The article is from 2008. Since then, the so called parallel DB goes no where, and Hadoop takes over the world. The main problem of traditional (OLAP) DBMS in

The article is from 2008. Since then, the so called parallel DB goes no where, and Hadoop takes over the world. The main problem of traditional (OLAP) DBMS in the era of big data is that ETL (Extract/Transform/Load) becomes the main bottle neck rather than complex queries, as big data is inherently semi-structured and noisy. MR is the tool to process big data.