I have modified my autoyast to a simpler partitioning scheme. This seems to work. However, eventually I will need to add support for a more advanced partitioning scheme with multiple drives. Do you suggest using the Autoinstallation tool, or editing the bare xml file? I have attached the modified autoinst.xml file. Below is the partitioning scheme I would like to implement. sda1 - /boot 100M - /swap 2GB - / 30GB - LVM max sda2 - LVM max volgrp1 - logical volume 1 max Sorry if I responded to the wrong address --Wayman On Fri, 2008-01-11 at 09:41 +0100, Uwe Gansert wrote:

On Donnerstag, 10. Januar 2008, Wayman Smith wrote:

...

/sbin/mkinitrd: line 3297: /etc/fstab: No such file or directory No '/' mountpoint spceified in /etc/fstab

your partitioning section looks not so good. /dev/sdb looks strange and can (should) be removed out of your profile. on sda you don't want to create any partition except for one in the middle of all others. That's a problem with autoyast. Set create to "false" for the root partition too or set all other partitions to create=true with a <use>all</use> maybe

...

It should also be noted that my autoinst.xml file fails RNG validation (https://bugzilla.novell.com/show_bug.cgi?id=211014#c21)

you can ignore that.

-- ciao, Uwe Gansert

Uwe Gansert, Server Technologies Team SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) Business: http://www.suse.de/~ug

On Fri, Jan 11, 2008 at 03:31:18PM -0500, Wayman Smith wrote:

I have modified my autoyast to a simpler partitioning scheme. This seems to work. However, eventually I will need to add support for a more advanced partitioning scheme with multiple drives. Do you suggest using the Autoinstallation tool, or editing the bare xml file? I have attached the modified autoinst.xml file. Below is the partitioning scheme I would like to implement.

You can already do more advanced partitioning for multiple drives (and differing numbers/types of drives) by creating the partitioning info in separate files and using the existing rules framework to read in the matching situation automatically.

sda1 - /boot 100M - /swap 2GB - / 30GB - LVM max sda2 - LVM max volgrp1 - logical volume 1 max

I'm doing some LVM stuff (on one disk) already. I'm using only some of the disk for the OS and leaving the rest available for xen instances. I attached the file for doing this if you're curious about how to get LVM going in autoyast. If you have a situation with sda and sdb ( assume you actually meant that above vs sda1 and sda2) you could use something based on my attached file with 2 sections that create the initial LVM devices, one with: <device>/dev/sda</device> and named something like: diskone and then second section exactly the same except for the /boot partition and the following differences: <device>/dev/sdb</device> and named something like: disktwo Then you can make your partitions on: <device>/dev/diskone</device> and <device>/dev/disktwo</device> You'll note from my example too, that you don't need to keep the OS partitions outside of the LVM group.. only /boot needs to be outside it so that grub (or lilo or any bootloader you like) can read the kernel and initrd. Even swap works find inside lvm. -- Mike Marion-Unix SysAdmin/Staff IT Engineer-http://www.qualcomm.com [Lisa and Bart used Gandpa's name on a cartoon script they wrote] Bart - "Didn't you wonder why you were getting checks for doing absolutely nothing?" Grandpa - "I figured 'cause the Democrats are in power again."

On Mittwoch, 16. Januar 2008, Wayman Smith wrote:

Thank you for the quick reply. I am in need for a little more clarification. So, I only need to create the "middle" partition?

creating a partition in the middle is dangerous with autoyast.

I read over the Partition section of the autoyast documentation (http://www.suse.com/~ug/autoyast_doc/CreateProfile.Partitioning.html) and generated another layout for my 2 drives (sda & sdb). I have attached it. It passes validation in kxmleditor. Does this look like it would work?

why do you have sdb in your profile? You don't want to create anything on it, you don't want to format anything and there is nothing to mount. You can remove it from the profile. for sda, set the root partition to create=false All four partitions must already exist on sda then, when autoyast starts. If that is not the case, the create=true to all of them. -- ciao, Uwe Gansert Uwe Gansert, Server Technologies Team SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) Business: http://www.suse.de/~ug -- To unsubscribe, e-mail: opensuse-autoinstall+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-autoinstall+help@opensuse.org

...

...

"Stephens, Bill PBSG" 24/01/2008 17:06 >>> Is there a way to set up 2 bond networks using 4 nic's through autoyast, without knowing the macaddresses ahead of time? Our config is 2 broadcom onboard nics, and a dual port Intel PCI nic. The goal is to bond the first ports (eth0 and eth2) to our intranet lan, and the second

AFAIK has each NIC vendor its own range of mac address. Thus if you know the vendor you can figured out the starting of the mac address and create a rule on that. I do not know if autoyast is able to create bonded network channels at all.. ports (eth1 and 3) to a private subnet. -- To unsubscribe, e-mail: opensuse-autoinstall+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-autoinstall+help@opensuse.org -- To unsubscribe, e-mail: opensuse-autoinstall+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-autoinstall+help@opensuse.org

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Michal Zugec wrote:

Stephens, Bill {PBSG} wrote / napísal(a):

...

Is there a way to set up 2 bond networks using 4 nic's through autoyast, without knowing the macaddresses ahead of time? Our config is 2 broadcom onboard nics, and a dual port Intel PCI nic. The goal is to bond the first ports (eth0 and eth2) to our intranet lan, and the second ports (eth1 and 3) to a private subnet.

This is possible from openSUSE-11.0 and later. Until this (the question is probably about SLES10-SPx) there is no way to have persisten names for network devices. You can use PERSISTENT_NAME option, but there are some problematic issues. You can use configuration names ifcfg-eth0 instead of ifcfg-eth-id-12:23:34:45:56:67 but order of ethX will be set up randomly.

You can use the PCI-IDs to rename the Interfaces: #chroot-script cat <<EOF > /mnt/etc/udev/rules.d/30-net_persistent_names.rules SUBSYSTEM=="net", ACTION=="add", ID=="0000:06:00.0", IMPORT="/lib/udev/rename_netiface %k eth0" SUBSYSTEM=="net", ACTION=="add", ID=="0000:06:00.1", IMPORT="/lib/udev/rename_netiface %k eth1" SUBSYSTEM=="net", ACTION=="add", ID=="0000:04:01.0", IMPORT="/lib/udev/rename_netiface %k eth2" SUBSYSTEM=="net", ACTION=="add", ID=="0000:04:01.1", IMPORT="/lib/udev/rename_netiface %k eth3" EOF Works with SLES10 SP1. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iD8DBQFH1W8qK86mLtuNvScRAl59AJ4xKioQebRAYqH4Iykx44wcO2KpuACfYvkE Fb8dh6YrTV0DrWA9F86rczo= =6c35 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-autoinstall+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-autoinstall+help@opensuse.org