Re: Live Migration of 32-bit Linux guest broken since 2.6.35-rc2

[Peter Lieven <pl@dlh.net>]

Avi Kivity wrote:
> On 06/08/2010 04:38 PM, Peter Lieven wrote:
>> Avi Kivity wrote:
>>> On 06/08/2010 04:28 PM, Peter Lieven wrote:
>>>>>
>>>>>> i will retry the case later today and send info register output.
>>>>>> what is the recommended value for nx (and why)?
>>>>>
>>>>> Enabled (so you get no-execute memory protection).
>>>>
>>>> do you have a guideline which flags should be identical to ensure 
>>>> proper live migration?
>>>> i would like to do an automatic compare before life migration to 
>>>> avoid crashes.
>>>
>>> All of them [1].  If you have an asymmetric machine, you can disable 
>>> those flags with the -cpu switch.
>>>
>>> Note the default qemu cpu disables many flags, so if one machine has 
>>> sse4.2 and the other doesn't, migration will still work without 
>>> disabling anything.
>> which cpu model do you suggest for 4 core xeon cpus? -cpu host ?
>
> -cpu host is good if you have identical machines and don't plan to add 
> new ones.
i will likely add new ones, but my plan would be to use qemu64 and then 
add all flags manually that
are common to all cpus in the pool.
would that be safe?

br,
peter

>
>> how important is the match of the virtualization extensions?
>
> The virtualization extensions on Intel aren't exposed to the guest, so 
> they don't need to match.
>


-- 
Mit freundlichen Grüßen/Kind Regards

Peter Lieven

..........................................................................................................

   KAMP Netzwerkdienste GmbH
   Vestische Str. 89-91 | 46117 Oberhausen
   Tel: +49 (0) 208.89 402-50 | Fax: +49 (0) 208.89 402-40
   mailto:pl@kamp.de | http://www.kamp.de

   Geschäftsführer: Heiner Lante | Michael Lante
   Amtsgericht Duisburg | HRB Nr. 12154
   USt-Id-Nr.: DE 120607556

.........................................................................................................