ABSTRACT
Traditional statistical methods are currently used to analyze the feature of hard drives failure, but difficult to model the failure feature for random events with small probability of hard drives failure or datasets with unbalanced and missing values. Therefore, this paper presents a change-point wavelet transform (CPWT) method, which can extract the indicative SMART change-point features of failed hard drives. Because wavelet has the ability of denoting local signal characteristics in the time domain and frequency domain, we use the wavelet transform method to extract the sudden change features of the time series, to reveal the characteristics of the change-point wavelet. Then, we use convolution network and long short-term memory (CNN-LSTM) model, to verify the effectiveness of the change-point wavelet extracted for failure prediction. The experimental results show that the change-point wavelet evaluates the health of the hard drives, the accuracy and precision of prediction are 91.56% and 90.78% respectively, confirming that the change-point wavelet has a certain predictive ability for hard drives failures, which provides a basis for cloud storage failure judgment and prediction.
- Narayanan I, Wang D, Jeon M, SSD failures in datacenters: What? when? and why?[C]//Proceedings of the 9th ACM International on Systems and Storage Conference. 2016: 1-11.Google Scholar
- Wang G, Zhang L, Xu W. What can we learn from four years of data center hardware failures?[C]//2017 47th Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN). IEEE, 2017: 25-36.Google Scholar
- Schroeder B, Gibson G A. Understanding drives failure rates: What does an MTTF of 1,000,000 hours mean to you?[J]. ACM Transactions on Storage (TOS), 2007, 3(3): 8-es.Google Scholar
- Yi Y, Xiao J, Wu S, Failure Order: A Missing Piece in Drives Failure Processing of Data Centers[C]//2019 IEEE 21st International Conference on High Performance Computing and Communications; IEEE 17th International Conference on Smart City; IEEE 5th International Conference on Data Science and Systems (HPCC/SmartCity/DSS). IEEE, 2019: 223-230.Google Scholar
- Zeydan E, Arslan S S. Cloud 2 HDD: Large-Scale HDD Data Analysis on Cloud for Cloud Datacenters[C]//2020 23rd Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN). IEEE, 2020: 243-249.Google Scholar
- Lu S, Luo B, Patel T, Making Drives Failure Predictions SMARTer![C]//18th {USENIX} Conference on File and Storage Technologies ({FAST} 20). 2020: 151-167.Google Scholar
- Ahmad W, Khan S A, Kim C H, Feature Selection for Improving Failure Detection in Hard Drives Drives Using a Genetic Algorithm and Significance Scores[J]. Applied Sciences, 2020, 10(9): 3200.Google ScholarCross Ref
- Queiroz L P, Rodrigues F C M, Gomes J P P, A fault detection method for hard drives drives based on mixture of Gaussians and nonparametric statistics[J]. IEEE Transactions on Industrial Informatics, 2016, 13(2): 542-550.Google ScholarCross Ref
- Zhang X, Liu Z, Wang J, Time–frequency analysis for bearing fault diagnosis using multiple Q-factor Gabor wavelets[J]. ISA transactions, 2019, 87: 225-234.Google Scholar
- Hsueh Y M, Ittangihal V R, Wu W B, Fault diagnosis system for induction motors by CNN using empirical wavelet transform[J]. Symmetry, 2019, 11(10): 1212.Google ScholarCross Ref
- Nishat Toma R, Kim J M. Bearing fault classification of induction motors using discrete wavelet transform and ensemble machine learning algorithms[J]. Applied Sciences, 2020, 10(15): 5251.Google ScholarCross Ref
- Liang P, Deng C, Wu J, Intelligent Fault Diagnosis of Rotating Machinery via Wavelet Transform, Generative Adversarial Nets and Convolutional Neural Network[J]. Measurement, 2020: 107768.Google Scholar
- Lee C Y, Cheng Y H. Motor Fault Detection Using Wavelet Transform and Improved PSO-BP Neural Network[J]. Processes, 2020, 8(10): 1322.Google ScholarCross Ref
- Mashhadi A R, Cade W, Behdad S. Moving towards real-time data-driven quality monitoring: a case study of hard drives drives[J]. Procedia Manufacturing, 2018, 26: 1107-1115.Google ScholarCross Ref
- Wang J, Bao W, Zheng L, An Attention-augmented Deep Architecture for Hard Drive Status Monitoring in Large-scale Storage Systems[J]. ACM Transactions on Storage (TOS), 2019, 15(3): 1-26.Google Scholar
- Arslan S S, Zeydan E. On the Distribution Modeling of Heavy-Tailed Drives Failure Lifetime in Big Data Centers[J]. IEEE Transactions on Reliability, 2020.Google Scholar
- Huang S, Liang S, Fu S, Characterizing drives health degradation and proactively protecting against drives failures for reliable storage systems[C]//2019 IEEE International Conference on Autonomic Computing (ICAC). IEEE, 2019: 157-166.Google Scholar
- Alistair CH Rowe and Paul C Abbott. 1995. Daubechies wavelets and mathematica.Computers in Physics 9, 6 (1995), 635–648.Google Scholar
- N. Thakur and C. Y. Han, "A Framework for Prediction of Cramps during Activities of Daily Living in Elderly", Proceedings of the International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2020), Fuzhou, China, June 12-14, 2020Google ScholarCross Ref
Recommendations
A New Image Denoising Method Using Wavelet Transform
IFITA '09: Proceedings of the 2009 International Forum on Information Technology and Applications - Volume 01Wavelet image denoising has been widely used in the field of image noise. After taking into account the objective and subjective results of the noise image, this paper presents a new image denosing method. Firstly, this method decomposes the noisy image ...
New Image Denoising Method Based Wavelet and Curvelet Transform
ICIE '09: Proceedings of the 2009 WASE International Conference on Information Engineering - Volume 01Wavelet multi-resolution analysis is widely used and achieves good results in image denoising fields. Wavelet transform has unique advantages in dealing with singular point, but has some disadvantages in dealing with the borders and the line features of ...
The optimal Mexican hat wavelet filter de-noising method based on cross-validation method
A new de-noising method based on parameter optimized Mexican hat wavelet was put forward in this paper. For the similar shape to the mechanical shock vibration signal, the Mexican hat wavelet is chosen as the mother wavelet and improved by the shape ...
Comments