Hard disk drives remain the most popular form of storage device used within a RAID array. Hard disks are constructed from four main components, all of which must operate correctly, in order for the drive to function properly. These four components consist of the hard disk platters, the read/write heads, motor assembly and the hard disk controller, each of which perform essential specific functions within the drive. If any of these components fails it can lead to erratic behaviour, which could cause additional problems.
If a failed drive is not replaced or the RAID does not provide redundancy a professional data recovery solution will be required. If a failed drive is allowed to remain operational within a RAID array, any further failures could lead to a catastrophic failure of the system, which could result in the loss of data.
Hard Disk Drive Platters
The platters are probably the most important part of the disk, as they contain the stored data in magnetic form. Platters are most often constructed from an aluminium, glass or ceramic substrate, where both sides are thinly coated with the magnetic recording layer. This recording layer is most often made using a vacuum deposition process called magnetron sputtering. This recording layer consists of a complex layer of metal alloys, which are laid down to optimise the magnetic media layer. A protective carbon-based layer is then put over upon which a lubricant overcoat is deposited. Buffing processes are then used to remove any defects, as extreme smoothness, durability, and perfection of finish in order to achieve the required properties of a usable hard-disk platter.
Drive Read/Write Heads
The read/write heads used in a drive move above the disk platter and convert the magnetic field stored on the platter into electrical current (reading the disk) or transform an electrical current to a magnetic field (write to the disk). The hard drive heads are said to ‘fly’ above the surface at a height which may be as small as 3 nanometres. The lower the ‘flying’ height which is used the higher the magnetic recording density which is achievable and thereby higher the capacity of drives which can be constructed. The slider which contains the read/write head uses an air bearing, which maintains the ‘flying’ height at a constant distance due to air pressure causing them to float above the surface. If the read/write head should hit the surface of the disk platter, it will result in a catastrophic head crash, after which data recovery will be required.
Spindle Motor Assembly
The spindle motor assembly, sometimes referred to as the spindle shaft, performs the important task of spinning the hard disk platters at a reliable and constant fixed speed. The speeds used in a drive depend on its application and can range from 4800 rpm through to 15000 rpm for high end SCSI hard disks. Heat and vibration are the biggest risk to the spindle motor which could cause a failure, especially within a RAID array as the disks are continuously in operation. Disk within a RAID array operate within the same environment, often acquired at the same time, so when one drive fails it is quite common for another to fail soon afterward. We receive many RAID systems where multiple hard disks have failed, leading to RAID data recovery being required.
Hard Disk Controller
Without the hard disk controller, the hard drive will not operate, as it performs several important functions, all of which are essential for the proper functioning of the hard disk drive. The controller ensures that the spindle motor assembly spins at the correct speed, the read/write heads are positioned correctly during each input/output operation. It also must operate as the interface between the hard disk drive and the RAID controller board or host adaptor installed in the computer.
Hard Disk Failures Need Data Recovery
When any component fails, it is essential that the drive is powered off as soon as practicable. This is to avoid a possible cascade of failures, which in the worst cases involves the read/write heads impacting with the disk platters. In a RAID without redundancy all disks should be sent for data recovery. In a RAID using redundancy it is normal practice to replace the drive and attempt to rebuild the data to the replacement disk. The risk of a further failure occurring during this rebuild period is high, so the original drive should be securely stored and labelled. This original drive could prove essential during data recovery, should be required following a rebuild failure. Failed drives especially those used in RAID arrays should be worked upon by a professional data recovery company, such as DiskEng, who can address all the problems encountered.