Fire, flood, earthquake—a modern information based business cannot afford to lose its data, nor access to its data. There are standard techniques to deal with normal hardware failures and minor data corruptions and losses. Nevertheless, there is always the possibility of some form of catastrophic failure which such routine precautions cannot address, and which could produce major data loss at worst, or at best system downtime. The traditional approach to data recovery has been to keep offsite tape backups. These can be used to restore to a known point then later transactions can be reapplied. SANs can help to reduce the downtime for the backup window to near zero, but the restore time can be many hours or even days. Use of RAID and/or storage virtualization has a big (negative) impact on data recovery and rebuild times.