Energy-Efficient Green Database Systems: Adaptive Resource Allocation in Cloud Platforms

This research attempts to develop a Green DBMS that is generally designed for the cloud computing environment, emphasizing dynamic resource allocation. Because of the spread of cloud computing, this concern about energy relating to data centers has increased. In this regard, the paper introduces a novel methodology with less energy consumption at the DBMS level of a cloud and uses a dynamic system to allocate resources. The system would use machine learning models to predict an optimal resource allocation to better balance performance with energy efficiency. We introduce a prototype that is always observing the performance and adjusting the energy consumption of the system, taking into account real-time unfolding demands for workloads. Extensive experiments show significant energy savings against static allocation strategies while maintaining system performance. This work is a contribution toward ongoing effort in achieving more sustainable cloud computing infrastructures.