Losing a server with a RAID configuration especially if you are a business with mission critical data or other such type of very important information will strike fear into the hearts of all responsible for data backup and preservation within the company, this situation is particularly embarrassing and potential litigious if the RAID or server contains backup Set’s of other clients data, in other words you are an online storage backup and management company, no worries you might think RAID is indestructible and never fails right?, Wrong, it does and often without any warning at all.
RAID Redundant Array of Independant disks has many configurations and permatations, lets take a brief look at some of the options;
RAID 0: This is not really a RAID in the classical view or description as there is absolutely no redundancy whatsoever, they data is striped in blocks across both disks from anything from 4k to 512k, although 64k is a common configuration, as the data is ‘split’ if one drive fails catastrophically so will your data set as only ‘half’ the data will be available on the working disk.
RAID 1: Data is written in an identical manner to both disks symotaneously and is classed as a mirror set.
RAID 5: Requiring 3 disks or more offer distributed parity across all disks if one disk fails it can be rebuilt from the remaining disks, this configuration is very popular as it offers a level of protection and some operating ead and write performance.
JBOD: Just a bunch of disks this is not a RAID configuration at all it simply is one large volume configuring all disks into one large system.
Let’s take a look at failures and causes of such to the server and storage setup.
External issues such as power fluctuations can seriously damage your hard drives, a very good preventative measure is a UPS uninterruptible power supply, this device will stop excess power from reaching your hard drive, also it will power the system in power outages, a truly excellent piece of kit for any computer user and modestly priced, especially when you consider the value of your data.
Heat from within the computer system can be a serious problem for hard disks as this can cause the hard drive to degrade, a simple but very effective hard drive cooling kit will keep the drive from over heating and keep the system operating a optimal performance.
I can not over emphasize the importance and need for the items mentioned above as they require little time and resources to implement and are inconsequential in terms of the value of the data.
Failure of the RAID may be instant and without warning at best you may have a S.M.A.R.T. warning or a basic RAID offline or RAID degraded error, please do not attempt to rebuild the RAID as this will compromise the data at best and potentially destroy the data at worst.
Each hard drive should be professional diagnosed to operability and then cloned from LBA 0 to the end of the user data, this should be performed on all hard drives, including any physical work that may be requyired in a clean room environment for physical damage to the HSA.
Rebuilding the file system using a hex type editor will then be the final stage with folder and file structure this can and often is a very complex area, discovering the drive order, start drive, block size, file system and user data area start, it may be NTFS with MFT master file table, or Linux, Unix Mac type structure an experienced engineer will be able to reconstruct manually such data.