Date of Publication :17th February 2024

Abstract: The high energy consumption has now become a critical and urgent issue for database community. Query optimizer are identified as one of the most energy hungriest component in Data Storage Systems (DSSs). This is mainly due to the users’ needs to ingest, store and process data very frequently. With the torch of environmental sustainability being waved and the exorbitant cost of energy, the development and application of energy reduction techniques within these systems is more urgent than ever. In this paper, we propose a tool baptized “GreenPipeline” that enhances the query execution engine of the PostgreSQL system by integrating energy constraints during Analytical Queries processing. This initiative aligns with the development of environmentally friendly databases and has two primary objectives: to evaluate the energy efficiency benefits achieved by incorporating an energy model into query processing and to provide users a means to estimate the system’s energy consumption without requiring physical equipment. More specifically, our primary goal is to integrate our energy cost model and an evaluation plan methodology into the core of the PostgreSQL optimizer. We offer an evaluation plan approach that enables the optimization of either energy efficiency, performance, or a compromise between the two. Preliminary results highlight the effectiveness of our tool in reducing energy consumption for various query sets from different benchmarks.

Reference :