On 09/09/2011 11:36 AM, magic31 wrote:
> KBR;2132921 Wrote:
>> Hi all,
>> After joining the ad domain and rebooting the xenbased w2k8R2 VM this
>> server has lost its disks and can not boot. We are using sles 10 sp2/3
>> with (Novell/SUSE Virtual machine driver pack).
>> Also there are issues in eventviewer which we solve by editing the
>> config.xen file and replacing "model rtl8138" with "type netfront"
>> The vm's are placed on a cluster ocfs2 volume handled by HA2
>> Strange thing is thatw e have several windows server virtual machine
>> running for years, it looks like this issue is started since windows
>> 2008 R2 (in a domain).
>> Any hints ideas are very welcome!
> Possibly the Windows disk has been marked offline for some reason?
> What exactly are you seeing when trying to boot the affected VM? Is
> Windows throwing an error, or does the VM not get passed the first boot
The problem is caused by a move of the vm and back again, the disk image
file is a sparse file which during the move was "unsparsed". The move
back again resulted in a disk which did not fit anymore on the original
location. So the disk was corrupt.
Now the only issue is that when we uninstall vmdp v1.3 the windows vm is
broke. If we upgrade to vmdp 1.7 first we can do an uninstall. It does
not matter which vmdp version we use we always have atapi, disk and acpi
error in teh event viewer. The acpi errors can be fixed by changing
"model=rtlxxx into type = netfront" but the disk and atapi errors stay.
Novell has stated that vmdp 1.7 is the way to go and the disk and atapi
errors are informational only.