--Apple-Mail=_80349546-76AA-4918-9B45-6194CB54F8F2 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=windows-1252 On 4 Oct 2013, at 19:55, emitor@gmail.com wrote:
Dan, I've made a whole new installation of oVirt 3.3 using the stable = repos, I've updated libvirt to 1.1.2-1 and configured a glusterFS = cluster. =20 I've just tried the live migration of the Windows Host, here i send = you the vdsm, libvir and qemu logs
=20 =20 2013/10/4 Dan Kenigsberg <danken@redhat.com> On Wed, Oct 02, 2013 at 01:36:56PM -0300, emitor@gmail.com wrote:
I've configured the host as a NFS cluster. I've imported the VMs, =
debian now it's migrating without problems, the pfSense now gives me = an error (Migration failed due to Error: Desktop does not exist) and =
Windows VM doesn't start in the new cluster after import it.
I've attached the logs of the vdsm =20 Thanks for providing your, but they begin too late. =20 The bug is still a mystery to me. Something garbles Vdsm's internal representation of virtual disks. And it's bad. =20 I would appreciate if you could dig backward in your logs and find the last successfull startup of a VM, say W2K8_cluster1 (86f10137-d917-4508-a07a-3f83e6f24a95). Then, browse forward to find =
hmm=85I see both PCI address for CDROM as we've been seeing on gluster = setups. So far no update from gluster AFAIK=85it seems to be broken on = their side somewhere=85 and I also see error on migrations dst which seems to be = another/different disk device mess up The W2K8 create VM call in the ovirt1's log is already wrong, can you = capture/include the initial create call? In the previous cases it was = always the case that the VM was created(from engine) with IDE address = but then after the first run it somehow changed to PCI on the first = status update from vdsm Thanks, michal the the the
first failure (be it on next startup of VM or during migration). =20 Best regards, Dan. =20 =20 =20 --=20 Emiliano Tortorella +598 98941176 emitor@gmail.com = <ovirt_logs_clusterGFS.tar.gz>____________________________________________=
Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
<blockquote type=3D"cite"><div dir=3D"ltr"> </div><div class=3D"gmail_extra"><br><br><div = class=3D"gmail_quote">2013/10/4 Dan Kenigsberg <span dir=3D"ltr"><<a =
--Apple-Mail=_80349546-76AA-4918-9B45-6194CB54F8F2 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=windows-1252 <html><head></head><body style=3D"word-wrap: break-word; = -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; = "><br><div><div>On 4 Oct 2013, at 19:55, <a = href=3D"mailto:emitor@gmail.com">emitor@gmail.com</a> wrote:</div><br = class=3D"Apple-interchange-newline"><blockquote type=3D"cite"><div = dir=3D"ltr">Dan, I've made a whole new installation of oVirt 3.3 using = the stable repos, I've updated libvirt to 1.1.2-1 and configured a = glusterFS cluster.<div><br></div><div>I've just tried the live migration = of the Windows Host, here i send you the vdsm, libvir and qemu = logs</div></div></blockquote><div><br></div>hmm=85I see both PCI address = for CDROM as we've been seeing on gluster setups. So far no update from = gluster AFAIK=85it seems to be broken on their side = somewhere=85</div><div>and I also see error on migrations dst which = seems to be another/different disk device mess = up</div><div><div><br></div><div>The W2K8 create VM call in the ovirt1's = log is already wrong, can you capture/include the initial create call? = In the previous cases it was always the case that the VM was = created(from engine) with IDE address but then after the first run it = somehow changed to PCI on the first status update from = vdsm</div><div><br></div><div>Thanks,</div><div>michal</div><div><br></div= href=3D"mailto:danken@redhat.com" = target=3D"_blank">danken@redhat.com</a>></span><br><blockquote = class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc = solid;padding-left:1ex"> <div class=3D"im">On Wed, Oct 02, 2013 at 01:36:56PM -0300, <a = href=3D"mailto:emitor@gmail.com">emitor@gmail.com</a> wrote:<br> > I've configured the host as a NFS cluster. I've imported the VMs, = the<br> > debian now it's migrating without problems, the pfSense now gives = me an<br> > error (Migration failed due to Error: Desktop does not exist) and = the<br> > Windows VM doesn't start in the new cluster after import it.<br> ><br> > I've attached the logs of the vdsm<br> <br> </div>Thanks for providing your, but they begin too late.<br> <br> The bug is still a mystery to me. Something garbles Vdsm's internal<br> representation of virtual disks. And it's bad.<br> <br> I would appreciate if you could dig backward in your logs and find = the<br> last successfull startup of a VM, say W2K8_cluster1<br> (86f10137-d917-4508-a07a-3f83e6f24a95). Then, browse forward to find = the<br> first failure (be it on next startup of VM or during migration).<br> <br> Best regards,<br> Dan.<br> </blockquote></div><br><br clear=3D"all"><div><br></div>-- <br><div = dir=3D"ltr"><b>Emiliano Tortorella</b><br>+598 98941176<br><a = href=3D"mailto:emitor@gmail.com" = target=3D"_blank">emitor@gmail.com</a></div> </div> = <span><ovirt_logs_clusterGFS.tar.gz></span>_________________________= ______________________<br>Users mailing list<br><a = href=3D"mailto:Users@ovirt.org">Users@ovirt.org</a><br>http://lists.ovirt.= org/mailman/listinfo/users<br></blockquote></div><br></body></html>= --Apple-Mail=_80349546-76AA-4918-9B45-6194CB54F8F2--