JPPF在构建高性能ETL过程中的应用  

Research of constructing JPPF based ETL process with high performance

作　　者：许力[1] 牟晓光[1] 马云存 

机构地区：[1]东软集团(大连)有限公司基础软件事业部,辽宁大连116023

出　　处：《计算机应用》2008年第S2期223-225,270,共4页journal of Computer Applications

基　　金：国家863计划项目(2006AA04Z166);国家科技支撑计划项目(2007BAH19B01)

摘　　要：数据仓库中数据规模的不断膨胀和数据实时性需求的提高使得对ETL过程性能要求越来越高。通过对并行处理框架JPPF分析,提出了利用JPPF构建高性能ETL系统架构方案和一种ETL数据处理任务提交算法。经过测试和性能对比证明了该方案在处理包含大规模计算任务的ETL过程优势明显。With the expansion of the data scale and increased real time data need in data warehouse,high performance ETL process is requested.According to the analysis of JPPF,put forward a solution of constructing high performance ETL structure by using JPPF and an ETL data procession tasks submission algorithm.After testing and performance comparison proves the prototype system has advantage in processing ETL process containing massive computation tasks.

关 键 词：数据装载工具 数据仓库 并行计算 网格计算 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]