Backup: Why and How? Part 1

By: Walter Metcalf
Date: 08/25/99

Backup is a touchy subject. Most people will feel somewhat guilty if you even mention the word because they either aren't backing up at all, doing it when it's convenient, or are attempting to do it but aren't sure they are doing it effectively. The purpose of this article is to try to convince you that a regular system-wide backup is essential, briefly mention three of the procedures suggested in the literature, and finally show you a procedure that is very secure and very painless--an unbeatable combination!

First then, let's look at why it's necessary. Before examining this further, however, let's look briefly at how hard drives are made. At their most basic, hard drives are a series of circular plates, not unlike flat dinner plates, stacked with a small space separating each of them from the other. When power is applied to the drive, tiny arms with double-sided magnetic heads at the end are inserted into the spaces between each plate. Understand that the distance between the heads and surfaces is less than 1/800 of the thickness of a human hair! The heads are aerodynamically shaped and are kept away from the surfaces by the the airflow between the head and the surface of the plate caused by the rotation of the surface in exactly the same way an airplane's speed keeps it from crashing into the earth. Finally there are some very sensitive, sophisticated electronics, including one of more microprocessors, controlling every aspect of the hard drive, data conversion, and communication with the computer. (For more information on the construction of hard drives, see Storage Basics, Part 2 Chapter 3 by Quantum.) What amazes me is not that drives crash occasionally, but that they work at all!

Hard drives can also be divided into two classes: those that have crashed, and those that will. I should define what I mean by "crash." Crashes can be "hard" or "soft." A hard failure involves physical damage of data of some part of the drive mechanism--a blown chip or severe damage to the heads or surfaces. In these cases you either trash the disk or send it to a data recovery expert--in any case the drive will not be useful again. A soft crash ( I include operator error in this category) involves mass data destruction, such as clearing the partition table, which serves as the drive's master data index. In these cases, the drive can be used again (perhaps after partitioning and formatting), but all or most of your data is gone. In both types of crash, the data is gone, and without a sold backup recovery, even if possible, is going to be expensive.

What's important to realize is that if you're running a business, the cost of the drive is irrelevant. Drives these days are cheap: for example, as this is being written, you can purchase a IBM 25 GB Ultra ATA drive for $271.95 at NECX direct order. Your data and the time it represents is worth much, much more that that.

A friend of mine services a small video store. The owner was making an attempt at one of the procedures discussed in Part 2 of this article, but then the monkey virus got into his system and rendered his entire hard drive unusable. Only after my friend asked for his backup, did they realize how woefully inadequate the owner's procedures had been. When the drive was finally reinitialized, all his records, rental, accounts due, payables, etc. were gone. For several days he had to record his rentals in a 3-ring binder! It was six weeks before his records were restored, although I don't what he's going to for a yearly statement for the tax department because some of his financial information was not recovered. In addition during much of that six week period he was completely at the mercy of his customers and his suppliers because he had to depend on them for much of his information.

Don't say, "It won't happen to me," because it will--it's only a question of when. It's happened to me all too many times! That's part, but only part, of why I am so adamant about backup procedures. There are just too many things that can happen to be complacent. Here are just a few:

  • Lightning strikes in the vicinity of your computer, which can cause massive power surges.
  • Power outages caused by your local hydro-electric company.
  • An inexperienced operator accidentally resetting or rebooting your system during a disk operation.
  • An operator issuing the wrong command and deleting key files. (I seem to be particularly prone to this kind of damage to my own system!)
  • A disk component simply dies because of wear and tear caused by old age.
  • A system/program malfunction causes your system drive to be unbootable.

I've had all the above happen and more. Before I developed a secure backup procedure it could several days before I was running again, and perhaps weeks before I'd managed to get most of my data back in place. Some of it is gone forever. Now usually after an hour or two at most, I'm back up and running.

I hope that by now, if you don't have a dependable backup system, you're convinced you need to implement a regular protection system. Next week we'll look at the some of the procedures commonly used, and then I'll describe one procedure in detail and give you some suggestions regarding software and hardware available to make the job painless.

Unless otherwise noted, all content on this site is Copyright © 2004, VOICE