[kernel-xen] Issue Booting DomU's from certain hardware.

Steven Haigh netwiz at crc.id.au
Thu Sep 27 19:48:18 EST 2012


I've had a couple of reports of people with issues using Debian and 
Fedora for DomU's having issues - however I haven't had any technical 
details to try and track this down.

I have pointed everyone who has contacted me to this list to see if we 
can nut it out.

On the Xen 4.2.0 machine, have a look at the logs in /var/log/xen and 
see if you can parse anything out of that. It almost seems like an 
incompatibility with the guest kernel to the new Xen hypervisor - 
however this shouldn't be the case.

In starting this, it may also be worthwhile to post your DomU config 
file as well so I can even start to think about reproducing things. This 
also goes for others who might have this same issue.

-- 
Steven Haigh

Email: netwiz at crc.id.au
Web: http://www.crc.id.au
Phone: (03) 9001 6090 - 0412 935 897
Fax: (03) 8338 0299

On 27/09/2012 1:21 PM, Joe Fialkowski wrote:
> UPDATE
>
> I was able to finally get the console connected. This is the only thing
> that happens before the DomU crashes:
>
> Using <class 'grub.GrubConf.GrubConfigFile'> to parse /boot/grub/grub.conf
> [root at xm3]#
>
> Here is the /boot/grub/grub.conf from my supermicro server:
>
> # 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 (hd0,2)
> #          kernel /vmlinuz-version ro root=/dev/md1
> #          initrd /initrd-[generic-]version.img
> #boot=/dev/sdb
> default=0
> timeout=5
> splashimage=(hd0,2)/grub/splash.xpm.gz
> hiddenmenu
> title CentOS (3.5.4-2.el6xen.x86_64)
>          root (hd0,2)
>          kernel /xen.gz dom0_mem=1024M cpufreq=xen dom0_max_vcpus=4
> dom0_vcpus_pin
>          module /vmlinuz-3.5.4-2.el6xen.x86_64 ro
> root=UUID=c1e20d4f-2c51-436d-b1d7-b9024482d5ce rd_NO_LUKS
>   KEYBOARDTYPE=pc KEYTABLE=us
> rd_MD_UUID=586bdbaf:6a424617:7f257043:fd4ba025
> rd_MD_UUID=7410ab1f:e7a157c6:c334ce62:18dd2fc9 SYSFONT=latarcyrheb-sun16
> crashkernel=auto rd_NO_LVM rd_NO_DM LANG=en_US.UTF-8 rhgb quiet
>          module /initramfs-3.5.4-2.el6xen.x86_64.img
>
>
> Also I noticed that this new server micro server is running a newer
> version of xen. I was not paying attention and the mirror was updated.
>
> Working server is running:
> xen-libs-4.1.3-2.el6.x86_64
> xen-4.1.3-2.el6.x86_64
> kernel-xen-3.5.4-2.x86_64
> xen-runtime-4.1.3-2.el6.x86_64
> kernel-xen-3.5.4-1.x86_64
> kernel-xen-release-6-3.noarch
> xen-licenses-4.1.3-2.el6.x86_64
> xen-hypervisor-4.1.3-2.el6.x86_64
>
> Non-working:
> xen-runtime-4.2.0-1.el6.x86_64
> kernel-xen-3.5.4-2.x86_64
> xen-licenses-4.2.0-1.el6.x86_64
> xen-hypervisor-4.2.0-1.el6.x86_64
> xen-4.2.0-1.el6.x86_64
> kernel-xen-release-6-3.noarch
> xen-libs-4.2.0-1.el6.x86_64
>
> Hope this helps shed any light on this.
>
> --
> Joe Fialkowski
> ZoomCloud.net
>
> On Wednesday, September 26, 2012 at 10:29 PM, Joe Fialkowski wrote:
>
>> Hello kernel-xen readers,
>>
>> I am new to this list, but I appreciate everything you folks have done
>> for all things Xen and hope to help out in anyway I can.
>>
>> I have been using Xen for a few months now and am running the 3.5.4-2
>> xen kernel on CentOS 6.3. Everything has been working great but I have
>> hit an issue that I am hoping the support from this list can help me
>> overcome.
>>
>> Here is the issue:
>>
>> I have 3 Servers. Two are HP DL120 G7's with Xeon E3-1230V2
>> Processors, and hardware raid. The Other server is a super micro with
>> the same CPU spec's but software raid.
>>
>> I have Xen PV Images for CentOS 6, Debian 6, Ubuntu 12.04 and Fedora
>> 17. These images work great on DomU's on the HP hardware. However on
>> the super micro server, my Ubuntu and Fedora Images will not boot.
>> They crash almost immediately and I cannot even get a console
>> connected in time (even using xm start -c).
>>
>> I am wondering if it has something to do with the hosts partitioning.
>> On the HP hardware it sees /dev/sda for the drive, however on the
>> Supermicro since its software raid, its /dev/md0 which is a raid 1
>> mirror of sub and sdc.
>>
>> I am not sure if that makes a difference or not, I am really shooting
>> in the dark. It may also be worth mentioning I am using SolusVM to
>> provision VPS's across the cluster of Servers.
>>
>> Any insight you can share is greatly appreciated.
>>
>> Kind Regards,
>>
>> --
>> Joe Fialkowski
>> Joe at ZoomCloud.net <mailto:Joe at ZoomCloud.net>
>>
>> _______________________________________________
>> kernel-xen mailing list
>> kernel-xen at lists.wireless.org.au <mailto:kernel-xen at lists.wireless.org.au>
>> https://lists.wireless.org.au/mailman/listinfo/kernel-xen
>
>
>
> _______________________________________________
> kernel-xen mailing list
> kernel-xen at lists.wireless.org.au
> https://lists.wireless.org.au/mailman/listinfo/kernel-xen
>


More information about the kernel-xen mailing list