Re: new Etch install fails to boot
On Sun, Jul 29, 2007 at 11:00:30PM -0400:
> [I wrote that my fresh Etch install calls grub and then stops.]
I've tried the remaining suggestions without luck and am now royally
confused.
1. I installed the grub-disk package and ran the following:
dd if=grub-0.97-i486-pc.ext2fs of=/dev/fd0 bs=1024 conv=sync; sync
There were no errors. I could mount the floppy and read the files. I set
the BIOS to boot off the floppy. The system touched the floppy but failed
to boot off it. I tried this with the box I'm working on and one that is
successfully running Etch. I repeated the test with a second floppy,
which also didn't boot.
2. I tried repeatedly to install grub from the CD in rescue mode. The
install seemed to succeed. I poked around with a rescue shell and
confirmed that files had been installed in /boot/grub. However, booting
the machine still gave the usual message from the dead, i.e.:
Verifying DMI Pool Data ..........
GRUB Loading stage1.5.
Read
3. I reformatted the partitions, reinstalled the base system, and tried to
install lilo. This came up:
LILO installation target:
/dev/hdf: Master Boot Record
/dev/hde1: new Debian partition
Other choice (Advanced)
Note that the first one is for hdf, which is the second drive. I have no
idea why. Anyway, I tried all of these, include Other with /dev/hde and
/dev/hde1. In all cases, I got "lilo-installer failed with error code 1".
Console 4 showed the following:
Setting up lilo (22.6.1-9.3)
mount: /dev/hde1 already mounted or /boot busy
dpkg: error processing lilo (--configure)
subprocess post-installation script returned error exit status 32
Still, I verified that /etc/lilo.conf was there, and there were no grub
files anywhere under /target (including under /boot). I finished the
installation anyway, and found it totally bizarre when a reboot produced
the same output as before, including "GRUB Loading stage1.5". It's as if
this string lives on the MBR and I'm unable to overwrite it.
Practically all I can think of now is trying to install Windows and see if
that's even possible. That's pretty desperate.
In case it's relevant, here's how fdisk showed the filesys in rescue mode:
Device Boot Start End Blocks Id System
/dev/hde1 * 1 12 96358+ 83 Linux
/dev/hde2 4819 5005 1502077+ 5 Extended
/dev/hde3 12 4818 38604195 83 Linux
/dev/hde5 4819 5005 1502046 82 Linux swap / Solaris
I do appreciate all of your suggestions and am sorry this is going on so
long.
Reply to: