subject: ABILITY OF DATA TO BE RECOVERED IN RAID SYSTEM FAILURES [print this page] ABILITY OF DATA TO BE RECOVERED IN RAID SYSTEM FAILURES
RAID data storage systems, or commonly explained as a Redundant Array of Inexpensive Disks, have been proven to satisfy the user who enjoys the processing speed and improvements in the capacity of large disks that it provides. Due to the multiple drives that are used to store the data, it's redundancy not only keeps the system running while one drive has failed but also in the event of a system malfunction aids the data recovery engineer in recovery efforts when multiple drive fail.
In layman's terms the mechanism of a RAID has two functions namely mirroring and striping. All RAID systems function in some compromise between the two, and as these two functions are interpolated this allows for larger multiple fault tolerances. Thus, the higher the RAID number, i.e. RAID 10 versus RAID 0, the higher the fault tolerance and data transfer speed is.
Mirroring simply makes a runtime copy of the primary drives' data onto a secondary drive. This drive keeps a complete backup' of all information so if one drive couldn't function properly; the other drive contains all the same information. Systems which use multiple drives in RAID topology (other than RAID 1) employ a data storage method called striping'. This means that if one drive in a striped RAID fails, it can be replaced by a new drive using stored equivalent or parity information. In other words the missing data will be rebuilt automatically. This superior back up functionality has lead to many organizations becoming interested in using RAID systems, and among the RAID topology RAID 5 has proved the most popular.
If more than one drive fails data retrieval becomes necessary and the success of retrieving data from multiple drive arrays might be lower. Therefore the key factor in determining successful recovery of user data is the stripe size which is normally 128K. However, in most RAID systems this could go down to 64K, and the size will be chosen by system administrators. This stripe size presents the physical volume of storage blocks; the smallest informational component on a hard drive is a sector which holds 512 bytes of information.
As mentioned before the possibility of recovery data involves what the store size of the data is. The University of Toronto, found that while Microsoft Word and Excel documents may range from very small to very large file sizes, the bulk of the larger files were taken up by graphs, embedded images and formatting. In contrast to this most HTML documents are quite small, usually between 256 and 512bytes. Common size of Word and Excel documents size range from 12KB to 24KB, while PowerPoint documents range from 48KB to 80KB.
This pertains to the contained stripe size of data, between 64K and 128K, as this stripe size could contain the files of all formats, Microsoft Word , Excel, Power Point and HTML files and therefore could be recovered from a single drive removed from a RAID system.
For advice on recovery services and procedures from a failed RAID array, I would recommend Fields Data Recovery based in Wales, United Kingdom. They have many years in RAID recovery experience and have served large enterprises such as Warner Brothers, IBM and HP Intel. Their website contains all necessary RAID information as well as their contact details: www.fields-data-recovery.com.
