IBM DeathStar / Down in the Dumps

I don’t usually write about my mood, but today I woke up feeling really down. Thankfully, I haven’t been sent into massive panic attacks over the issues described below. (One of the ironic points of living with an anxiety disorder and being a computer Geek is that a trigger point for my anxiety is problems with my own personal computer.) I am grateful that God convinced me to make a critical backup at just the right time, so everything important is intact and backed up.

It looks like I’m going to have to replace the old IBM drive on my primary computer. I noticed last night that it was an IBM DeskStar drive, which my friend Charlie Dunn said was so prone to failure they nicknamed them “DeathStars” when he was working in IT.

At first it looked like I had a problem with overheating—and that still may be the problem—although it seems to be getting worse. Last weekend Nichelle compalined that her e-mail was missing for a period of several months, but after rebooting the computer, everything was fine. I took the precaution of backing up our photos (most of which are already on CD as yet another backup), financials, and documents folders to a nonused partition on a drive I’d added in December.

A couple of days ago, I logged in to the computer and was notified that numerous updates were ready to be installed to Java and Windows XP. These updates had already been installed, and it was rather obvious what had happened—the primary drive, from which we have been booting for nearly all of the lifetime of the computer had disappeared, and the machine rebooted to an imperfect Windows setup on the second drive. (That setup was toasted, apparently, by a bad power supply, now replaced, that caused some data loss.)

Part of the time the IBM drive isn’t being recognized at all. After a certain amount of time—perhaps due to overheating, perhaps due to another type of drive failure, it won’t be recognized at all. Rebooting the system sometimes catches it and boots to the original drive, especially if the whole system has been off for a while. On a really hot day, or after a period of high disk activity, the drive will fail again and the system will reboot to the bad partition.

I’d like to try copying the image over to a new drive, using Ghost or whatever comes with the new drive (I’m leaning toward Seagate), as it would get me instantly up-and-running without having to reinstall. Of course, I do not have that much hope that the old drive will “hang in there” long enough for such a data-reading-intense operation to happen.

If that doesn’t work, I’m willing to bite the bullet and do a reinstall of XP Pro. However, there is a drive size limitation (I think it’s 120 GB) in the initial release of XP Pro (pre SP-1), which is what I have. Installing from an SP2 CD would be even sweeter, and less prone to failure than upgrading to SP2 from a previous version.

:: sigh ::

Editor’s note added June 13, 2005:

It looks like the drive is only flaking out when it overheats. I think it may fail soon, but letting it cool down with the machine off so far makes it recover. In the mean time, I’m keeping my data backed up daily, and when the drive does fail, I can boot to my “February” installation, and still run all my apps, so replacing the drive is now less urgent, and I can plan for this at my convenience.

Editor’s note added June 21, 2005:

No, it’s not necessarily related to overheating. Or, if it is, the drive is “overheating” at a temperature far below what its temperature sensor would indicate shoud be a problem.

Last night I was running at a reasonably cool 35 degrees C, when the whole things went kablooie. At first I had the usual trouble recognizing the drive. Then, when I finally got it to recognize, it was dog slow, and told me that Windows couldn’t boot because windows/system32/system was missing or corrupt. I had just finished backing up everyone’s e-mail, pictures, etc., except mine.

This morning, after leaving the system off overnight, I got the drive to be recognized by leaving it running for a few minutes, and everything seemed to be fine.

So heat may be playing a factor, but I think it’s time to just retire the drive.

I’m also looking at putting together a new system that would kick butt for as little as $500, but I also need to get this one back to a reliable state.

Editor’s note added June 27, 2005:

Wednesday night after church, I went out and grabbed a 120 GB Seagate Barricuda drive to replace the boot drive. On Wednesday, the IBM DeathStar drive seemed to fail completely, and I was looking at rebuilding my system.

On Thursday evening, I decided to give the IBM drive a few more tries, and got it to recognize at startup!

Hurriedly, I rebooted with the Seagate drive, hoping to be able to copy my IBM drive over to the Seagate drive, which would become the new boot drive, and save me from having to reinstall everything.

Seagate’s drive setup software is supposed to be easier to use than Western Digital’s. I actually didn”t find that to be the case. They both are about the same, but there were some nonintuitive points that I had to keep hitting the help file and documentation for. Once nice thing (which I did not need) is that Seagate lets you print installation instructions that are customized for what you’re actually doing.

It took hours for the drive to copy, mostly due to the read errors (all were being corrected, but it was having to perform each read multiple times) that SMART was reporting. Copying also appeared to be slow because of Seagate’s verification process, but getting an exact drive copy is far more important than getting a fast copy.

I only found three problems:

  1. Trying to access the System Restore tab crashed. It turned that is always caused by a drive copy process, and all I had to do to fix it was disable the System Restore service, then re-enable it.
  2. The next problem was a missing FrontPage icon (the link would work, but the icon itself was the default rectangle). It was looking for something under installs that clearly wasn’t around any more. I simply replaced the shortcuts with new ones.
  3. Lastly, the icons on my own logon are weird. Unless I run in 32-bit color mode, I get ugly 2d, 8-bit icons. It seems to happen only on my account. I find it disturbing, but it can’t be described as much more than a cosmetic issue. Some of the menu colors are off as well. Again, it’s no big deal, but it does bother me.

Other than that, the machine is working perfectly. My friend Phil Luchon came over and brough his machine. We played WarCraft and StarCraft. The Lego Star Wars game, which the kids play every moment we will allow (one hour on weekedays, 2 on weekends), which had been very prone to crashing and lockups, has not had a single failure in 3 days.

Everything involving disk access is much faster on the new drive: booting, loading big programs, virtual memory swapping, etc.

3 Replies to “IBM DeathStar / Down in the Dumps”

  1. Yikes! I think reinsalling the OS is the way to go. Hopefully you can get your hands on an SP2 install disk, but would your serial number work with that?

  2. Should you need to re-install the OS.. which is probably a good thing.. and if you can get your hands on a machine that works with a CD burner in it.. You can slipstream SP2 onto / into XP so you can install it all at once. See http://www.winsupersite.com/showcase/windowsxp_sp2_slipstream.asp

    I’ve done it.. works great.

    Also.. I did recently read that those IBM Drives were being recalled or you could get money back on them. Not sure of the details, mostly because it didn’t pertain to me.. but fire up google baby.

  3. Yes, thanks for the tips on SlipStreaming. I actually started the process a month or two ago, but couldn’t finish it on my XP Box, because I never installed CD burner software (Why bother on XP?) that would let me make the CD bootable.

    I think what I’m going to try to do is get a ghost of the current O/S on a new drive, install a boot manager (like BootIt), and be able to choose between my ghosted boot and my under-construction setup.

    I’ll have plenty of room on the replacement drive, and I won’t then have the fear that I’ll have missed backing up something that later turns out to be vital.

    Plus I can continue to work uninterrupted on one boot while getting everything installed on the other.

    I doubt my IBM drive is still eligible for a refund. I’d have to check my records, but it was an OEM purchase (came with the machine), and is at least 3 years old.

Leave a Reply

Your email address will not be published. Required fields are marked *