Abstract: With the trend of global low-carbon, as well as data-centric computing trends, studying the energy-saving green database systems has become a hot issue of government, business and academia. However, traditional database systems mainly focus on performance, and have little consideration on energy metrics, including energy efficiency and energy proportionality. In this paper, based on the requirement analysis on green database systems, some key issues on this topic were explored, and two critical problems were emphasized, namely the energy efficiency problem for database systems, as well as the energy proportionality problem for database clusters. Furthermore, some future directions on energy-aware green database systems were pointed out to provide some new insights for the research into this new area.

Key words: energy, green database, energy efficiency, energy proportionality

摘要： 随着全球的低碳化趋势以及以数据为中心的计算趋势,研究节能的绿色数据库系统已成为政府、企业和学术界共同关注的问题。然而,已有的数据库系统与算法大都针对性能目标而设计,缺乏对能耗的感知与处理,包括能耗有效性以及能耗同比性等问题的考虑。围绕绿色数据库系统的发展需求,对绿色数据库技术中的关键问题进行了讨论,重点阐述了数据库系统能耗有效性,以及面向数据库节点集群的能耗同比性等问题。在此基础上,给出了能耗感知的绿色数据库技术的未来发展方向,为该方向的进一步深入研究提供新的参考。

关键词: 能耗, 绿色数据库, 能耗有效性, 能耗同比性