7:26 p.m.
----- Original Message -----
From: "Carlos A. Diaz" <carlos_d99(a)yahoo.com>
To: "Nir Soffer" <nsoffer(a)redhat.com>
Cc: users(a)ovirt.org, "Liron Aravot" <laravot(a)redhat.com>
Sent: Monday, September 8, 2014 7:58:33 PM
Subject: Re: [ovirt-users] Issues with Storage domain
Thanks Nir,
Attached is a Snippet of the log, that essentially keeps going and going with
a few alerts of the engine trying to migrate some VMs unsuccessfully which
is pretty much related to the fact that the storage is not available; so I
filtered those out.
I cannot tell anything by these snippets. I need the full logs, since the last
vdsm was started and connected successfully to this storage domain.
Can you mount manually your NFS storage domain?
You can find the mount command that vdsm uses in vdsm log. You should be
able to copy and paste it to the shell in recent versions. In older versions,
you may need to tweek the quoting to make it work.
Does storageDomainId = 00000000-0000-0000-0000-000000000000 sounds right?
Anyways, I am looking at older logs to see if the domain ID changed, I will
keep you posted, if you need anything else let me know, and thanks again :)
Carlos
On Sunday, September 7, 2014 7:26 PM, Nir Soffer <nsoffer(a)redhat.com> wrote:
----- Original Message -----
> From: "Carlos A. Diaz" <carlos_d99(a)yahoo.com>
> To: users(a)ovirt.org
> Sent: Friday, September 5, 2014 9:36:42 PM
> Subject: [ovirt-users] Issues with Storage domain
>
> Hello everyone, its my first time here; so I hope I am on the right place
>
> I had a problem where one of our SysAdmins cloned our NFS production export
> in order to create a new NFS export, but he didn't changed the FSID, so we
> ended up with 2 nfs exports using the same ID. I was able to roll back and
> recover the data, once I had corrected everything I went to the hosts and
> unpaused the VMs from the CLI (I lost the engine in the process), so now we
> are back up and running.
>
> My problem right now is that oVirt says that our main storage domain is
> "Down" and I can't bring it up, so essentially I cannot start any VMs
on
> that domain, which hosts 90% of our production environment; which is also
> scary because I am not sure how to recover from this error without
> affecting
> my production environment.
>
> My guess is that at some point the MD5 hash changed, or something in the
> metadata for that Domain, but I don't know if this is the problem and if
> recreating the signature is possible without breaking my existing
> configuration.
>
> Is there anyway to mark it as UP from the database or at any other level
> without breaking (more) the config?
Hi Carlos,
Please attach engine.log and vdsm.log, hopefully we can understand from
the logs why you cannot activate the storage domain.
Thanks,
Nir