I am trying to use Rocks's anaconda to install on a Xen virtual
machine. This fails when anaconda does not detect the existing
disk partitions. I would not expect it to detect the xen virtual block
device since there is no PCI or IDE bus. The hda1 and hda2 partitions
exist and have the same major/minor numbers as an IDE disk.
What I want to know is how to manipulate the anaconda/kickstart config so
anaconda will mknod /dev/hda1 and /dev/hda2 and start writing to them
even though it never detected my hda disk.
The disk is partitioned and formated already. I do have this much in
the kickstart config:
part / --noformat --onpart hda1
part swap --noformat --onpart hda2
..but since Anaconda thinks there are no disks, installation doesn't get
past the "partitionobjinit" step in installclass.py.
partitionObjectsInitialize fails at diskset.checkNoDisks(intf)
because there are no disks in probedList.
Can some %pre code just inform anaconda it has /dev/hda1 and
/dev/hda2 even though it didn't detect anything?
Or can a driver disk be hacked to do this?
David Becker
becker@xxxxxxxxxxx
