07 September 2004
On Saturday night, my notebook locked up, so I rebooted it, and it just failed with a kernel OOPS somewhere in the VM code.
I suspected a corrupted swap space, so I booted the Knoppix CD, recreated my swap space (
mkswap -c), and sure enough, there were a couple bad blocks in the middle. It flagged them accordingly and works around them.
Following that, I needed to do a whole
reiserfsck --rebuild-tree on my root partition before the machine would successfully boot. Luckily the bad blocks are in the swap (which can flag and skip them) instead of the reiserfs (which cannot).
Once bad blocks start showing up, the drive is on its way out, so I created a login at IBM's website, logged a request, they called me back about 30 minutes later, but I missed the call. I called back on Monday night around 10pm, talked to a tech, and they're shipping me a new drive, no charge, and a return-postage box included. Disregarding the hassle of hardware failure, IBM is making the whole thing pretty easy -- no sitting on hold, and no low-tier tech support insisting I conduct a bunch of tests I already did.