On 08/09/2013 02:32 PM, Michael R. Hines wrote:

>>>               "total-time":12345,
>>> +            "setup-time":12345,
>> Theoretically, if setup-time == total-time, then we are still in the
>> setup phase, right?  The example might be more realistic if it uses a
>> smaller number for setup time.  But at the end of the day, I don't
>> actually care enough to force a respin, when we are this close to the
>> 1.6 deadline and this deserves to be in 1.6.
>>
> 
> Ooops - well, when I was reading the file - all the numbers seemed like
> place-holders, not realistic numbers.

Indeed, which is why I'm okay with no respin.

> 
> This documentation is kind of hard to maintain in the first place -
> something like this is more the job of a tool like 'virt-test' which
> could actually exercise QMP and dump the output during a test cycle in
> case the user actually wanted to see what the migration life cycle
> actually looks like.

Yes, I'd love to have more of this file generated (why are we forced to
duplicate efforts into both the .json and the .hx files again?)

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org