On Fri, 2005-11-04 at 19:49 -0500, Sam Drinkard wrote:
This afternoon, after my fiasco with the other machine, I decided to bring the drive home and install it in the primary machine here, boot from the normal drive, then mount the 4.1 i386 drive so I could make changes and check things. What happened next was totally unexpected. After decompressing the kernel and starting the boot, there were all kinds of error messages scrolling off the screen, and so fast, I could not see what it was all about, but the boot halted at some point. After seeing that, I decided maybe it was not a good idea after all, but I'd done the same thing with an i386 version 3.5 with no problems.
Is there some reason why I should not be able to boot from my normal drive, CentOS 4.2 x86_64, but still have the 4.1 i386 version on a secondary drive, even tho it was not automounted?
Yes ... If you allowed anaconda to label the drives, then both of the drives probably have the same label.
You can change your /etc/fstab to mount the device names and not the labels (and make sure your /boot/grub/grub.conf uses device names and not lables)
or change the labels on the second drive using e2lable and either change it back before you boot from it later, or change the references in the /boot/grub/grub.conf and /etc/fstab.
Also, while on the subject, I found out that a FreeBSD drive is not mountable under Linux. It recognized the drive as a Sun ufs drive. Don't suppose there is any way around that problem is there?
I think that the centosplus kernel can read, but not write to some versions of FreeBSD (ufs) file systems.