ISSN 1000-1239 CN 11-1777/TP

Journal of Computer Research and Development ›› 2019, Vol. 56 ›› Issue (9): 1821-1831.doi: 10.7544/issn1000-1239.2019.20180670

Previous Articles     Next Articles

Energy-Efficiency Query Optimization for Green Datacenters

Xing Baoping1, Lü Mengyuan1, Jin Peiquan1,2, Huang Guorui3, Yue Lihua1,2   

  1. 1(School of Computer Science and Technology, University of Science and Technology of China, Hefei 230027); 2(Key Laboratory of Electromagnetic Space Information, Chinese Academy of Sciences, Hefei 230027); 3(Unit 31002, People’s Liberation Army of China, Beijing 100081)
  • Online:2019-09-10
  • Supported by: 
    This work was supported by the General Program of the National Natural Science Foundation of China (61672479).

Abstract: Reducing energy consumption and building green datacenters has been one of the major needs of modern large-scale datacenters. In green datacenters, a key research issue is how to lower the energy consumption of database systems while keeping stable performance. This issue is called energy efficiency, and has become a new research frontier recently. Energy efficiency of database systems is defined as using little energy to accomplish as many operations as possible. High energy efficiency means that database systems can use less energy while processing a fixed number of operations. In other words, it uses less energy but achieves the same performance. In this paper, we propose a method for energy-efficient query optimization. First, an operator-level power model is established based on the regression analysis method, which can accurately predict average power consumption during query execution for a given query. Next, a new cost model is proposed for query optimizer, which considers both energy and performance aspects. The new cost model uses a new factor to obtain a better tradeoff between performance and energy costs. A testbed is built for measuring energy consumption of database systems, and the TPC-H and TPC-C benchmarks are used to evaluate the performance of our proposal. The results show that the proposed power model achieves higher precision than existing methods. In addition, the proposed performance-degrade factor can provide flexible trade-offs between performance and energy. Moreover, by setting up an appropriate performance-degrade factor, better energy efficiency can be achieved than the original PostgreSQL.

Key words: green datacenter, energy efficiency, query optimization, cost model, power model

CLC Number: