AN EFFICIENT MACHINE LEARNING APPROACH FOR VIRTUAL MACHINE RESOURCE DEMAND PREDICTION

Published 28 Feb 2019 •  vol 123  • 


Authors:

 

Jitendra Kumar, Department of Computer Applications, National Institute of Technology Kurukshetra, India
Ashutosh Kumar Singh, Department of Computer Applications, National Institute of Technology Kurukshetra, India

Abstract:

 

In an ever increasing business competition among cloud service providers, it has become a challenging task to ensure the customer satisfaction. A service provider always seek for a mechanism to manage the resource efficiently along with improved quality of service (QoS) and service level agreements (SLAs). In this paper, we present a machine learning approach to estimate the virtual machines' resource demands that can contribute in scaling decisions. The proposed scheme exploits the neural network to anticipate the future demands and learns the synaptic weights using a population based optimization algorithm inspired by blackhole phenomenon of the nature. The approach is tested on benchmark data of Google cluster trace and a significant improvement is observed in the forecast accuracy over existing prediction models. The scheme reduced the root mean squared error upto 83.40% and 81.88% for CPU and memory request trace. The efficacy of the method is also validated using a statistical evaluation using Friedman test.

Keywords:

 

Forecasting, Cloud Resource Demand, Google Cluster Trace, Neural Network, Statistical Analysis

References:

 

[1] Eileen Smith, Michael Shirer. “Worldwide Public Cloud Services Spending Forecast to Reach $160 Billion This Year.” According to IDC; 2018. Available from: https://www.idc.com/getdoc.jsp?containerId=prUS43511618.
[2] Amiri M, Mohammad-Khanli L. “Survey on prediction models of applications for resources provisioning in cloud.” Journal of Network and Computer Applications 82 (2017): 93–113.
[3] Liu N, Li Z, Xu J, et al. “A Hierarchical Framework of Cloud Resource Allocation and Power Management Using Deep Reinforcement Learning.” IEEE 37th International Conference on Distributed Computing Systems (ICDCS); Atlanta, Georgia, USA, 5–8 June 2017. IEEE, 2017, pp. 372–382.
[4] Qiu F, Zhang B, Guo J. “A deep learning approach for VM workload prediction in the cloud.” 17th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (SNPD).” Shanghai, China, 30 May-1 June 2016. IEEE, pp. 319–324.
[5] Kumar J, Goomer R, Singh AK. “Long Short Term Memory Recurrent Neural Network (LSTMRNN) Based Workload Forecasting Model for Cloud Datacenters.” Procedia Computer Science 125 (2018): 676–682.
[6] Zhang Q, Yang LT, Yan Z, et al. “An Efficient Deep Learning Model to Predict Cloud Workload for Industry Informatics.” IEEE Transactions on Industrial Informatics 14.7 (2018): 1–9.
[7] Zhang Y, Yao J, Guan H. “Intelligent Cloud Resource Management with Deep Reinforcement Learning.” IEEE Cloud Computing 4.6 (2017): 60–69.
[8] Patel YS, Misra R. “Performance Comparison of Deep VM Workload Prediction Approaches for Cloud.” Progress in computing, analytics and networking, Springer, Singapore; 2018m, pp. 149–160.
[9] Kumaraswamy S, Nair MK. “Intelligent VMs prediction in cloud computing environment.” International Conference On Smart Technologies For Smart Nation (SmartTechCon), Bengaluru, India, 17-19 August 2017, pp. 288–294.
[10] Wamba GM, Li Y, Orgerie AC, et al. “Cloud Workload Prediction and Generation Models.” 29th International Symposium on Computer Architecture and High Performance Computing, SBAC-PAD, Brazil, 17-20 October 2017, pp. 89–96.
[11] Kumar J, Singh AK. “Workload prediction in cloud using artificial neural network and adaptive differential evolution.” Future Generation Computer Systems 81 (2018): 41–52.
[12] Kumar J, Singh AK. “Dynamic resource scaling in cloud using neural network and black hole algorithm.” Fifth International Conference on Eco-friendly Computing and Communication Systems (ICECCS), Bhopal, India, 8-9 Dec 2016, pp. 63–67.
[13] Mason K, Duggan M, Barrett E, et al. “Predicting host CPU utilization in the cloud using evolutionary neural networks.” Future Generation Computer Systems 86 (2018): 162–173.
[14] Hatamlou A. “Black hole: A new heuristic optimization approach for data clustering.” Information Sciences 222 (2013): 175–184.
[15] Reiss C, Tumanov A, Ganger GR, et al. “Heterogeneity and Dynamicity of Clouds at Scale: Google Trace Analysis.” ACM Symposium on Cloud Computing, San Jose, CA, USA, 14-17 October 2012, pp. 1–18.
[16] Friedman M. “The use of ranks to avoid the assumption of normality implicit in the analysis of variance.” Journal of the American Statistical Association 32.200 (1937): 675–701.
[17] Finner H. “On a monotonicity problem in step-down multiple test procedures.” Journal of the American Statistical Association 88.423 (1993): 920–923.

Citations:

 

APA:
Kumar, J. & Kumar Singh, A. (2019). An Efficient Machine Learning Approach for Virtual Machine Resource Demand Prediction. International Journal of Advanced Science and Technology (IJAST), ISSN: 2005-4238(Print); 2207-6360 (Online), NADIA, 123, 21-30. doi: 10.33832/ijast.2019.123.03.

MLA:
Kumar, Jitendra, et al. “An Efficient Machine Learning Approach for Virtual Machine Resource Demand Prediction.” International Journal of Advanced Science and Technology, ISSN: 2005-4238(Print); 2207-6360 (Online), NADIA, vol. 123, 2019, pp. 21-30. IJAST, http://article.nadiapub.com/IJAST/Vol123/3.html.

IEEE:
[1] J. Kumar and A. Kumar Singh, “An Efficient Machine Learning Approach for Virtual Machine Resource Demand Prediction.” International Journal of Advanced Science and Technology (IJAST), ISSN: 2005-4238(Print); 2207-6360 (Online), NADIA, vol. 123, pp. 21-30, Feb. 2019.