State reduction method of Markov chain calculation for a highly-available storage device

This thesis proposes data reduction method for Markov chain to calculate availability of a storage device. Components of a highly-available storage device are redundant to prevent service loss and data loss. The number of the states of the storage device increases exponentially following with increment of components. By this proposed method, the number of the states can be reduced from 424 to 81.

[1]  Jiri Schindler,et al.  Beyond MTTDL: A Closed-Form RAID 6 Reliability Equation , 2014, TOS.

[2]  James S. Plank,et al.  Mean Time to Meaningless: MTTDL, Markov Models, and Storage System Reliability , 2010, HotStorage.

[3]  Eduardo Pinheiro,et al.  DRAM errors in the wild: a large-scale field study , 2009, SIGMETRICS '09.

[4]  Kazunori Ueda,et al.  Reliability Analysis of Highly Redundant Distributed Storage Systems with Dynamic Refuging , 2015, 2015 23rd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing.

[5]  Monte Carlo,et al.  Markov Chain , 2017, Encyclopedia of Machine Learning and Data Mining.