Disk drives are known to fail at a higher rate during their first year of operation than during the remaining years of their useful lifetime. We propose to use the free space that normally exists on new disks to minimize the risk of data loss during that first year. Our technique applies to disk arrays that mirror their data on two disks. Whenever a disk fails, the array will reorganize itself by storing a new copy of the disk that failed on one or more disks that have free space. This will protect the data against any single disk failure until the failed disk gets replaced and the system reverts to its original state. A Markov analysis of the behavior of a small system consisting of two pairs of mirrored disks indicates that our technique can reduce the probability of a data loss during the first year of operation of the system by at least 75 percent provided the disks have 34 percent of spare space.

MEMS-based storage is a new, non-volatile storage technology currently under development. It promises fast data access, high throughput, high storage density, small physical size, low power consumption, and low entry costs. These properties make MEMS-based storage into a serious alternative to disk drives, in particular for mobile applications. The first generation of MEMS will only offer a fraction of the storage capacity of disks; therefore, we propose to integrate multiple MEMS devices into a MEMS storage enclosure, organizing them as a RAID Level 5 with multiple spares, to be used as the basic storage building block. This paper investigates the reliability of such an enclosure. We find that Mean Time To Failure is an inappropriate reliability metric for MEMS enclosures. We show that the reliability of the enclosures is appropriate for their economic lifetime if users choose not to replace failed MEMS storage components. In addition, we investigate the benefits of occasional, preventive maintenance of enclosures. © 2004 IEEE.

The emergence of more reliable, often much faster solid-stage storage devices is revolutionizing most aspects of data storage technology. Here we address the impact of these new storage devices on the techniques used to ensure data survival. In particular, we show that the higher bandwidth-to-capacity ratios of the new Storage Class Memory devices will make self-adaptive storage solutions much more attractive.

We propose increasing the survivability of data stored in two-dimensional RAID arrays by causing these arrays to reorganize themselves whenever they detect a disk failure. This reorganization will rebalance as much as possible the redundancy level of all stored data, thus reducing the potential impact of additional disk failures. It remains in effect until the failed disk gets repaired. We show how our technique can be applied to two-dimensional RAID arrays consisting of n 2 data disks and In parity disks and show how it can increase the mean time to data loss of the array by at least 200 percent as long as the reorganization process takes less than half the time it takes to replace a failed disk. © 2007 IEEE.

We present a disk array organization that adapts itself to successive disk failures. When all disks are operational, all data are mirrored on two disks. Whenever a disk fails, the array reorganizes itself, by selecting a disk containing redundant data and replacing these data by their exclusive or (XOR) with the other copy of the data contained on the disk that failed. This will protect the array against any single disk failure until the failed disk gets replaced and the array can revert to its original condition. Hence data will remain protected against the successive failures of up to one half of the original number of disks, provided that no critical disk failure happens while the array is reorganizing itself. As a result, our scheme achieves the same access times as a mirrored organization under normal operational conditions while having a much lower likelihood of loosing data under abnormal conditions. In addition it tolerates much longer repair times than mirrored disk arrays. © Springer-Verlag Berlin Heidelberg 2006.

While mean time to data loss (MTTDL) provides an easy way to estimate the reliability of redundant disk arrays, it fails to take into account the relatively short lifetime of these arrays. We analyzed five different disk array organizations and compared the reliability estimates obtained using their mean times to data loss with the more exact values obtained by directly solving their corresponding Markov model. We observed that the conventional MTTDL approach generally provided a good estimate of the long-term reliability of arrays— with the exception of non-repairable arrays—while significantly underestimating the short-term reliability of these arrays. 12

Storage class memories (SCMs) constitute an emerging class of non-volatile storage devices that promise to be significantly faster and more reliable than magnetic disks. We propose to add one of these devices to each group of two or three RAID level arrays and store on it additional parity data. We show that the new organization can tolerate all double disk failures, between 75 and 90 percent of all triple disk failures and between 50 and 70 percent of all failures involving two disks and the SCM device without incurring any data loss. As a result, the additional parity device increases the mean time to data loss of the arrays in the group it protects by at least 200-fold. ©2010 IEEE.

Adding flash memory to the storage hierarchy has recently gained a great deal of attention in both industry and academia. Decreasing cost, low power utilization and improved performance has sparked this interest. Flash reliability is a weakness that must be overcome in order for the storage industry to fully adopt flash for persistent storage in mission-critical systems such as high-end storage controllers and low-power storage systems. We consider the unique reliability properties of NAND flash and present a high-level architecture for a reliable NAND flash memory storage system. The architecture manages erasure-coded stripes to increase reliability and operational lifetime of a flash memory-based storage system, while providing excellent write performance. Our analysis details the tradeoffs such a system can make, enabling the construction of highly reliable flash-based storage systems.

Large archival storage systems experience long periods of idleness broken up by rare data accesses. In such systems, disks may remain powered off for long periods of time. These systems can lose data for a variety of reasons, including failures at both the device level and the block level. To deal with these failures, we must detect them early enough to be able to use the redundancy built into the storage system. We propose a process called "disk scrubbing" in a system in which drives are periodically accessed to detect drive failure. By scrubbing all of the data stored on all of the disks, we can detect block failures and compensate for them by rebuilding the affected blocks. Our research shows how the scheduling of disk scrubbing affects overall system reliability, and that "opportunistic" scrubbing, in which the system scrubs disks only when they are powered on for other reasons, performs very well without the need to power on disks solely to check them. © 2004 IEEE.