A.I. Ustimov - Master\'s student, Department of Information Processing and Control Systems, Bauman Moscow State Technical University
With increasing of volumes of data being processed, problems with RDBMS using are appeared. One possible solution are new technologies based on parallel proccessing. This article compares execution time of request to 3D data warehouse in RDBMS MySQL and MapReduce/Hadoop environments. In Hadoop environment a request can be executed in parallel on multiple data warehouse nodes. But this technology doesn’t support data indexing by default. Otherwise, in MySQL environment a request is executed on single node, but it can be optimised with using of indexing. So, what option will be faster is not predictable in this case.
Result of this comparison was data fragmentation and parallel processing in Hadoop prevailing under optimising and indexing in RDBMS MySQL.