lenny install: some problems with aptitude, and/or kernel?

[Toni Mueller]

Hello,


I am still struggling with my FAI install, using 3.2.20 to install
Lenny/amd64 from a Lenny/amd64 machine, and have encountered these
problems so far (putting the lvm issue aside for a moment):


1. Installing a kernel using a class and "PACKAGES aptitude" fails with
a message about a missing "initrd = yes" statement in
${target}/etc/kernel-img.conf

The problem itself is discussed eg. here:

http://www.mail-archive.com/debian-kernel@lists.debian.org/msg46746.html

This is from software.log:

-------------------- cut
Unpacking linux-image-2.6.26-2-amd64 (from .../linux-image-2.6.26-2-amd64_2.6.26-17lenny2_amd64.deb) ...
Ok, Aborting, since the bootloader might not handle an initrd/initramfs image.
dpkg: error processing /var/cache/apt/archives/linux-image-2.6.26-2-amd64_2.6.26-17lenny2_amd64.deb (--unpack):
 subprocess pre-installation script returned error exit status 1
-------------------- cut


I worked around this problem by creating this custom script to install
the kernel later:

-------------------- cut
#!/bin/sh


if [ ! -f ${target}/etc/kernel-img.conf ]; then
    touch ${target}/etc/kernel-img.conf
fi

ainsl ${target}/etc/kernel-img.conf "do_initrd = Yes"

chroot ${target} aptitude -y install grub linux-image-2.6.26-2-amd64

exit $?
-------------------- cut

This is dead ugly, should imho not be required, and different solutions
are much appreciated!



2. aptitude sometimes (?) fails to install a lot of packages, leaving a
rather broken system behind, and sometimes even deletes (!!!) freshly
installed packages despite the fact that there are no conflicts. This
breaks eg. my scripts to install custom packages by checking code out
of a revision control system, then compile it, breaks big time.

Example:

Writing extended state information...
Reading task descriptions...
The following NEW packages will be installed:
... lots of packages, then:
The following packages will be REMOVED:
  bacula-common{u} console-common{u} console-data{u} initramfs-tools{u} 
  kbd{u} klibc-utils{u} libklibc{u} libvolume-id0{u} mysql-client-5.0{u} 
  udev{u} 



I can't see how this is anywhere near feasible, or how to easily work
around it.




Kind regards,
--Toni++