Maybe the quickest way to find out is to install manually via PXE. Use the same disk structure etc. you use in the kickstart, stay as close as possible. If that succeeds compare the kickstart log for that install with your kickstart file. It's possible that you have some directive in there that wasn't 100% correct, but failed gracefully before and now it doesn't. I assume you have already watched one of the failing kickstart installs to rule out any missing files or other glitches that would be visible on the install screen that runs thru? I also assume that you at least verified that the partitions are there and filled with reasonably looking files.

my /etc/grub.conf looks fine

it doesn't contain the boot config.

Kai

Jerry Geis wrote:

ok - so I am getting the kernel panic in my (kickstart PXE invironment). I have used this environment since cents 5.2 The PXE starts up 5.6 gets installed that all works fine. then after it reboots I get the kernel panic cant mount root (0,0) This is just one disk SATA nothing special. Same hardware I have always used.

I not have booted into rescue mode. my /etc/grub.conf looks fine. everything else looks fine.

Maybe check /boot/grub/device.map? It *better* have (hd0) /dev/sda and grub *better* have root (hd0,0), not root (0,0).

mark "nevre makes typoes"

On Tuesday, April 19, 2011 05:26:18 PM Jerry Geis wrote:

Jerry Geis wrote:

...

ok - so I am getting the kernel panic in my (kickstart PXE invironment). I have used this environment since cents 5.2 The PXE starts up 5.6 gets installed that all works fine. then after it reboots I get the kernel panic cant mount root (0,0) This is just one disk SATA nothing special. Same hardware I have always used.

I not have booted into rescue mode. my /etc/grub.conf looks fine. everything else looks fine.

I have tried 3 different machines now. All three get the same result.

I have not changed anything in my kickstart environment from 5.5 to 5.6 other than change numbers from 5.5 to 5.6.

What am I looking for? as I mentioned grub looks fine.

Thanks,

Jerry

I looked at all the grub stuff in the editor mode and it looked fine. root was hd0,0 etc...

The "root (hdXXX,YY)" line was never in question (without it no kernel had been found and you'd not gotten the "unable to mount root" panic.

What could have been wrong was the "root=..." part of the kernel line in grub. But as it stands now it seems more likely that mkinitrd failed during install somehow..

I booted in rescue mode and was looking around. ls -l /boot all files are there however the initrd-2.6.18-238.9.1.el5.img is only 50 bytes not the normal 2.6M

How is this getting corrupted???

Sounds like the installer for some reason did not succeed to generate an initrd for your setup. Any clues if you run mkinitrd on the machine manually? (preferably with verbose option)

Except for this manual invocation of mkinitrd it would also be nice to (as a previous poster suggested) run the install a bit more interactively to see what goes wrong.

/Peter

On Tuesday, April 19, 2011 05:45:15 PM Jerry Geis wrote:

Ok - after further looking - I have found this in /root/install.log

error: nash-5.1.19.6-68.el5_61: Header V3 DSA signature BAD, key ID e8562897

I am using a local repo for installations:

Seems the installer fetched a broken pkg from your local repo. I'm a bit surprised the install didn't stop/die there but finished.

In my updates directory nash shows as ls -l nash* -rw-r--r-- 1 537 537 1177729 Apr 14 9:52 nash-5.1.19.6-68.el5_6.1.x86_64.rpm

Is this file corrupt?

If the signature is "BAD" it's defined as corrupt.

/Peter

I can always just remove it and try again. however I'd like to know what happened so it doesnt again.

Any thoughts? Thanks!

Jerry