I will start with this question in the newbie section.

I have, what I thought, was a stable install of Debian. Up until tonight all was well.
Here are specs first:

Dual P3 550 running the stable 2.6.8-SMP 686 kernel on Sarge.

A few weeks ago I installed VMware sucessfully. I had an XP vm running as well as a DOS 6.22 vm. The catalyst for this event, I think, is my desire to have sound in the DOS vm.

To this point the machine has not had a sound card. With an uptime around 80 days, I turned off the PC and installed a Creative CT4810 (Esonique). Turning the PC back on everything went seemingly according to plan.

Upon booting to the bash prompt (this machine does not run X) I typed lspci | grep audio to ensure Debian saw the card. It returned details about a CT4810 and something akin to a memory address. I thought all was well.

I fired up vmware server console on my Windows machine to begin my DOS vm adventures, but I could not connect. I then loaded the web interface (vmware-mui) from Windows and all looked ok. I started the DOS vm and quite strangely the whole Debian PC restarted.

Now, each time Debian boots 1 of 2 things happens:

1) The boot process hangs permanently "Checking quotas..." (which is immediately preceded by "Mounting local file systems..."

2) The boot process hangs briefly at the same place... but then spits out a variety of things
Stack:
Call Trace:
Code:
... all with too much for me to type here right now.
<0> Kernel panic: fatal exception in interrupt
In interrupt handler - not syncing

I have removed the sound card and the booting remains the same... I only have the 1 kernel to boot with as I removed the default 2.4.x kernel that is installed initially by the Debian installer.

I would like to know if there is a way for for me to salvage this install, without having to install on a new drive, mount this one and then bring everything over.

UPDATE

I guess after so many mounts of a filesystem fsck is forced to run...

So it ran, prompted me to login and remount the drive as rw. I did this and ran fsck again. It found some things and I answered 'Y' to all questions... and then dumped me back to the same interrupt error.

But.. now when I reboot it hangs much earlier:

chroot: cannot run command '/sbin/init': No such file or directory
Kernel panic: Attempted to kill init!

My guess at this point is that the kernel, as well as other files are corrupt on the hard drive.

I have the Debian net-installer CD. Is there a way to copy the 2.4.x kernel files to the drive and then I can edit GRUB to load the different kernel? Am I destined to reinstall on a new drive and bring over what I can from this one?

Thanks.

Use a LiveCD or other Linux bootable CD to boot up your system and then run fsck on the drive.

Thanks for the reply.

I have discovered that 1 stick of RAM seems to have been the cause all along. Until I removed the problematic 64mb stick I couldn't boot into any live CD or alternate OS on this machine

I ended up doing something similar to your suggestion...

I installed a base Debian install on a spare hard drive. Then I mounted the original, problem drive. The entire contents of /sbin was gone. The folder was empty. Something must have been screwed up when it ran fsck and had the bad stick of RAM in it. I copied all files from /sbin on the good install to the /sbin on the other drive and prayed.

Everything started seemingly normally. I am aware there are some symlinks which did not get copied but I will deal with that as needed. At least it will boot now and all init scripts and services seem to run properly.

Thanks again.