|Brien M. Posey|
Although I don't normally do consulting work, I have helped out friends and relatives by doing some work on their networks. On these occasions, I have found that often companies, especially smaller ones, will choose to use a high-end PC as a network server rather than going with a system specifically designed to be a server. A PC will run Windows 2000 Server with no problem. By using a PC rather than a true server, you'll save tens of thousands of dollars, but you'll sacrifice things like hardware redundancy, fault tolerance and some level of performance.
There is a way to get a PC to perform more closely to the performance level of a true server, though. Any time you install a hard disk into a Windows 2000 Server, whether it's a true server or a PC, you have the option of defining the disk as either a basic disk or a dynamic disk. A basic disk is just a normal hard disk with nothing special about it. If you convert a basic disk to a dynamic disk, you can use software to give your makeshift server a serious performance boost.
Dynamic disks allow you to do things like spanning volumes between multiple hard disks, a useful technique when you find a volume to be running low on space. You can also stripe a volume. Striping a volume means that the volume spans multiple hard disks but that each file is actually spread over the disks in the stripe set. This means that performance is dramatically increased because files are read from and written to multiple hard disks simultaneously. For example, if you had a stripe set consisting of three hard disks, then one third of the file would be on each disk. Therefore, the file could be read in one third the time that it would take to read the same file off of a nonstriped disk because all of the drives are being accessed simultaneously.
If fault tolerance is what you're after, dynamic disks offer a couple of methods of protecting your data. One such method involves mirroring the drive. The idea is that anything written to one drive is simultaneously written to another drive as well. Therefore, if a drive dies, you have a backup that's readily available. Another fault tolerant mechanism is called RAID 5. RAID 5 is striping with parity. This means that a stripe set is created, but each drive in the stripe set contains enough extra information that if any one drive in the stripe set were to fail, the stripe set can continue to function. The downside to this is that you lose the capacity of one drive's worth of information. For example, if four hard drives were used in the set, all four hard drives would actually be used, but your stripe set would have only as much storage space as three drives could accommodate. Of course the flip side to this is that you could create a normal stripe set and have the full capacity of each drive at your disposal, but if anything were to go wrong with any of the drives in the set, the entire stripe set would be lost.
Each of these dynamic disk structures has its own specific hardware and software requirements. You can look at your existing disk configuration by entering the DISKMGMT.MSC command at the Run prompt. This will launch the Disk Management console. This console will allow you to view your current disk configuration and to create any of the disk configurations I've discussed.
About the author:
Brien Posey, CEO of Posey Enterprises, is a freelance technical writer and has been working with computers for about 15 years. Before going freelance, Brien served as the Director of Information Systems for a large, nationwide healthcare company. He has also served as a network engineer/security consultant for the Department of Defense. You can access Brien's Web site, which contains hundreds of his articles and white papers, at www.brienposey.com.
This was first published in March 2003