Parallel data base processing method and its executing device
作者： Kazuo Masai , 一夫 正井 , 豊 細矢 , Yutaka Hosoya
DOI:
关键词: SQL 、 Real-time computing 、 Value (computer science) 、 Order (business) 、 Computer science 、 Processing methods 、 Base (topology) 、 Expression (computer science) 、 Schedule 、 Table (database)
摘要: PURPOSE: To improve the through-put of whole system by predicting load requested processing in parallel DB so that respective processors can be equalized. CONSTITUTION: Predicted execution time each engine processor to executed is found out based upon a calculation expression corresponding structured query language(SQL) estimating information 133 obtained optimization and allowed as predicted written table 132. The sum all processors, added estimation value recorded an SQL 131 request analyzing schedule processor. When referred another processor, considering also which not yet executed. Then access outputted operate data. After completing analysis, scheduler 124 adjusts determines efficient order COPYRIGHT: (C)1996,JPO
lens.org 参考文章(0)