Data recovery from redundant complex data storages
Most complex data storage systems ensure data safety by means of redundant recording distinguished by three types: mirroring (providing an exact extra copy of data), parity (application of data conversion for redundancy) and redundancy based on parity and GF-algebra method.
Mirroring is applied in RAID level 1 and its combinations with other data distribution methods. With this method several RAID-components contain the copies of information. Thus, data can be recovered from any copy.
his method is applied to RAID levels 3, 4, 5 and 7. Data can be recovered if only one of the components fails. Generally, to recover information lost from such RAID-systems it's necessary to use data from all other components and parities. This RAID differs in algorithm of data distribution across the components. RAID 3, 4 and 7 use separate components to store parity and the data are evenly distributed across the rest of the components. In RAID level 5 the data and parity are equally distributed across all components. When recovering data from these systems, one should take into account data distribution algorithm and the quantity of active components. If more than one component fails, data recovery is possible only after the components are repaired. It's recommended not to conduct this operation on one’s own and rather take the device to a specialized service center.
Reed-Solomon code (redundancy based on parity and GF-algebra)
This method is introduced in RAID level 6. It provides a possibility to recover lost information even after failure of two components. This is a hybrid system using two algorithms to ensure redundancy. If one component fails, data can be recovered in the same way as in RAID5 using data from other components and parity. The second redundancy algorithm allows recovering lost data if one more component fails.
Equal data distribution across all the components of RAID-systems level 3, 4, 5, 6, 7 enables recovering lost information from a separate component. Before starting data recovery from a RAID-system components assembly is required. UFS Explorer allows virtual assembling of these RAID levels imitating data distribution algorithm applied by controller of the given massive and thus reading information on the file system as it is. If components of these arrays fail, UFS Explorer enables data recovery using redundant information.