🚀 go-pugleaf

Bit thanks to all those who helped me out in this saga, it is _such_ a relief
to have lilo up and running again.


Graham


Here is the solution to the problem.

For starters here is a copy of an fdisk -l showing the new hardisk as follows :

>    Disk /dev/hda: 16 heads, 63 sectors, 6232 cylinders
>    Units = cylinders of 1008 * 512 bytes
>    
>       Device Boot  Begin   Start     End  Blocks   Id  System
>    /dev/hda1           1       1     813  409720+   6  DOS 16-bit >=32M
>    /dev/hda2   *     814     814    1017  102816    6  DOS 16-bit >=32M
>    /dev/hda3        1018    1018    1148   66024   82  Linux swap
>    /dev/hda4        1024    1149    6232 2562336   83  Linux native
>    Partition 4 does not start on cylinder boundary:
>         phys=(1023, 15, 63) should be (1023, 0, 1)

The problem with this scenario was twofold.

In my /etc/lilo.conf file I have the line ===> boot=/dev/hda4, so lilo was
being installed on a cylinder that was above the 1024 limit. This would have
grave effects on the ability for lilo to perform, as well as the fact that I
had hda2 marked as the active partition.

Double stupid.

What follows is an fdisk -l on the old hard drive.

>   Disk /dev/hdc: 16 heads, 63 sectors, 1654 cylinders
>   Units = cylinders of 1008 * 512 bytes
>   
>      Device Boot  Begin   Start     End  Blocks   Id  System
>   /dev/hdc1           1       1     691  348232+   6  DOS 16-bit >=32M
>   /dev/hdc2         692     692     935  122976    6  DOS 16-bit >=32M
>   /dev/hdc3         936     936     968   16632   82  Linux swap
>   /dev/hdc4   *     969     969    1654  345744   83  Linux native

You will notice that hdc4 (hda4 in the old scheme) is marked as active, this 
worked fine because in this case hda4 starts on cylinder 969, well within the
ability of lilo to see it, as well as the fact that hda4 is marked as active.

Seeing as the /etc/lilo.conf has the line boot=/dev/hda4 in it, the old hardisk
setup is fine.


What was the solution ?


I tried installing lilo on /dev/hda2 (boot=/dev/hda2 in the /etc/lilo.conf file)

This should have worked as /dev/hda2 starts on cylinder 814. 

Unfortunately doing this destroyed /dev/hda2 as a valid dos partition. 

What I thought then to do was to wipe /dev/hda2 and then install it as an
extended partition. (It was originally a primary partition). After this was
done I create a dos partition within the extended partition (now hda5).

I re-ran lilo which happily installed itself to /dev/hda2, this time around
however the dos partition hda5 was left intact.

Now lilo works flawlessly.


Big Thanks to all those Helped.
Graham Daveney
graham@daveney.matra.com.au


P.S. I did have to do a fdisk /mbr underneath dos to get the first dos partition
working again (hda1).