One of the major benefits of RAID 5 arrays is that it is possible to recreate the data that is lost when single disks within the array fail. While in theory this all looks nice and easy and that all that is required when a raid drive fails is to plug in a new one and wait for the data to recreate itself, in practice it takes a much greater understanding of the system in order to be able to carry out a successful RAID rebuild.
There are two main types of failure that can happen within a RAID system. One is a physical drive failure that only affects a single drive and allows the utilisation of a replacement drive. In contrast to this, a logical corruption cannot be sorted out by replacing a drive within the disk array. In the event of a physical failure there are two ways in which the data from the damaged drive can be replaced.
The first of these is in pre-failure replacement. A RAID array is set up to monitor a number of parameters that can indicate when a drive will shortly fail and so the drive is then highlighted as requiring a replacement. If another drive is available immediately this can be used to copy the data from the failing drive before it fails and while it is doing this it will also become the active drive so any new data will be written to the replacement drive rather than the failing drive.
If the failing drive breaks before the above can be carried out then post-failure replacement will take place. In this case the data on the failed drive is lost and must be rebuilt from the existing parity data that is stored on the rest of the drives within the array. This is not a simple swap as there are many calculations that have to be carried out in order to successfully rebuild the data so that it is usable. If these calculations are wrong then the rebuilt data will be corrupt and will not be accessible.
It is not only incorrect calculations that can lead to corruption of the rebuilt drive. If an inappropriate disk is used to replace the drive that has failed then not only will the data on the failed disk be corrupted, any recently written data will also be affected as the redundancy in the array is restored by the system. This type of corruption may not be immediately noticeable as you are likely to find that come files appear to be unaffected and will open normally. If you find later that your files are not opening properly your system will already be corrupted and your data is likely to have been permanently lost.
So if you are attempting to carry out a RAID rebuild consult our RAID recovery engineers who will be able to advise you on the best practice for recovering the data from your failed drive without corrupting it. Our advice would be to image all disks before you attempt a rebuild, that way you are safe guarding your data if the rebuild does not complete successfully. If this is not possible, and you maybe have already completed a rebuild that has failed then contact us for advice and a fee no obligation quote. We will use our 14 years of knowledge and experience to provide a quote and timescale for a professional recovery of your data.