This is a worryingly common misconception, and there are several reasons why this could lead to a catastrophic loss of data, which could risk the long term future of a company. RAID (Redundant Array of Independent Disks) by its name suggests a built in redundancy, but the level of protection differs between the different architecture schemes.
Redundancy is a Good Thing
The RAID array scheme used determines the level of redundancy, and data protection it affords. RAID 0, which is simple data striping, provides no level of data protection, but provides very fast access. RAID 1 provides mirroring, while RAID 5 and 6 provide data parity as a means of data protection. A RAID 0 scheme can be combined with another architecture, but in all cases it is still important to have an effective backup solution in place.
What Happens After a Disk Failure?
Apart from RAID 0 striping, the RAID array can still function after a disk fails, and in the case of RAID 6, with the failure of two disks, known as degraded mode. It is alarming how often a RAID is allowed to run in degraded mode without any action being undertaken to rectify the problem. Failure to take action will sooner or later lead to a serious issue. In rare cases it has been known for a failed drive to start working after a RAID has been power cycled, which has the potential to corrupt the integrity of the file system if the data held on the drive is out-of-date. It is therefore important to always replace a failed immediately.
Failures in RAID arrays are the subject of much research, which indicates clearly that when one disk fails, the chances of another hard drive failing within the next 24 hours is very high. RAID arrays contain several disks, usually maintaining those drives in an identical environment. Combine this with the fact that most disks placed in a RAID array are obtained from the same source and batch, the lifetime of the disks are likely to be very similar. The disks in RAID 6 arrays are now often acquired from different sources, but it still leaves the identical environment as a potential issue.
A Rebuild is Fast and Easy
It should be the case that replacing a failed drive and rebuilding the array should be fast and easy. Apart from the potential issues outlined, with the potential failure of further drives, care must be exercised to avoid damaging the data held on the working disks. With multi-terabyte disks commonplace in RAID arrays, it now takes a lot longer to rebuild the array onto a replacement drive, which increases the chances of an additional disk failure causing the rebuild to fail.
Your data is of the utmost importance, so be very careful who you allow to attempt the rebuild of the RAID array. We have seen cases of well meaning IT professionals, who have caused damage to the integrity of the array, by replacing the wrong drive, and rendering the data inaccessible. Any procedure that is undertaken, should be properly documented, with all disks removed and being rebuilt, labelled clearly. If you are not certain what to do, it is best to avoid even attempting a rebuild, and trust your data to a professional data recovery company. Most mistakes happen as the result of panicking, and failing to use clear rational thought. If you don’t have a backup plan in operation, such actions could in the worst cases lead to a total loss of your data.