This is somewhat more of a hardware question but it might interest someone
I was installing (with the pci/ide disks) on a system that has very been
running 2.1 for 9 or 10 months (I built it when I loaded 2.1 on it, so its
recent hardware western digital ide hd, asus p5a, k6-2 450), went through
it, rebooted, and was in dselect picking which packages to install. I was
interupted and had to go away for a few hours, when I came back, I didn't
quite realize at what step I was at in the install, and so I noticed the
2.2.17pre6 kernel image, and thought, oh I'd like that too, and selected it.
So halfway through installing the packages, it replaced the kernel, and
still not thinking about what I was doing, moved the modules directory and
hit Y to install the new kernel. Then it continues installing the other half
of the packages.
Almost did that is. It began to get drive write errors, and quite shortly
the system locked up.
I had to cut the power, and start again. But when I began the install over
again, even rewrote the partition table, etc, I kept getting the same write
errors from the harddrive. I then formated the disk with and old dos
bootdisk, and still got the errors when reinstalling. I then rebooted from
the rescue disk, and did a dd if=/dev/zero of=/dev/hda and wiped the disk
totally. made one big partition, fscked it, and did a read check, but stil
got the same write errors during an install attempt.
Pissed, I went to bed, and got up in the morning and the install everything
I can only assume being powered off for 7 or 8 hours caused whatever was
wrong to go away. Is this reasonable? Is there something on the drive, a
buffer of somekind, that could have been hosed by the kernel-crash that was
able to survive a 30second power off, but went away during the long
The last thing to ask me about is hardware, so I'm rather puzzeled.
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.
Share information about yourself, create your own public profile at