iSWoM: The Incremental Storage Workload Model Based on Hidden Markov Models

详细信息    查看全文

• 作者：Tiberiu Chis (18)
  Peter G. Harrison (18)
  
• 刊名：Lecture Notes in Computer Science
• 出版年：2013
• 出版时间：2013
• 年：2013
• 卷：7984
• 期：1
• 页码：142-157
• 全文大小：213KB
• 参考文献：1. Baum, L.E., Petrie, T.: Stastical Inference for Probabilistic Functions of Finite Markov Chains. The Annals of Mathematical Statistics聽37, 1554鈥?563 (1966) CrossRef
  2. Baum, L.E., Petrie, T., Soules, G., Weiss, N.: A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains. The Annals of Mathematical Statistics聽41, 164鈥?71 (1970) CrossRef
  3. Viterbi, A.J.: Error bounds for convolutional codes and an asymptotically optimum decoding algorithm. IEEE Transactions on Information Theory聽13, 260鈥?69 (1967) CrossRef
  4. Krough, A., Brown, M., Mian, S., Sjolander, K., Haussler, D.: Hidden Markov Models in Computational Biology. Journal of Molecular Biology, 1501鈥?531 (1994)
  5. Burge, C., Karlin, S.: Prediction of complete gene structures in human genomic DNA. Journal of Molecular Biology, 78鈥?4 (1997)
  6. Ashraf, J., Iqbal, N., Khattak, N.S., Zaidi, A.M.: Speaker Independent Urdu Speech Recognition Using HMM (2010)
  7. Harrison, P.G., Harrison, S.K., Patel, N.M., Zertal, S.: Storage Workload Modelling by Hidden Markov Models: Application to Flash Memory. Performance Evaluation聽69, 17鈥?0 (2012) CrossRef
  8. Florez-Larrahondo, G., Bridges, S., Hansen, E.A.: Incremental Estimation of Discrete Hidden Markov Models on a New Backward Procedure (2005)
  9. Rabiner, L.R., Juang, B.H.: An Introduction to Hidden Markov Models. IEEE ASSP Magazine聽3, 4鈥?6 (1986) CrossRef
  10. Rabiner, L.R.: A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition. IEEE聽77, 257鈥?86 (1989) CrossRef
  11. Chis, T.: Hidden Markov Models: Applications to Flash Memory Data and Hospital Arrival Times (2011)
  12. Zhang, X., Riska, A., Riedel, E.: Characterization of the E-commerce Storage Subsystem Workload. In: Proceedings of Quantitative Evaluation of SysTems (2008)
  13. RUBiS Implementation: http://rubis.objectweb.org/
  14. On-Line Transaction Processing (OLTP) Benchmark: http://www.tpc.org/tpcc
  15. Kumas, Z., Keeton, K., Becker-Szendy, R.: I/O Workload Characterization. In: CAECW 2001, before HPCA-7 (2001)
  16. Transactional Web e-Commerce Benchmark: http://www.tpc.org/tpcw
  17. Chis, T., Harrison, P.G.: Incremental HMM with an improved Baum-Welch Algorithm. In: Proceedings of Imperial College Computing Student Workshop (2012)
  18. Stenger, B., Ramesh, V., Paragois, N., Coetzee, F., Buhmann, J.M.: Topology free Hidden Markov Models: Application to background modeling. In: Proceedings of the International Conference on Computer Vision, pp. 297鈥?01 (2001)
  19. Keaton, K., Veitch, A., Obal, D., Wilkes, J.: I/O characterization of commercial workloads. In: CAECW 2000 (2000)
  
• 作者单位：Tiberiu Chis (18)
  Peter G. Harrison (18)
  
  18. Department of Computing, Imperial College London, Huxley Building, 180 Queens Gate, London, SW7 2RH, UK
  
• ISSN：1611-3349

文摘

We propose a storage workload model able to process discrete time series incrementally, continually updating its parameters with the availability of new data. More specifically, a Hidden Markov Model (HMM) with an adaptive Baum-Welch algorithm is trained on two raw traces: a NetApp network trace consisting of timestamped I/O commands and a Microsoft trace also with timestamped entries containing reads and writes. Each of these traces is analyzed statistically and HMM parameters are inferred, from which a fluid input model with rates modulated by a Markov chain is derived. We generate new data traces using this Markovian fluid, workload model. To validate our parsimonious model, we compare statistics of the raw and generated traces and use the Viterbi algorithm to produce representative sequences of the hidden states. The incremental model is measured against both the standard model (parameterized on the whole dataset) and the raw data trace.