From jahudson at novell.com  Tue Jun  9 09:33:44 2015
From: jahudson at novell.com (Jared Hudson)
Date: Tue, 09 Jun 2015 09:33:44 -0600
Subject: [sles-beta] SLES11SP4RC2 x86_64 autoyast installation - problem
 with reinstall
In-Reply-To: <1E287A5DDB506E478DE9CAE3C8492D01F94F22A5@BRDC-MBX-1.network.internal>
References: <1E287A5DDB506E478DE9CAE3C8492D01F94F22A5@BRDC-MBX-1.network.internal>
Message-ID: <5576C108020000AD0001EF1F@prv-mh.provo.novell.com>

I believe what you are experiencing is that UEFI sees the GPT partition is damaged since your dd command only removes the primary GPT header and not the secondary GPT header. It then restores the primary GPT header using the secondary GPT header information.

Your problem is most likely due to this command 'dd if=/dev/zero of=/dev/sda bs=1M count=1' being insufficient. With an MBR partitioning scheme that command would be sufficient however with a GPT partition there are GPT records stored at the beginning and end of the disk. See http://en.wikipedia.org/wiki/GUID_Partition_Table#/media/File:GUID_Partition_Table_Scheme.svg

You will need to zero out at least the first and last 17408 bytes of the disk. You can probably use /sys/block/sda/size to see the size then subtract 34 512-byte sectors and use dd to seek to that location and zero another 34 512-byte sectors. If you had access to the sgdisk command, which is included in SLES 12, you could more easily just run sgdisk --zap-all /dev/sda to remove both a MBR and GPT partition table.

There's also the Autoyast XML <initialize> tag which when set to true (by default it's false) should wipe the partition table as well:
<initialize config:type="boolean">true</initialize>
See http://users.suse.com/~ug/autoyast_doc/configuration.html#CreateProfile.Partitioning

Jared H.


Jared H. Hudson
Dedicated Support Engineer
jahudson at novell.com 
479-657-4309 Office
479-283-4677 Cellular


>>> On 6/3/2015 at 03:06 PM, in message
<1E287A5DDB506E478DE9CAE3C8492D01F94F22A5 at BRDC-MBX-1.network.internal>, Joel
Barbieri <Joel.Barbieri at merge.com> wrote: 

> Finally got a chance to work with SLES11SP4.  Cut our shrunken distro. 
>  
> Mostly I see problems with re-installations.  The autoyast profile is set to  
> "blow-up" the server each time, but stops with some sort of problem dealing  
> with the partitions on a "second" install. 
>  
> Hardware is Intel NUC Core i3 in UEFI mode with installation over the wire.   
> Problem is consistent.  Will try on a Dell r610 in UEFI tomorrow to see if  
> the problem follows the hardware or software.  I have had similar grief in  
> SLES11SP3 when switching a system from UEFI to BIOS mode where I have to nuke  
> the drive manually to execute a re-install from scratch, but never when  
> re-installing without switching BIOS modes. 
>  
> To nuke a drive, I just escape to an available shell [alt-f2, or whichever is  
> available], dd if=/dev/zero of=/dev/sda bs=1M count=1;reboot -f.  Don't forget  
> the "-f" on the reboot.  This is the only way I have succeeded on Dell r610 at  
> switching from UEFI back to BIOS mode.  Without it, I estimate the kernel or  
> RAID card or some other driver re-writes the UEFI/GPT partition table back to  
> disk, thus not effectively "cleaning" it. 
>  
> I wouldn't call this a show stopper, but it is a major nuisance for product  
> testing for me.  It wouldn't be so bad except for Server Manufacturers have  
> made the BIOS/UEFI initialization so long that it physically hurts. UEFI was  
> supposed to be faster, but apparently not on servers.  Does great on laptops  
> and many modern desktops. 
>  
> I gathered the /var/log/YaST directory in case SuSE would like to look at  
> why it bombed.  I may deconstruct also if I have time. 
>  
> Let me know if this is SR worthy by description. 
>  
> I don't remember of SLES12 was more amenable to always succeeding in second  
> installs.  I may have just kept them all in UEFI.  I know on the r610s it  
> re-installed repeatedly just fine. 
>  
> -Joel 
>  
> Joel Barbieri 
> Merge Healthcare 
>  
> ________________________________________ 
> From: sles-beta-bounces at lists.suse.com [sles-beta-bounces at lists.suse.com] on  
> behalf of urs.frey at post.ch [urs.frey at post.ch] 
> Sent: Tuesday, May 26, 2015 7:17 AM 
> To: sles-beta at lists.suse.com 
> Subject: [sles-beta] SLES11SP4RC2 x86_64 autoyast installation & online  
> upgrade OK 
>  
> Hi 
>  
> Just installed a HP Proliant BL460cGen8 and a HP Proliant BL460cG7 with  
> SLES11-SP4 x86_64 using autoyast 
> Installation works OK. Profiles working under SLES11-SP3 are working the same  
> under SLES11-SP4 
>  
> Upgrades using zypper dist upgrade from SLES11-SP3 kernel version  
> 3.0.101-0.47.52-default do work OK. 
>  
> Thank you very much SUSE for this stable work 
> Best regards 
>  
> Urs Frey 
> Post CH AG 
> Informationstechnologie 
> IT Betrieb 
> Webergutstrasse 12 
> 3030 Bern (Zollikofen) 
> Telefon : ++41 (0)58 338 58 70 
> FAX     : ++41 (0)58 667 30 07 
> E-Mail:   urs.frey at post.ch<mailto:urs.frey at post.ch> 
>  
>  
>  
> _______________________________________________ 
> sles-beta mailing list 
> sles-beta at lists.suse.com 
> http://lists.suse.com/mailman/listinfo/sles-beta 
>  
>