Scheduling Algorithm In Parallelizarion Of Xquery Programs

With the development of the Internet, XML is becoming the standard of data exchange and data sharing on the Internet. And it plays an increasingly important role in data storage on the Web and elsewhere. XQuery, as a standard XML query language, has attracted significant attentions in recent years. It is used to extract the information from the XML data source. The functions of XQuery in the XML is similar to SQL in a traditional relational database.XQuery is a functional programming language, which has quite a complicated program structure. Moreover, due to the semi-structured nature of XML, how to improve the efficiency of the XML query becomes a tough problem. To improve efficiency of XQuery processing, researchers came up with several approaches. XML algebra is investigated to optimize query processing. Twig pattern query captures the semi-structure characteristic of XML, outstanding twig pattern evaluation algorithms appeared and gained widespread attention. In order to improve query performance, the compiler technology used to optimize the program.On the other hand, with the rapid development of computer hardware technology, the multi-core computer is becoming increasingly popular. Various parallel approaches have been proposed to improve the performance of the implementation of XQuery. However, most existing algorithms are only designed for twig pattern matching, XPath or document processing.In order to achieve the high-performance processing of XML queries. In this paper, we propose a contains three parallel strategies parallel processing solution that is against the XQuery language. The solution includes an XQuery automated parallel processing framework to support parallel processing. We propose a novel approach for tasks scheduling in XQuery’s automatic parallel implementation. The approach solves the integrated scheduling problem on parallel multiprocessor system, which includes three strategies, i.e. task parallelism, data parallelism and pipeline. The scheduling algorithm also tries to balance load distribution.The experimental results show that our approach improves the performance and has good memory efficiency.