[Users] all-in-one install fail

[Itamar Heim]

On 05/20/2013 04:46 PM, Ryan Wilkinson wrote:
> So cpu in fact does pass validation?  Could my issue be the network
> interface name - p1p1?  Usually it's em1 but on this machine for some
> reason was named differently..

current check is for cpu family, it doesn't cover VT being disabled by 
bios (which is what i understood you failed on, and simply need to 
enable at bios and do a cold boot)

>
>
> On Mon, May 20, 2013 at 3:05 AM, Sandro Bonazzola <sbonazzo at redhat.com
> <mailto:sbonazzo at redhat.com>> wrote:
>
>     Il 20/05/2013 10:02, Dan Kenigsberg ha scritto:
>      > On Mon, May 20, 2013 at 10:34:35AM +0300, Itamar Heim wrote:
>      >> On 05/20/2013 10:01 AM, Sandro Bonazzola wrote:
>      >>> Il 17/05/2013 17:48, Itamar Heim ha scritto:
>      >>>> On 05/17/2013 06:01 PM, Sandro Bonazzola wrote:
>      >>>>> Hi,
>      >>>>>   looking at the logs seems you're trying to install
>     all-in-one on a
>      >>>>> host that doesn't support virtualization:
>      >>>>>
>      >>>>> 2013-05-17 07:57:56,749 ERROR
>      >>>>> [org.ovirt.engine.core.bll.InstallerMessages] (VdsDeploy)
>     Installation
>      >>>>> ovirt.wilk.local: Failed to execute stage 'Setup validation':
>     Hardware
>      >>>>> does not support virtualization
>      >>>>>
>      >>>>> This seems to be the reason for
>      >>>>>
>      >>>>> Error: Host was found in a 'Failed' state.
>      >>>> wasn't this the first step in AIO - to validate this?
>      >>>>
>      >>> AIO ask libvirtd about supported cpus, this is the log about
>     the cpu
>      >>> validation:
>      >>>
>      >>> 2013-05-17 07:54:31::INFO::all_in_one_100::203::root:: libvirt
>     service
>      >>> is started
>      >>> 2013-05-17 07:54:31::DEBUG::setup_sequences::59::root:: running
>      >>> getSupportedCpus
>      >>> 2013-05-17 07:54:31::DEBUG::all_in_one_100::393::root::
>     Attempting to
>      >>> load the caps vdsm module
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::403::root::
>     Supported CPU
>      >>> models are:
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::405::root::
>     model_athlon
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::405::root::
>     model_Opteron_G1
>      >>> 2013-05-17 07:54:32::DEBUG::setup_sequences::59::root:: running
>     getCPUFamily
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::414::root::
>     Determening the
>      >>> CPU family supported by the host
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Westmere with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Nehalem with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Penryn with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Conroe with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Opteron_G3 with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Opteron_G2 with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::434::root:: comparing
>      >>> model_Opteron_G1 with supported models
>      >>> 2013-05-17 07:54:32::DEBUG::all_in_one_100::436::root::
>     Supported model
>      >>> family is: AMD Opteron G1
>      >>>
>      >>> so it passed cpu validation.
>      >> ok, that's one part of cpu validation for the cpu model. VDSM (or
>      >> the old bootstrap code) was also supposed to have one for VT enabled
>      >> in bios.
>      > The old bootstrap code indeed has virtEnabledInCpuAndBios().
>      > ovirt-host-deploy has a parallel VdsmEnv.CHECK_VIRT_HARDWARE optional
>      > check.
>      >
>      > HTH,
>      > Dan.
>
>     Ok, I'll add this check in AIO plugin.
>     Has anybody already opened a bug about this issue?
>
>     --
>     Sandro Bonazzola
>     Better technology. Faster innovation. Powered by community
>     collaboration.
>     See how it works at redhat.com <http://redhat.com>
>
>