Dear freinds:

I have a hard problem with GRUB and I donīt know how I can resolving it.

I'm doing a master dvd to mastering a lot of servers-computers (about 2000) ditributed.
The architecture is like this one:

Fujitsu-Siemens Primergy S3/S4/S5. They have a OS RedHat Enterprise 5.3. three partitions on the HD (hard-RAID-1). The partition 1 it's for the /boot, the partition 2 is for LVM (/root, /swap and others, all in LVM) and the partition 3 is a security partition to allow me to have a maintenance system on line.

For this job I use the SystemRescueCD 1.1.5 release with the parted, partimage, and others tools. The third partition has a SystemRescueCd installed as it's described in link:

Sysresccd-manual-en Easy install SystemRescueCd on harddisk - SystemRescueCd

To install the grub I have used the easy way:

install-grub /dev/sda

And all was ok, but my problem come now.

My company has a lot of IBM servers xSeries-205 (more than 1000 although they will be exchange on few time). I have to expand the dvd-master in order to can update they at Red Hat Enterprise-5.3.

The architecture is:

IBM xSeries 205
Intel Pentium 4 2.66 Ghz.
1 Gb RAM
40 Gb Hard Disk in two disk on mirror with this car: RAID bus controller: Silicom Image, Inc, PCI0680 Ultra ATA-133 Host Controller (rev 02)

Yeah, I know those servers are a lot of ancients but while the company change they, I have to install the master.

To make the master I'm using the SystemRescueCd-1.1.5., like I told before and I bootload de RescueCd with this parameters:

altker32 setkmap=es dodmraid dolvm2 rootpass=XXXXXXXX eth0=172.20.128.10/17 gateway=172.20.128.2.

Thanks to dmraid option I can see the device:

/dev/dm-0 -> /dev/mapper/lsi_bhgeiicjeea

I also can part it, making the logical volumes and use the partimage to write the file system. The real problem is when I install the grub. To do it I write on my script these sentences:

mount /dev/mapper/lsi_bhgeiicjeea /boot

grub --batch --device-map=/boot/grub/device.map --config-file=/boot/grub/menu.ls
t << EOT
root (hd0,0)
setup (hd0)
quit
EOT

The configuration of config file is:

# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,0)
# kernel /vmlinuz-version ro root=/dev/CEPSA/ROOT
# initrd /initrd-version.img
#boot=/dev/mapper/lsi_bhgeiicjeea
default=0
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux Server (2.6.18-128.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-128.el5 ro root=/dev/CEPSA/ROOT rhgb quiet
initrd /initrd-2.6.18-128.el5.img
title Red Hat Enterprise Linux Server monousuario (2.6.18-128.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-128.el5 ro root=/dev/CEPSA/ROOT rhgb quiet single
initrd /initrd-2.6.18-128.el5.img
title SystemRescueCd desde disco duro
root (hd0,2)
kernel /sysrcd/altker32 subdir=sysrcd setkmap=es
initrd /sysrcd/initram.igz
boot

And it's installed without any problem. Finally I dispose to reboot the server.

Once the server is up and I enter at grub menu, the server can boot on RedHat Server environment and Red Hat server monouser mode but when I want to boot on the SystemRescue mode, the grub say me:

Booting comman-list

root (hd0,2)
Filesystem type is ext2fs, partition type 0x83
kernel /sysrcd/altker32 subdir=sysrcd sermap=es

Error 2: Bad file or directory type

Press ana key to continue...

At this point everybody can see the Error 2 of GRUB is unjustified.

AT the Fujitsu-Siemens all servers have the same configuration and I can booting the System-Rescue partition with no problem.

First I posted the problem at Ssytem-RescueCd Forums and they answer me:

"It you don't see messages from the kernel, it means it's in an early stage of the boot process, so it's a grub problem."

I have also attempted to install the grub through the order:

mount /dev/dm-0 /boot
grub-install /dev/dm-1

But Grub respond whith these sentence:

/dev/mapper/lsi_bhgeiicjeea does not have any corresponding BIOS drive.

I have been studying the problem with the GNU GRUB manual and it says at the section 3.3:
Installing GRUB using grub-install pag. 8

This procedure is definitely less safe, because there are several ways in which your computer can become unbootable. For example, most operating systems donīt tell GRUB how to map BIOS drives to OS devices correctly (GRUB merely guesses the mapping. This will succeed in most of cases, but not always).

Please notice that on Fujitsu Server I can reach and boot the System-Rescue on hard disk without any problem. It bootload well and I achieve a perfectly maintenance environment. So I know it isn't a problem of the SystemRescue installation disk. Even more, if I installing Grub on an Fujitsu-Server through the script before then I reach the same problem with Grub an error 2.

And it's all folks.

If some body can help me I will be very grateful.

luispoo