On Fri, Mar 20, 2015 at 12:34:59PM +0100, Francesc Guasch wrote:
> On Fri, Mar 20, 2015 at 10:03:20AM +0000, Stefan Hajnoczi wrote:
> 
> Hi Stefan, thank you very much for answering me.
> 
> > On Wed, Mar 18, 2015 at 04:53:28PM +0100, Francesc Guasch wrote:
> > > I have three Ubuntu Server 14.04 trusty with KVM. Two of
> > > them are HP servers and one is Dell. Both brands run fine
> > > the KVM virtual servers, and I can do live migration between
> > > the HPs. But I get I/O errors in the vda when I migrate to
> > > or from the Dell server.
> > > 
> > > I have shared storage with NFS, mounted the same way in all
> > > of them:
> > > 
> > > As soon as it starts in the origin console I spot I/O error
> > > messages, when it finishes I got them in the console in the
> > > destination server. The file system is read only and I have to
> > > shut it down hard.
> > > 
> > >     end request I/O error, /dev/vda, sector 8790327
> > 
> > "origin console" == guest's console?
> 
> Yes, I mean I open two consoles with virt-manager, one in
> the origin host and another one in the destination
> > 
> > I/O errors starting while the guest is still running on the migration
> > source host is strange.  I wonder if something happened to the NFS file
> > related to file permissions or SELinux labels?
> 
> I think I found something checking SELinux. ls -Z and getfattr
> return nothing. But ps -eZ showed something very different
> in the Dell server.
> 
> This is in the HP server:
>     /usr/sbin/libvirtd      1034 ?        11:51:44 libvirtd
>     libvirt-09540b5d-82     701  ?        05:28:40 qemu-system-x86
>     unconfined              1    ?        00:01:00 init
> 
> In the Dell server init is confined in lxc and there are also
> lxc-start processes.
> 
>     /usr/sbin/libvirtd      1622 ?        05:07:07 libvirtd
>     libvirt-8a0f9087-32d... 29926 ?       00:00:01 qemu-system-x86
>     lxc-container-default   1774 ?        00:00:00 init
>     /usr/bin/lxc-start      1763 ?        00:00:00 lxc-start
>     
> There is also LXC installed in that server ! Maybe that is messing
> with kvm. The qemu processes look fine to me but there is a chance
> the problem comes from there.
> 
> I could move the LXC somewhere else or I can keep it there to
> try to fix this issue. What do you advice I should do now ?

I suggest asking on the libvirt mailing list: libvirt-l...@redhat.com

Attachment: pgpk8pbvBvTGx.pgp
Description: PGP signature

Reply via email to