Hi,
We'll have SUSE Linux 10.1 beta1 soon - the last SUSE Linux before
SLES10. With this eMail I want to encourage you to take part of
the beta test of SUSE Linux. Especially for you, as administrator of
an autoinstallation environment, it's important to find
problems and bugs during the beta phase. As soon as the 10.1 is gold,
we can't do any fixes anymore for autoyast because that requires a
patch for the installation system which is on the CD/DVD.
So official online updates for autoyast are hardly an option then.
To make changes in autoyast a little bit more visible for you I've
started a brief overview of what has changed in autoyast since 10.0
The current state (post alpha4) of what has changed can be found here:
http://www.suse.de/~ug/autoyast_changes.html
If you want to download the latest test version of SUSE Linux, you can
download it here:
http://www.opensuse.org/Development_Version
If you think you've hit a bug you can either write a mail to this list
or (even better) you can use bugzilla to report a bug:
https://bugzilla.novell.com
Thanks in advance for testing SUSE Linux.
--
ciao, Uwe Gansert
Uwe Gansert, Server Technologies Team
SUSE LINUX Products GmbH, Maxfeldstrasse 5, D-90409 Nuernberg, Germany
e-mail: uwe.gansert(a)suse.de, Tel: +49-(0)911-74053-0,
Fax: +49-(0)911-74053-476, Web: http://www.suse.de
So can I assume that technically it is ok if the system does not reboot
after disk1? I was just concerned about the change in behavior.
My systems seem ok, they are functioning properly all RPMS appear to be
installed, and I assume that the installer is getting all dependencies
taken care of.
I have to do http installs because of firewall restrictions. So I am ok
with the reboot.
I have this in my mode section...
<mode>
<installation config:type="boolean">true</installation>
<confirm config:type="boolean">false</confirm>
<forceboot config:type="boolean">true</forceboot>
<interactive_boot config:type="boolean">false</interactive_boot>
<interactive config:type="boolean">false</interactive>
<reboot config:type="boolean">true</reboot>
<upgrade config:type="boolean">false</upgrade>
</mode>
Thanks for getting back to me.
Todd
-----Original Message-----
From: Uwe Gansert [mailto:ug@suse.de]
Sent: Thursday, January 05, 2006 6:53 AM
To: suse-autoinstall(a)suse.com
Subject: Re: [suse-autoinstall] Custom media, no first reboot
On Thursday 05 January 2006 14:33, Ness, Todd wrote:
> I have JDS which in based off SLES8.
>
> Any idea of how to get the initial reboot to happen after Disk1 of JDS
> and Disk1 of my patches are installed?
I hope got you right but even then, I'm not sure if I can help.
First you can try to add:
<install>
<general>
<mode>
<forceboot config:type="boolean">true</forceboot>**
</mode>
</general>
</install>
And if this does not help, you can try to install from an NFS source (I
guess you are using ftp or http or something like this at the moment?)
With NFS the RPMs are not getting copied to the local machine and so you
don't need that much memory.
--
ciao, Uwe Gansert
Uwe Gansert, Server Technologies Team
SUSE LINUX Products GmbH, Maxfeldstrasse 5, D-90409 Nuernberg, Germany
e-mail: uwe.gansert(a)suse.de, Tel: +49-(0)911-74053-0,
Fax: +49-(0)911-74053-476, Web: http://www.suse.de
--
To unsubscribe, e-mail: suse-autoinstall-unsubscribe(a)suse.com
For additional commands, e-mail: suse-autoinstall-help(a)suse.com
I have asked this before and either missed the reply or did not get an
answer.
I have JDS which in based off SLES8.
I have a bunch of new patches, plus some other RPMs that I want to add
to the install media, and I am have made my own yast directory with
installorder and order file, everything works mostly as expected. But,
one thing is different. During the very first boot of the system
installation starts, and all of the RPMS get installed on this one boot,
there is not a boot after the RPMS from Disk1 have been installed.
I don't think this is causing me any issues, the StarOffice RPM is the
only RPM that installs (during network installs only) after the initial
boot.
Any idea of how to get the initial reboot to happen after Disk1 of JDS
and Disk1 of my patches are installed?
Todd
Hi all,
I just ran into a bizzare problem: I'm currently creating/testing our
auto-install, hence, I'm re-running it many, many times on two PCs.
So far, I always used a floppy with "info" and "autoinst.xml" and a copy
of CD 1 to boot. I would then select "Install", the installer would find
"info" on the floppy and the rest would happen magically from there. In
fact, I have reinstalled the second of the two PCs that way just
yesterday.
Today, however, I'm trying to reinstall the first PC - and am getting
nowhere. It starts the install, I can see the message "searching for
info file" (or suchlike) on the screen and the floppy drive spins up.
Then it proceeds into the graphical installer (my auto-install uses
text-only) and prepares for auto-installation (so it *must* have found
the info file as such) - but hangs, asking for a profile for the
machine. If I give "floppy:///autoinst.xml" as answer to that, it will
happily continue, though it tries to install from CD-ROM (instead of via
network, as defined in the info file). If I choose "Network Install"
manually right at the start, I can get the autoinstall working.
It looks as if it *does* find 'info' on the floppy (hence starts in
auto-install mode), but for some odd reasons refuses to read the actual
contents. I'm completely puzzled as to why this would happen, as I have
installed that very machine several times that way.
Hence, if there is anyone out there with a brainwave as to where to look
to fix (or even just investigate) this, I'd be most grateful...
