Page cache and Block cache. Queries in realapplications are aggregate. Subsequent queries are often related with previous queries. Usingthe cache data appropriately and answering the subsequent queries with the help of semanticinformation, can reduce the executions of similar queries and improve the performance ofdata accessing.This paper starts by analyzing the shortcoming of cache structures in MySQL. Since theDBMS can’ t satisfy the needs of applications, this paper design and develop a caching systemthat make use of the semantic information. The caching system have several advantages:1) It can use semantic information in the SQL to decide whether the cache data cansatisfy the query. It makes use of the data in the cache as much as possible;2) It can take advantage of the computing power of the multi-process to reduce the timein the query processing;3) It can manage the cache data according to the statistic information of the data queriesand increase the hit rate of the caching system;Our experiments in Chapter5show that the system has a great performance to the queriesthat data are changeless. It can use the cache to reduce the response time of the queryprocessing and use the multi-process to accelerate the query procedure.