Based On Multi-dimensional Compression Parallel Array Storage Structure Online Analytical Processing Operation Algorithm

Data warehouse is a new research field, and domestic researching is on the beginning phase0 So there are few results about it0 We believe that data warehouse can not only be an application base on the database system, it can be a system itself as well Our research and practice have testified this opinion0 This paper presents a serial of sequential and parallel data operation algorithms for data warehouse using a new store method named Compressed Multidimensional Array or CMA for short0 These algorithms operate directly on the compressed data without the need to decompress the data first0 These algorithms are applicable to data warehouse that are compressed using variety of data compression methods 0 This paper designs and realizes the parallel data operation algorithms using two kinds of compression methods and three data distribution manners under the Linux and PVM environment and designs and realizes the sequential algorithms under the same environment at the same time 0 The analytical and experimental results show that these data operation algorithms are much more efficient than other algorithms What抯 more we propose a communication mode0 This mode can support the parallel communication between machines which can decrease the time cost for data communication0...