From Performance’s Designing Feedback in the Light of Energy Efficiency Improving: How & Proofs?

At the core of any Data Storage Systems (DSS), ongoing research efforts are dedicated to refining algorithms for effective data storage and retrieval. The primary objective of any data management systems (DBMS) is improving the performance of data management such optimizing queries response times. Decades of research have borne fruit in the form of performance-oriented optimization techniques, revolutionizing DSS usability and enhancing the user experience significantly. However, the explosion of digital data and the rise of various data-driven services, combined with the ongoing decline in processing equipment costs, are contributing significantly to the heightened energy consumption of data storage systems. With the pervasive incorporation of digital services into our daily lives, energy management has become a major challenge for digital companies to mitigate their environmental footprint and reduce their energy cost. Considering this perspective, several initiatives have been undertaken, such as the development and exploitation of renewable energies, or the redefinition of the Database (DB)’s primary quality objective: from the performance-oriented approaches to an energy-integrated approaches. In this paper, we suggest leveraging the advancements in performance characterizing to integrate energy efficiency into DSS. Initially, we assess the energy efficiency of a particular optimization technique, specifically, data partitioning applied during physical design. Next, we evaluate the energy efficiency of the query processor system by introducing a multi-objective formulation that considers two non-functional constraints: performance and energy consumption. Lastly, we present a simulator that incorporates an energy model, enabling the generation of query plans optimized for energy efficiency.