[ovirt-users] Storage domain not in pool issue

Adam Litke alitke at redhat.com
Wed Apr 8 14:04:57 UTC 2015


On 08/04/15 09:48 +0000, VONDRA Alain wrote:
>Hi,
>Is there any news ??
>I there anybody who can help me ?
>This long silence get me more and more worried.
>I saw that Federico Simoncelli is working on a new feature "StoragePool Metadata Removal", but I don't have any advice from him or anybody else.
>I'd be really happy to know if somebody works on this kind of problem, I guess that I'm not the only one with this issue, and potentially this can be a disaster bug that could be a bad thing for the reputation of the oVirt platform.
>Don't forget that my Data Center is out of order since March 13th, and this is not trustable for a production environment.
>Let me know what you think about it ?

Hi,

Yesterday I tried to reproduce your situation with a local oVirt setup
I have.  I then tried to manually alter the metadata to force detach
the storage domain from the old pool.  My steps involved the use of
the vgchange command to remove and add VG tags to the volume group
that holds the storage domain.  Unfortunately I was not successful in
reviving the domain.  I'm trying to find some help for you since I
think I have reached my limit for this situation.

How familiar are you with LVM?  If you can activate the VG for your
storage domain and then activate all of its LVs, you may be able to
scrape the disks off of the storage using the qemu-img command.  If
you place those disks on an export domain then you could import them
into your working storage domain and use them to recreate new VMs.
This is a clunky and labor intensive solution but could work if we are
unable to revive the domain in its entirety.

Let me know if you'd like to pursue this approach.

-- 
Adam Litke



More information about the Users mailing list