Research on parallel query technology of pXCube model based on cloud pattern

Abstract

XML data cube has characteristics of large scale data and complex join operations during query process. In general, due to the large amount of fact data, the efficiency of join query is inefficient. Cloud computing is a high-performance computing environment, and parallel query technology can be adopted for XML data cube to improve efficiency of multi-dimensional query, when the cube is based on cloud pattern. A kind of XML data cube model-pXCube based on path computation is analyzed and parallel query method based on cloud pattern is proposed for the model. In particularly, Non-symmetrical strategy is used in pXCube's fact and dimension partition and the vector is uniform. On the basis, the query algorithm is provided. Experimental data shows that pXCube query parallel based on the cloud pattern has a good speedup and it is well-suitable for decision support query.