Re: Bug#341135: Problem installing on I2O disks
On Tuesday 29 November 2005 06:16, Markus Boas wrote:
> Am Montag 28 November 2005 17:00 schrieb David Goodenough:
> > Package: Installation
> > Version: debian-testing-i386-businesscard.iso downloaded 28th Nov 2005
> >
> > I have now tried both the AMD-64 and I386 versions of the installer
> > trying to install Debian on a box with a DPT (now Adaptec) I2O
> > controller.
> >
> > The AMD-64 build tries to load the wrong driver, as the dpt_i2o driver
> > does not exist for AMD-64, rather the more generic and newer i2o_block
> > driver is needed.
> >
> > Once the correct driver is loaded the disks will partition apparently
> > correctly (I have no other machine which will accept these disks so I can
> > not tell if they have been formatted correctly).
> >
> > But they will not load.
> >
> > The i386 script talks about /dev/sda, but there is no such entry in the
> > /dev tree. The nearest seems be /dev/scsi/host0/bus0/target0/lun0/part1.
> >
> > If I switch to VC2 and issue the mount command manually I get a
> > message back saying "mount: Mounting /dev/... on /target failed: Invalid
> > arguement when I issue the command:-
> >
> > mount /dev/scsi/host0/bus0/target0/lun0/part1 /target
> >
> > The same if I include -t ext3.
> >
> > On VC4 there are some errors saying:-
> >
> > EXT3-FS error (device sda1): ext3_check_descriptors: Inode table for
> > group 9 no in group (block 32780)!
> > EXT3-FS group descriptors corrupted!
> >
> > Formatting the disk from VC2 makes no difference, and gives no errors
> > during the formatting. Waiting for the cache to be flushed after
> > formatting for some 10s of minutes also makes no difference.
> >
> > Thanks in advance
> >
> > David
>
> I have take an extra ide disk, build an new kernel boot this and make an
> debstrap from this. Works perfect.
>
> Markus
I tried this with the AMD-64 kernel, and the formatting hung. I will try it
with the I386 kernel and see if it works. At least I now know this is not
just me or the particular hardware I have got. But obviously it needs fixing
sometime, so if anyone has any ideas as to what might be needed to debug
this I only have the machine available for testing for a limited time.
David
Reply to: