XQuery并行实现中的任务调度算法

Task scheduling of XQuery language parallelization

为提高XQuery语言的处理性能,针对XQuery并行实现中的任务调度问题,提出一种适用于共享内存多线程环境的调度算法。在一种新型调度策略的指导下,能够利用XQuery语言中存在任务并行性、数据并行性和流水线并行性的特点,提高程序并行执行效率;针对流水线并行执行方式,建立一种流水线局部并行自动机模型,通过利用流水线中各节拍之间的空闲等待时间,提高系统资源的利用率。通过实验验证了该算法的可行性和有效性。

To improve the performance of XQuery, a novel plementation was proposed. The scheduling problem in the approach for tasks scheduling in the XQuery＇s automatic parallel ira- shared memory multithread environment was solved and three strate- gies including task parallelism, data parallelism and pipeline parallelism were used. For the optimization of pipeline, an automa- ton model was established to reduce the idle time between pipeline stages. The validity and efficiency of the presented approach was demonstrated by experimental results.