RAID Storage Systems with Early-warning and Data Migration

This paper proposes a new RAID-based system, we defined as REM: RAID system with Early-warning and data Migration. REM can predict RAID potential fault according to disk health degree and array health. Once the fault has been predicted, it takes data migration to protect the data, which allocates reserved space in every disk and arranges to redirect the data in the non-repairable sectors to the reserved space; when the number of the bad sectors in the disk exceeds threshold or the disk is in poor health, the system will copy data of the unreliable disk to a new one; when an array of multiple arrays early-warning, data can be moved into other backup array of the multiple arrays. Test results show that REM can enhance the reliability of system and reduce the entire system’s performance. Keywordsearly-warning; data migration; reliability;

[1]  Joseph F. Murray,et al.  Improved disk-drive failure warnings , 2002, IEEE Trans. Reliab..

[2]  Joseph F. Murray,et al.  Machine Learning Methods for Predicting Failures in Hard Drives: A Multiple-Instance Application , 2005, J. Mach. Learn. Res..

[3]  Greg Hamerly,et al.  Bayesian approaches to failure prediction for disk drives , 2001, ICML.

[4]  Randy H. Katz,et al.  A case for redundant arrays of inexpensive disks (RAID) , 1988, SIGMOD '88.

[5]  Joseph F. Murray,et al.  Hard drive failure prediction using non-parametric statistical methods , 2003 .

[6]  Scott A. Brandt,et al.  Reliability mechanisms for very large storage systems , 2003, 20th IEEE/11th NASA Goddard Conference on Mass Storage Systems and Technologies, 2003. (MSST 2003). Proceedings..