Most personal or industrial data is simply stored in files and accessed via simple programs. This approach works well for small applications but is not generic and does not scale. New applications require new software, and classical software can hardly cope with huge data sets. Database management systems (DBMS) have been built to provide a generic solution to this issue. Usually, a DBMS enforces a clear distinction between how the data is stored on disk and how it is accessed via queries. When querying a database, one should not be concerned about how and where the data is physically stored, but only with the logical structure of the data. This concept is called the data independence principle. This article highlights the query optimization in generic DBMS.