Energy Cost Minimization in Cloud Datacenter

In contemporary business environments, a growing number of enterprises are transitioning their operations from on-premise servers to cloud infrastructure. This shift has given rise to heightened computational demands, leading to substantial daily energy consumption in datacenters. The consequential strain on energy resources has prompted extensive scrutiny within the context of energy efficiency. This study delves into the intricate realm of energy-conscious resource management in cloud datacenters, specifically addressing the inclusion of green energy sources characterized by unpredictable capacity. A resilient decentralized resource management framework is proposed, leveraging reinforcement learning techniques. The methodology draws upon historical knowledge, eliminating dependencies on request arrival patterns and energy supply fluctuations. Empirical findings demonstrate the efficacy of the proposed approach, revealing significant cost reductions in comparison to established benchmark algorithms.