Cheerio,
Thomas
Hi,
here's sth. which didn't work in some earlier versions (don't remember when
I last tried it) but is working at least in 10.1 alpha4. I can now use
this rule in rules.xml to chose a profile for each host based on its
name.
This works only if the host gets its name via dhcp, so this is sth.
especially nice for pxe-based installations. I provide this as example
for other people and hope that someone can make use of it :-) This
is a solution for the non-working <hostname> matching, because the
<hostname> tag contains the IP and not the hostname, even when
using pxe/dhcp.
<rule>
<custom1>
<script>
<![CDATA[
#!/bin/sh
. /var/lib/dhcpcd/dhcpcd-eth0.info
echo -n $HOSTNAME
]]>
</script>
<match>*</match>
<match_type>exact</match_type>
</custom1>
<result>
<profile>profiles/part_@custom1@.xml</profile>
<continue config:type="boolean">true</continue>
</result>
</rule>
I use this to store partition information per host in a file
part_<hostname>.xml in the profile subdir. No need to change
the rules.xml anymore, just create the profile with the partition
information. In these profiles, I always keep partitions, formatting
only / and /var. In case I really really want to destroy and recreate
partitions for a specific host, I add another rule afterwards, e.g.
<rule>
<custom1>
<script>
<![CDATA[
#!/bin/sh
. /var/lib/dhcpcd/dhcpcd-eth0.info
echo -n $HOSTNAME
]]>
</script>
<match>riemann</match>
<match_type>exact</match_type>
</custom1>
<result>
<profile>profiles/part_riemann_createnew.xml</profile>
<continue config:type="boolean">true</continue>
</result>
</rule>
or, IP-based:
<rule>
<hostaddress>
<match>141.x.x.x</match>
<match_type>exact</match_type>
</hostaddress>
<result>
<profile>profiles/part_riemann_keep.xml</profile>
<continue config:type="boolean">true</continue>
</result>
</rule>
to override the common selection.
Have fun!
cu,
Frank
--
Dipl.-Inform. Frank Steiner Web: http://www.bio.ifi.lmu.de/~steiner/
Lehrstuhl f. Bioinformatik Mail: http://www.bio.ifi.lmu.de/~steiner/m/
LMU, Amalienstr. 17 Phone: +49 89 2180-4049
80333 Muenchen, Germany Fax: +49 89 2180-99-4049
* Rekursion kann man erst verstehen, wenn man Rekursion verstanden hat. *
Hi,
checking out opensuse 10.1a4, I see that the boot/ directory now
has two subdirectories with different initrds and kernels for i386
and x86_64.
I wonder what that means for autoyast installation via pxe. Is the intention
to setup two different pxe configurations with two different kernels/initrds
for 32 and 64bit hosts? In contrast to just one initrd on 10.0 where the
installer would chose 32 or 64bit installation from a common source
automatically?
cu,
Frank
--
Dipl.-Inform. Frank Steiner Web: http://www.bio.ifi.lmu.de/~steiner/
Lehrstuhl f. Bioinformatik Mail: http://www.bio.ifi.lmu.de/~steiner/m/
LMU, Amalienstr. 17 Phone: +49 89 2180-4049
80333 Muenchen, Germany Fax: +49 89 2180-99-4049
* Rekursion kann man erst verstehen, wenn man Rekursion verstanden hat. *
Hi,
how can I change the default filesystem type "reiserfs" to for example "xfs"
with autoyast?
The effect should be among other things for example that the yast-GUI
auto-proposes this filesystem when you create a new logival volume with LVM.
Thanks in advance.
--
Lust, ein paar Euro nebenbei zu verdienen? Ohne Kosten, ohne Risiko!
Satte Provisionen f�r GMX Partner: http://www.gmx.net/de/go/partner
Can you create an autoyast.xml file from an existing system (with minimal
effort) that you want other servers to mimic?
Does that make sense?
Thanks,
LDB
It is possible install the Yast module "cd creator" of the Suse
enterprise server in the Suse professional.
If it will go as I do?
Att.
Rodrigo Albert Fedosi
Suporte TI
Tel: (31) 3419-8854
Fax: (31) 3419-8803
rodrigo(a)bhz.jamef.com.br
www.jamef.com.br
Hello,
I'm having 3 questions :
1. I've tried to install SLES9 SP2 on an IBM x306 however i was having problems with the installation cuz it did not recognize the raid controller and it showed both physical drives instead of the logical drive. I had to manually install this server using a driver update disk. Needless to say, but i also want to automatically deploy SLES9 SP2 on this server. I've checkout this document : http://mirror.mcs.anl.gov/suse-people/hvogel/Update-Media-HOWTO/Update-Medi… It states that i will need to put a file 'driverupdate' in the root of my repository. (I assume that's the driver update disk image i can download from IBM's support site). How can i use several update disks if they all should be named 'driverupdate'?
2. Say I have an http repository which contains the installation sources for SLES9 SP2. One day Novell releases a critical security patch and I'd like to integrate this patch in my repository so i don't need to install these patches after installing a server. Is there a way to do this (besides doing this with post-install scripts that just run rpm-commands)? Replacing the package in the repository won't work i think, cuz (if i'm not wrong) there is a list of packages on the install sources.
3. Currently i'm using a file in pxelinux.cfg which contains references to the kernel- and initrdimage and also the linuxrc command line parameters. AutoYasts documentation mentiones an info file that contains these parameters. Where do i place such an info file? Can i combine the info file and the command line parameters? eg i could put all parameters that are not unique for a given server in the info file and then put the server specific option in the file in pxelinux.cfg.
Thanks a lot!
Johan