Cube Computation in Distributed Environment Using CM-Sketch Algorithm

Now a day's large amount of data is generated in structured and unstructured format. The volume of data stored is very large and huge in terms of terabyte, petabyte n sometime in zettabyte. So to analyses such huge data there is need to improve traditional RDBMS techniques. Data cube is commonly used operation in large amount of data that stores huge volume of data, analyzed and find out hidden information from data.
This paper addresses number of issues of constructing cubes for massive amount data. CM-Sketch algorithm used to partition data across nodes for cube construction. CM sketch algorithm performs ordering of dimension that minimizes the computation time of cube. After partitioning, cube generation algorithm used for cube construction over node. Experimental results from implementation of algorithm shows its effectiveness. For analysis of experimental result we consider parameters like cube construction with varying data size, parallelism and number hierarchies of dimension for cube construction.