Abstract
Nowadays, the electrical energy efficiency is one of the most challenging issues in the area of ITs. DBMS have been pointed out as one of the major energy consumers. The reduction of their energy consumption becomes an urgent priority. Two aspects have to be considered in order to reduce this consumption: (i) the DBMS hosting the database applications and (ii) the Eco-design of these applications. Note that the first aspect got more attention than the second one. In this paper, we attempt to consider both aspects in the context of data warehouses (\(\mathcal {DW}\)). Firstly, we propose a generic framework integrating the energy in query optimizers of DBMS hosting already designed \(\mathcal {DW}\). An instantiation of this framework has been done on PostgreSQL DBMS. Secondly, and thanks to the variability that has been widely studied by the community of software, we propose to go back to the logical phase of the \(\mathcal {DW}\) life cycle and see how the energy may be integrated and then evaluate its impact on the physical phase. This variation is possible due to the relationships (hierarchies) that may exist among the properties of that \(\mathcal {DW}\). Finally, intensive experiments are conducted to evaluate the effectiveness and efficiency of our findings on PostgreSQL and Oracle DBMS.