[Users] oVIrt 3.1 - Xeon E5530 - Wrong cpu identification

Dan Kenigsberg danken at redhat.com
Sun Aug 12 11:25:19 UTC 2012


On Thu, Aug 09, 2012 at 05:22:09PM +0100, Ricardo Esteves wrote:
> [root at blade4 ~]# virsh -r cpu-compare /tmp/cpu.xml
> Host CPU is a superset of CPU described in /tmp/cpu.xml
> 
> -----Original Message-----
> From: Dan Kenigsberg <danken at redhat.com>
> To: Itamar Heim <iheim at redhat.com>, gal at redhat.com
> Cc: Ricardo Esteves <ricardo.m.esteves at gmail.com>, users at ovirt.org
> Subject: Re: [Users] oVIrt 3.1 - Xeon E5530 - Wrong cpu identification
> Date: Thu, 9 Aug 2012 18:57:29 +0300
> 
> 
> On Thu, Aug 09, 2012 at 05:17:14PM +0300, Itamar Heim wrote:
> > On 08/09/2012 04:17 PM, Ricardo Esteves wrote:
> > >
> > >Ok, i fixed the ssl problem, my ovirt manager machine iptables was
> > >blocking the 8443 port.
> > >
> > >I also reinstalled the lastest version of the node
> > >(ovirt-node-iso-2.5.1-1.0.fc17.iso), but ovirt manager still doesn't
> > >recognize the CPU.
> > >
> > >The host status remains Non Operational :
> > >
> > >Host localhost.localdomain moved to Non-Operational state as host does
> > >not meet the cluster's minimum CPU level. Missing CPU features :
> > >model_Nehalem
> > 
> > danken - libvirt reports nehalem, yet vdsm reports conroe?
> 
> Interesting...
> 
> If yo put
>     <cpu match="minimum"><model>Nehalem</model><vendor>Intel</vendor></cpu>
> in /tmp/cpu.xml
> 
> what does
> 
>     virsh -r cpu-compare /tmp/cpu.xml
> 
> report?

(top posting make it very difficult to follow a long thread)

I have a hunch that this is something that has been fixed by
    http://gerrit.ovirt.org/#/c/5035/

    Find vendor for all cpu models, including those based on another
    cpu module.

it is accpeted upstream, but unfortunately did not make it into the ovirt-3.1 release.
Could you apply the patch to Vdsm and see if it fixes the reported cpu level?

Dan.



More information about the Users mailing list