Backing Up Data
Recently a friend who has a one-man business realized that his computer’s hard drive was full. He purchased an external hard drive, moved all the data to it, and erased all the original data.
Now, it is a good rule of thumb that if a new electronic device is going to fail, it is likely to do so within the first fifty hours. In this case, when he turned on the computer the next day it did not recognize the new drive. He called the manufacturer, who said it was obviously a Microsoft problem related to the USB driver, and not their problem. He called Microsoft, who spent a substantial amount of time walking him through replacing every driver that could be replaced. Still no joy.
He then connected the drive to someone else’s computer, which also couldn’t see the drive. This was a clue that it was a drive problem, not a computer problem. At this point he did a query online, and discovered that quite a number of people had been suffering this same problem with this particular model drive.
He ended up taking the drive to a data recovery company. They will attempt to re-build the drive in a clean room, and move the data to a DVD.
If he is lucky enough to recovery all the data, it will cost him a minimum of $1000, plus of course the cost of the drive he just bought, plus the cost of whatever other drive he gets. If he does not recover the data he is, for all practical purposes, out of business, as he has lost virtually every business record from his entire career.
Our friend’s situation is not much different from that of many small businesses, the only difference being that his luck ran out. So what did he do wrong, and what could he have done to minimize this catastrophic – an possibly irrecoverable – loss of information?
For a start, he needed to back up his data. Backing up data has two parts. First, you need to have at least two copies of all the data, and at least one of those copies needs to be stored far away. Why is this? Well, imagine that you had a business in New Orleans, and kept one copy of your backup data on the bookshelf and the other at your home, neither of which survived Katrina. Fat lot of good the backups would do you.
As an example, we are a small business, and have relatively little data, all of which is stored in encrypted virtual drives. The encrypted virtual drives are regularly copied to an on-line storage facility (we currently use iBackup – see the August 2006 issue of ÆGIS or http://www.ibackup.com). In addition, these files are copied onto flash drives that we carry with us. In addition, they are regularly copied onto DVDs, with a copy stored in our office and a copy mailed to the other side of the country. An awful lot would have to go wrong to make all our information permanently unavailable.
As noted above, if an electronic part is going to fail, it often does so within the first 50 hours. This means that you would be prudent to avoid committing yourself to a new computer or hard drive, getting rid of the old one, until it has been running for a few days. Additionally, you can help assure its continued functioning by using an uninterruptible power supply, better known as a UPS.
A UPS is a device that allows you to run the computer off of a battery, which is constantly being recharged. If power fails, it sounds an alarm, and you can shut down the machine. Or, with most UPS, you can connect it to the serial port of the computer, and if the power fails it will graciously shut down the computer for you. The advantage of a UPS is that the computer receives consistent power – it is running off a battery – independent of any fluctuations of power coming into your home or office. As it turns out, one of the big causes of mystery computer malfunctions is power fluctuation. Note that there are also standby power supplies that switch to the battery if the power fails. These are the more common, and less expensive. They do condition the line, and certainly are fine if you can’t find a cost effective online UPS.
Additionally, hard drives have a tendency to eventually fail: They are, after all filled with rapidly moving parts. Most hard drive vendors provide programs that will allow you to check the functioning of the hard drive. We ourselves use SpinRite (http://www.grc.com/sr/spinrite.htm – see the September 2004 issue of ÆGIS). In essence, SpinRite reads each sector of data and re-writes it. If there is a problem with the sector, it marks the piece as bad and relocates the data elsewhere. This has saved us on several occasions where a hard disk had gone bad, allowing us to recover the data and keep running at least long enough to back up everything and put in a new drive. If one track is bad, we take note. If more start appearing, we get a new drive.
How about a brand new drive? The first thing we do, even before we load software, is run SpinRite. If the drive passes, we start loading our software. We then run SpinRite every week. We recommend you do the same.