基于MapReduce 的关系型数据仓库并行查询 - Journal of.PDFVIP

第32 卷第5期 东 北 大 学 学 报 ( 自 然 科 学 版 ) Vol32, No. 5 201 1 年 5 月 Journal of Northeastern U niversity( Natural cience) M ay 2 0 1 1 MapReduce 师金钢, 鲍玉斌, 冷芳玲, 于 戈 ( , 110819) : MapReduce , ChunkDB MapReduce , ChunkDB MapReduce , MapReduce MapReduce ChunkDB : MapReduce; ; ; ; : TP 311. 13 : A : 1005-3026( 2011) 05-0626-04 Parallel Query for a Data Warehouse Utilizing MapReduce SHI Jin-gang , BA O Yu-bin, LEN G Fang-ling, Y U Ge ( chool of Information cience Engineering, Northeastern University, henyang 110819, China. Corresponding author : HI Jin-gang, E-mail: shijingang @ ise. neu. edu. cn) Abstract: MapReduce is a highly efficient distributed and parallel computing framework, allow ing users to readily manage large clusters in parallel computing. But the MapReduce framework is not compatible w ith traditional relational databases. This paper proposes a distributed relational database ChunkDB based on the chunk structure, and extends and redesigns the MapReduce framew ork to ensure compatibility w ith the ChunkDB database. Thus, scalability, ease of operation, the high parallelism of MapReduce were integrated with the advantages, including indexing, query optimization of a relational database. The ChunkDB database based on M apReduce provided fast and efficient parallel query for data w arehouse applications. Key words: MapReduce; data w arehouse; parallel computing; distributed database; query optimization M apReduce[ 1- 3] , 1 ChunkDB M apReduce M apReduce Hadoop[ 5- 6]