Originally published in BugNet, October 1998
and reprinted on the MSNBC Technology page

I almost didn’t make my BugNet deadlines this month, because of a death in the family — my hard drive died this week.

It started acting sickly late on a Friday, but I wrote it off as the result of a long stressful week. I didn’t check again till Sunday, and I could tell that it was struggling. Monday morning it lapsed into a coma. I was able to keep it going on life support (Safe Mode) for over two days, and used the time to extract all the data I could onto every available Zip drive and floppy. It officially passed away Wednesday evening. I was at its side when it happened; I heard the drive power down and no amount of coaxing could get it back up again. It had gone to the great bit bucket in the sky.

It probably wasn’t a bug that killed it, but the post-mortem may not come for two weeks, that’s how far backed up the Service Department is. But much of what I learned along the way is very instructive, for bugs or incompatibilities can affect your drives, as well as your valuable data.

Backup, Backup, Backup

Those are the three rules of safe computing. I keep all my data files in one folder and its subfolders, and I was pretty good about making backups of this, using an Iomega Zip drive and their One-Step Backup software, but I wasn’t really fanatical. I have owned five different hard drives over ten years, and this was the first one that crashed. (All the others were in no-name clone computers — this one was in my first name brand, a Compaq.). And I had just done a full backup of all my data file. But one nagging thought occurred to me. I had never actually restored anything off my Zip drives, so maybe it really wasn’t there. I have an old laptop computer, too, so I quickly tried to restore some files, and there they were. It was a relief, but it was probably something that should have been tried sooner.

When all the bad troubles started Monday, I was getting all kinds of error messages. First there were different errors in drivers. The messages metastasized into errors in core things like KRNL386 and USER.EXE. And finally, blue screens of death in Windows itself. I knew these were crucial clues, and I had been writing them all down. Now it was time to call Compaq Technical Support to find out what it means.

Clueless in Houston

Well, I needn’t have bothered. The Compaq technician made a fast stab at looking through his technical database (it took all of a minute) and then he came up with two possible diagnoses. Either all of my system files and Registry had become totally corrupted, or I had a bad hard drive. His solution for the first: he would send me a recovery disk that would reformat the hard drive, wipe out all its contents, and rewrite the system to its state on the day I took it out of its box. It called to mind the colonel in Viet Nam who said, “We had to destroy the village in order to save it.”

Well, everything would be lost anyway if the drive was shot, so that’s when I began my Safe Mode rescue operation, that ended when the drive gave up the ghost. It was a frantic effort, because it suddenly dawned on me how much stuff was in here other than backed up data files. There were passwords hidden in cookies; there were settings that it had taken me months of tweaking to get right; there were old e-mail messages that contained important information; there were about 200 bookmarks of web sites. I hadn’t accounted for any of this in my data backup scheme. And I would have to find all the device drivers for the printer, Zip drive and extra parallel port.

The Aftermath: Dealing with Loss

I couldn’t find the start-up disk for my Zip drive, and the Iomega Web site, while it had an extensive list of drivers, didn’t indicate which was needed for people dumb enough to lose their start-up disk. I called Iomega, and not only did they tell me, but the Tech Support person gave me some tips on how to use the Zip drive in DOS mode as part of the data rescue operation. (And since I had to wait on hold for nine minutes, my call was free. Thank you, lawsuit!) While I had recorded the various DNS settings and switches for my two ISP accounts, I realized that I hadn’t written down the local access number for one of them, and my only record of the password for the other was a row of asterisks in a dialog box. That meant two more calls. And I have been re-registering at a number of web sites where I had neglected to write down my user name and password. So write things down, before they are lost.

I ended up getting a new laptop to “tide me over” (got to deal with my loss some way) and got a surprise when I went to install Microsoft Office 97 from my official CD. There was a rather curt message from Microsoft on my screen, informing me that this was an upgrade version, and there didn’t seem to be any of the precursor products on my computer. After fruitlessly searching through five floppy disk holders in the closet, I finally found my Word 95 diskettes in a file cabinet drawer. Sticking Disk 1 in the floppy assuaged their suspicion that I was stealing from Bill.

The Moral of the Story

You need to be prepared for disaster, be it from a software bug or a hardware flaw. So here is what you need to do:

	Make back-ups on a regular schedule
	Keep all your user names, passwords, and phone numbers in a safe, paper-based place.
	Keep all of the drivers and configuration information for your peripherals together.
	Not only do you need all of your original software disks in case you need to re-install, you may also need the version before that.
	Know how to boot into safe mode, and know what to look for when you get there.

Well, that’s enough for now. I’ve got to go out and review some disaster recovery software.