Improving energy efficiency in data centers by controlling task distribution and cooling

Yusuke Nakajo, Jayati Athavale, Minami Yoda, Yogendra Joshi, Hiroaki Nishi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The rapid growth in cloud computing, the Internet of Things (IoT), and data processing via Machine Learning (ML), have greatly increased our need for computing resources. Given this rapid growth, it is expected that data centers will consume more and more of our global energy supply. Improving their energy efficiency is therefore crucial. One of the biggest sources of energy consumption is the energy required to cool the data centers, and ensure that the servers stay within their intended operating temperature range. Indeed, about 40% of a data center’s total power consumption is for air conditioning[1]. Here, we study how the server air inlet and outlet, as well as the CPU, temperatures depend upon server loads typical of real Internet Protocol (IP) traces. The trace data used here are from Google clusters and include the times, job and task ID, as well as the number and usage of CPU cores. The resulting IT loads are distributed using standard load-balancing methods such as Round Robin (RR) and the CPU utilization method. Experiments are conducted in the Data Center Laboratory (DCL) at the Georgia Institute of Technology to monitor the server outlet air temperature, as well as real-time CPU temperatures for servers at different heights within the rack. Server temperatures were measured by on-line temperature monitoring with Xbee, Raspberry PI, Arduino, and hot-wire anemometers. Given that the temperature response varies with server position, in part due to spatial variations in the cooling airflow over the rack inlet and the server fan speeds, a new load-balancing approach that accounts for spatially varying temperature response within a rack is tested and validated in this paper.

Original languageEnglish
Title of host publicationASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018
PublisherAmerican Society of Mechanical Engineers (ASME)
ISBN (Electronic)9780791851920
DOIs
Publication statusPublished - 2018 Jan 1
EventASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018 - San Francisco, United States
Duration: 2018 Aug 272018 Aug 30

Other

OtherASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018
CountryUnited States
CitySan Francisco
Period18/8/2718/8/30

Fingerprint

Energy efficiency
Servers
Cooling
Program processors
Temperature
Resource allocation
Internet protocols
Air intakes
Anemometers
Cloud computing
Air conditioning
Fans
Learning systems
Electric power utilization
Energy utilization
Wire
Monitoring
Air
Experiments

Keywords

  • Data Center
  • Load Balancing
  • Wireless Sensor System

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Hardware and Architecture

Cite this

Nakajo, Y., Athavale, J., Yoda, M., Joshi, Y., & Nishi, H. (2018). Improving energy efficiency in data centers by controlling task distribution and cooling. In ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018 American Society of Mechanical Engineers (ASME). https://doi.org/10.1115/IPACK2018-8305

Improving energy efficiency in data centers by controlling task distribution and cooling. / Nakajo, Yusuke; Athavale, Jayati; Yoda, Minami; Joshi, Yogendra; Nishi, Hiroaki.

ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018. American Society of Mechanical Engineers (ASME), 2018.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Nakajo, Y, Athavale, J, Yoda, M, Joshi, Y & Nishi, H 2018, Improving energy efficiency in data centers by controlling task distribution and cooling. in ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018. American Society of Mechanical Engineers (ASME), ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018, San Francisco, United States, 18/8/27. https://doi.org/10.1115/IPACK2018-8305
Nakajo Y, Athavale J, Yoda M, Joshi Y, Nishi H. Improving energy efficiency in data centers by controlling task distribution and cooling. In ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018. American Society of Mechanical Engineers (ASME). 2018 https://doi.org/10.1115/IPACK2018-8305
Nakajo, Yusuke ; Athavale, Jayati ; Yoda, Minami ; Joshi, Yogendra ; Nishi, Hiroaki. / Improving energy efficiency in data centers by controlling task distribution and cooling. ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018. American Society of Mechanical Engineers (ASME), 2018.
@inproceedings{b26cdb46cf5f470faa7523950316f6e2,
title = "Improving energy efficiency in data centers by controlling task distribution and cooling",
abstract = "The rapid growth in cloud computing, the Internet of Things (IoT), and data processing via Machine Learning (ML), have greatly increased our need for computing resources. Given this rapid growth, it is expected that data centers will consume more and more of our global energy supply. Improving their energy efficiency is therefore crucial. One of the biggest sources of energy consumption is the energy required to cool the data centers, and ensure that the servers stay within their intended operating temperature range. Indeed, about 40{\%} of a data center’s total power consumption is for air conditioning[1]. Here, we study how the server air inlet and outlet, as well as the CPU, temperatures depend upon server loads typical of real Internet Protocol (IP) traces. The trace data used here are from Google clusters and include the times, job and task ID, as well as the number and usage of CPU cores. The resulting IT loads are distributed using standard load-balancing methods such as Round Robin (RR) and the CPU utilization method. Experiments are conducted in the Data Center Laboratory (DCL) at the Georgia Institute of Technology to monitor the server outlet air temperature, as well as real-time CPU temperatures for servers at different heights within the rack. Server temperatures were measured by on-line temperature monitoring with Xbee, Raspberry PI, Arduino, and hot-wire anemometers. Given that the temperature response varies with server position, in part due to spatial variations in the cooling airflow over the rack inlet and the server fan speeds, a new load-balancing approach that accounts for spatially varying temperature response within a rack is tested and validated in this paper.",
keywords = "Data Center, Load Balancing, Wireless Sensor System",
author = "Yusuke Nakajo and Jayati Athavale and Minami Yoda and Yogendra Joshi and Hiroaki Nishi",
year = "2018",
month = "1",
day = "1",
doi = "10.1115/IPACK2018-8305",
language = "English",
booktitle = "ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018",
publisher = "American Society of Mechanical Engineers (ASME)",

