Abstract—Cloud storage is a crucial component of cloud computing, where cloud storage providers supply a large amount of on-demand data storage so that users can save and retrieve their data anytime anywhere in the cloud. Any service interruption or failure could have negative and serious impact on the reputation of cloud service providers. Therefore, it is significant to model and design reliability for cloud storage systems. The RAID (Redundant Array of Independent Disks) technology is one of the most effective ways to achieve data reliability. There are several standard RAID levels with different fault tolerance techniques. In this paper, we model and evaluate reliability of RAID-6 configuration implemented in the cloud infrastructure (referred to as cloud-RAID-6). A hybrid analytical modeling method is proposed, which combines a Markov model for analyzing a single disk state probabilities and a multi-valued decision diagram model for analyzing the entire cloud-RAID-6 system state probabilities. The suggested method is applicable to identical and non-identical disks within the cloud-RAID-6. The proposed methodology is illustrated through examples.
Index Terms—Cloud-RAID-6, Markov model, multi-valued decision diagram, lattice structure, multi-state, reliability.
The authors are with the Department of Electrical and Computer Engineering, University of Massachusetts Dartmouth, MA 02747 USA (e-mail: qliu1@umassd.edu, lxing@umassd.edu).
[PDF]
Cite: Qisi Liu and Liudong Xing, "Reliability Modeling of Cloud-RAID-6 Storage System," International Journal of Future Computer and Communication vol. 4, no. 6, pp. 415-420, 2015.
