Reducing The Energy Consumption Energy-Efficient Query Processing Node in Web Search Engines
Keywords:
Energy consumption, CPU Dynamic Voltage and Frequency Scaling, Web search engines.Abstract
Web search engines are made by thousands out of question handling hubs, i.e., servers devoted to process client inquiries. Such numerous servers expend a lot of energy, for the most part responsible to their CPUs, however they are important to guarantee low latencies, since clients expect sub-second reaction times (e.g., 500 ms). In any case, clients can scarcely see reaction times that are quicker than their desires. Henceforth, we propose the Predictive Energy Saving Online Scheduling Algorithm (PESOS) to choose the most proper CPU recurrence to process an inquiry on a for every center premise. PESOS goes for process questions by their due dates, and use abnormal state scheduling data to decrease the CPU energy utilization of a question handling hub. PESOS constructs its choice in light of inquiry effectiveness indicators, evaluating the preparing volume and handling time of a question. We tentatively assess PESOS upon the TREC ClueWeb09B gathering and the MSN2006 inquiry log. Results demonstrate that PESOS can decrease the CPU energy utilization of a question preparing hub up to 48% contrasted with a framework running at most extreme CPU center recurrence. PESOS beats moreover the best in class contender with a 20% energy saving, while the contender requires a fine parameter tuning and it might brings about in wild inertness infringement.
References
- L. A. Barroso, J. Clidaras, and U. H¨olzle, The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines, 2nd ed. Morgan & Claypool Publishers, 2013.
- I. Arapakis, X. Bai, and B. B. Cambazoglu, "Impact of response latency on user behavior in web search," in Proc. SIGIR, 2014, pp. 103–112.
- The Climate Group for the Global e-Sustainability Initiative, "Smart 2020: Enabling the low carbon economy in the information age," 2008.
- U.S. Department of Energy, "Best Practices Guide for Energy-Efficient Data Center Design." [Online].
- D. C. Snowdon, S. Ruocco, and G. Heiser, "Power Management and Dynamic Voltage Scaling: Myths and Facts," in Proc. of Workshop on Power Aware Real-time Computing, 2005.
- D. Brodowski, "CPU frequency and voltage scaling code in the Linux kernel."
- C. Macdonald, N. Tonellotto, and I. Ounis, "Learning to predict response times for online query scheduling," in Proc. SIGIR, 2012, pp. 621–630.
- M. Jeon, S. Kim, S.-w. Hwang, Y. He, S. Elnikety, A. L. Cox, and S. Rixner, "Predictive parallelization: Taming tail latencies in web search," in Proc. SIGIR, 2014, pp. 253–262.
- S. Kim, Y. He, S.-w. Hwang, S. Elnikety, and S. Choi, "Delayeddynamic-selective (dds) prediction for reducing extreme tail latency in web search," in Proc. WSDM, 2015, pp. 7–16
- M. Catena, C. Macdonald, and N. Tonellotto, "Load-sensitive cpu power management for web search engines," in Proc. SIGIR, 2015, pp. 751–754
Downloads
Published
Issue
Section
License
Copyright (c) IJSRCSEIT
![Creative Commons License](http://i.creativecommons.org/l/by/4.0/88x31.png)
This work is licensed under a Creative Commons Attribution 4.0 International License.