Disk subsystems are therefore designed to provide redundancy. This means that the subsystem will try to insure that it has at least two copies of the data. This might mean attaching a second disk to a server disk controller, and writing data to both disks at once. That way, if one disk fails, the other will have a good copy of the data. Software or hardware can be used to direct this process. Netware calls this arrangement disk mirroring. If both an additional controller and disk drive are added, the system can be protected against a controller failure as well. Novell calls this disk duplexing. A side benefit of these duplicate disks is that it becomes possible to split a disk access between two disks, retrieving half the data from each one. This speeds up the disk access.

What does it do?
One of the most common terms in connection with disk subsystems is RAID and RAID levels. RAID stands for Redundant Array of Inexpensive Disks. This is an extension of the disk duplication described above. In a RAID array, a group of hard disks are assembled together and arranged through controller intelligence, or software control, to provide a level of data duplication. RAID arrays may also provide duplicate controllers, and multiple power supplies. Another frequent design feature is the ability to swap a defective disk drive or controller without powering the system down.

RAID levels are typically referred to as RAID-1 through RAID-5, though some creative marketing efforts have produced some references to higher levels. The levels refer to the sophistication of the data redundancy. RAID-5 is more complex than RAID-1.

How is it used?
A disk subsystem is obviously the heart of a file server. Without disk storage, a server has no files to serve. Beyond file storage, a disk subsystem may be designed to improve reliability of the server by providing fault tolerance and increased access speed. A more recent development in storage technology is termed Hierarchical Storage Management (HSM). This technology combines hard disks (low capacity and high speed) with optical storage (higher capacity and lower speed) and tape (highest capacity, lowest speed). Based on the frequency of access, a set of files may be migrated to and from different media automatically. This insures that files with frequent usage are stored on the fastest access device, while less-used files are stored on slower but less-costly media. It also provides a means to automatically restore files to faster media based on demand.

Where do I get more information?
Contact your local CBV Office.