Problem with setup_harddisks

[Axel Freyn]

On Sat, May 03, 2008 at 08:19:56PM +0200, Michael Tautschnig wrote:
> Hi!
> 
> [...]
> > 
> > As disk-config, we are using:
> > disk_config disk1
> > primary  /boot/         10-100                  ; boot
> > primary  -              preserve2
> > logical /               9000-14000      rw,errors=remount-ro ;  -j ext3
> > logical swap            400-600         rw
> > logical -               preserve7
> > 
> > >From that, the following partition.hda - file is created:
> > 
> > unit: sectors
> > 
> > /dev/hda1 : start=        63, size=     79569, Id= 83, bootable
> > /dev/hda2 : start=     80325, size=  20482875, Id=  c
> > /dev/hda3 : start=  20563200, size=  57560895, Id=  5
> > /dev/hda4 : start=         0, size=         0, Id=  0
> > /dev/hda5 : start=  20563263, size=  20635713, Id= 83
> > /dev/hda6 : start=  41199039, size=    907137, Id= 82
> > /dev/hda7 : start=  42106428, size=  36017667, Id=  c
> > 
> > which looks quite reasonable - except may bo for the partition /dev/hda4?
> > 
> Hmm, yes. I don't know whether this is a bug in setup_harddisks or some problem
> with your prior partitioning. To rule out the former, could you try adding
> another primary partition (maybe as small as 16MB, just for the sake of
> debugging this? It shouldn't affect the 7th partition).
Sorry, that's not so easy - thanks to our special hardware to protect the
Hard-disks in Win98 I can't create hda4 without destroying the logical
partitions ...
> 
> > 
> > But setup_harddisks seems to have a problem with that:
> > 
> > Creating partition table: LC_ALL=C sfdisk -q  /dev/hda <
> > /tmp/fai/partition.hda
> > Warning: The partition table looks like it was made
> >   for C/H/S=*/255/63 (instead of 77504/16/63).
> > For this listing I'll assume that geometry.
> >
> [...]
> 
> Well, this might be the other possible cause: sfdisk might be confused about the
> geometry; usually, this requires adding the --force option to sfdisk (by adding
> export sfdisk="--force" to an appropriate .var file in class/).
"--force" didn't change anything for our problem.

Thank you for your remarks and proposals - we decided to take the "safest" way:
creating a hook to switch of the partitioning completely. Like that, everything
is working fine.

Axel