What is it about?
Data centers are large scale, energy-hungry infrastructure serving the increasing computational demands as the world is becoming more connected in smart cities. The emergence of advanced technologies such as cloud-based services, internet of things (IoT) and big data analytics has augmented the growth of global data centers, leading to high energy consumption. This upsurge in energy consumption of the data centers not only incurs the issue of surging high cost (operational and maintenance) but also has an adverse effect on the environment. Dynamic power management in a data center environment requires the cognizance of the correlation between the system and hardware level performance counters and the power consumption. Power consumption modeling exhibits this correlation and is crucial in designing energy-efficient optimization strategies based on resource utilization. Several works in power modeling are proposed and used in the literature. However, these power models have been evaluated using different benchmarking applications, power measurement techniques and error calculation formula on different machines. In this work, we present a taxonomy and evaluation of 24 software-based power models using a unified environment, benchmarking applications, power measurement technique and error formula, with the aim of achieving an objective comparison. We use different servers architectures to assess the impact of heterogeneity on the models' comparison. The performance analysis of these models is elaborated in the paper.
Featured Image
Photo by Mert Guller on Unsplash
Why is it important?
We create a taxonomy and comparative evaluation of state-of-the-art software based server power consumption models under a unified experimental setup. With the proliferation of data centers to serve big data analytics, IoT and smart applications in smart cities, the issue of energy consumption becomes critical. Two significant findings are that: a) the power models' performance is dependent on the applications' type used, and b) the relative performance of these models remains on different server architectures.
Perspectives
It is a great pleasure to write this article and this work I conducted from the heart and with great energy to save the planet its energy. This study was born from many years of experience in energy savings. I experienced real-time running data centers and was astonished by the high energy consumption. With the proliferation of data centers as smart cities are emerging, the issue of energy consumption becomes very crucial. This topic touches everyone on this planet, from environmental pollution destroying our planet and creating health diseases to economic impacts. This paper investigates the power models in the literature for the best model to save energy based on an objective comparison. I hope you find this article a pleasure to read and a contribution to save the planet.
Prof. Dr. Leila Ismail
United Arab Emirates University
Read the Original
This page is a summary of: Computing Server Power Modeling in a Data Center, ACM Computing Surveys, May 2021, ACM (Association for Computing Machinery),
DOI: 10.1145/3390605.
You can read the full text:
Resources
Computing Server Power Modeling in a Data Center: Survey, Taxonomy, and Performance Evaluation
Published in ACM Computing Survyes: Computing Server Power Modeling in a Data Center: Survey, Taxonomy, and Performance Evaluation Leila Ismail and Huned Materwala
Linear Power Modeling for Cloud Data Centers: Taxonomy, Locally Corrected Linear Regression, Simulation Framework and Evaluation
Linear Power Modeling for Cloud Data Centers: Taxonomy, Locally Corrected Linear Regression, Simulation Framework and Evaluation Leila Ismail and Eyad H. Abed
Contributors
The following have contributed to this page
