When it comes to professional RAID recovery, the approach taken depends on which form of RAID has been configured. For the types of RAID that have been optimised for throughput, the data should really have been intended for very short-term purposes, and initial checks should be performed on the back ups or longer term storage to see if recovery is worthwhile (depending on the amount of data lost). If the RAID has been used wrongly, you may have been storing files that really should have been kept somewhere more stable and recovery requires every drive to be individually recovered. For a RAID optimised for storage robustness (most commonly a RAID 5 arrangement), you can sometimes get away with recovering every drive but one, but this strongly depends on the exact RAID level configured by the system administrator. Experience shows that you cannot depend on the memory of the administrator for these settings – they need to be recalculated from scratch based on the actual content of the drives. Best practice is to attempt to recover every drive as you cannot be certain about the sequence of failure of the bad member drives. It is never wise to use the original RAID controller as this may well be implicated in the failure, it shouldnt be trusted.