}

TY - GEN

T1 - Improving energy efficiency in data centers by controlling task distribution and cooling

AU - Nakajo, Yusuke

AU - Athavale, Jayati

AU - Yoda, Minami

AU - Joshi, Yogendra

AU - Nishi, Hiroaki

PY - 2018/1/1

Y1 - 2018/1/1

N2 - The rapid growth in cloud computing, the Internet of Things (IoT), and data processing via Machine Learning (ML), have greatly increased our need for computing resources. Given this rapid growth, it is expected that data centers will consume more and more of our global energy supply. Improving their energy efficiency is therefore crucial. One of the biggest sources of energy consumption is the energy required to cool the data centers, and ensure that the servers stay within their intended operating temperature range. Indeed, about 40% of a data center’s total power consumption is for air conditioning[1]. Here, we study how the server air inlet and outlet, as well as the CPU, temperatures depend upon server loads typical of real Internet Protocol (IP) traces. The trace data used here are from Google clusters and include the times, job and task ID, as well as the number and usage of CPU cores. The resulting IT loads are distributed using standard load-balancing methods such as Round Robin (RR) and the CPU utilization method. Experiments are conducted in the Data Center Laboratory (DCL) at the Georgia Institute of Technology to monitor the server outlet air temperature, as well as real-time CPU temperatures for servers at different heights within the rack. Server temperatures were measured by on-line temperature monitoring with Xbee, Raspberry PI, Arduino, and hot-wire anemometers. Given that the temperature response varies with server position, in part due to spatial variations in the cooling airflow over the rack inlet and the server fan speeds, a new load-balancing approach that accounts for spatially varying temperature response within a rack is tested and validated in this paper.

AB - The rapid growth in cloud computing, the Internet of Things (IoT), and data processing via Machine Learning (ML), have greatly increased our need for computing resources. Given this rapid growth, it is expected that data centers will consume more and more of our global energy supply. Improving their energy efficiency is therefore crucial. One of the biggest sources of energy consumption is the energy required to cool the data centers, and ensure that the servers stay within their intended operating temperature range. Indeed, about 40% of a data center’s total power consumption is for air conditioning[1]. Here, we study how the server air inlet and outlet, as well as the CPU, temperatures depend upon server loads typical of real Internet Protocol (IP) traces. The trace data used here are from Google clusters and include the times, job and task ID, as well as the number and usage of CPU cores. The resulting IT loads are distributed using standard load-balancing methods such as Round Robin (RR) and the CPU utilization method. Experiments are conducted in the Data Center Laboratory (DCL) at the Georgia Institute of Technology to monitor the server outlet air temperature, as well as real-time CPU temperatures for servers at different heights within the rack. Server temperatures were measured by on-line temperature monitoring with Xbee, Raspberry PI, Arduino, and hot-wire anemometers. Given that the temperature response varies with server position, in part due to spatial variations in the cooling airflow over the rack inlet and the server fan speeds, a new load-balancing approach that accounts for spatially varying temperature response within a rack is tested and validated in this paper.

KW - Data Center

KW - Load Balancing

KW - Wireless Sensor System

UR - http://www.scopus.com/inward/record.url?scp=85057238771&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057238771&partnerID=8YFLogxK

U2 - 10.1115/IPACK2018-8305

DO - 10.1115/IPACK2018-8305

M3 - Conference contribution

AN - SCOPUS:85057238771

BT - ASME 2018 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems, InterPACK 2018

PB - American Society of Mechanical Engineers (ASME)

ER -