Downtime on computer networks is an expensive proposition for companies relative to size; the agony of not being able to work with the data can cost companies around $70k to $80k per hour in many cases. Some of the problems may have wiped out portions of data and the hard drives may have to be handed over to a data recovery service, many of whom can affect recovery in less than 24 hours.

Redundancy is crucial when storing data so larger companies would use RAID, which is an example of storage virtualization to increase uptime. Multiple disk drive components are combined into a logical unit where data is distributed in different levels, called RAID levels. RAID 1 disks can be read by the operating system as if it were one single disk; if there is damage the data can be recovered using data recovery software but when RAID 5 and other distributed arrays fail the recovery procedure become complex.

The modern drives of RAID arrays have error recovery algorithms that can recover and re-map the data it can't read, but it can take up to a minute; many RAID controllers will drop a drive from the array after only eight seconds so the array becomes vulnerable as it drops good drives that it didn't give enough time to go through the internal error recovery. Problems can also arise from failures of the automatic re-boots and a myriad of other complex issues with the various incarnations of RAID.

An operator with the skills to set up and monitor a RAID array should be able to prevent failures by detecting problems ahead of time and knowing which drives are failing. Of course the operator should know the RAID configuration inside and out, know how to replace drives and be able to initiate a rebuild of the array. By verifying that the array is not in a degraded state, uptime is maintained and data loss is prevented. Even then the data can be at risk from software failures and virus attacks. Every data loss is unique so RAID data recovery services must carry out a custom retrieval. They won't fix the RAID array however.

For those who find the cost of RAID impractical, remote backup, also known as off-site backup gives a company access to an enterprise class backup solution that can give protection against data loss mortality. It works by replicating between two data centers so that the data is always up-to date and available. This is a cost-effective method to avoid human error and the emergency expense for disk recovery should natural disasters like hurricanes, fires, and floods destroy tapes and disks.

Sabotage of servers carried out by disgruntled employees on their way out or inside subversion for illegal profit or to hide information is responsible for 70% of data loss. If the operator or someone else with access to the RAID array that has malevolent intent decides to get rid of data, a catastrophe is easy to create, and possibly beyond the capabilities of forensic file recovery experts. Someone familiar with the systems will also destroy the backup files; RAID is not a substitute for backup, it's intended to increase uptime rather than to be a data archiving method.