[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH RFC v2 5/7] libxl/vNUMA: VM config parsing functions



On Thu, Oct 10, 2013 at 5:32 PM, Dario Faggioli
<dario.faggioli@xxxxxxxxxx> wrote:
> On gio, 2013-10-10 at 12:25 -0400, Elena Ufimtseva wrote:
>> On Thu, Oct 10, 2013 at 4:38 AM, Dario Faggioli
>> > Well, let's make it like this: if you provide 2 values, it acts as you
>> > say above; otherwise you have to specify all of them. If you should have
>> > specified 4 (or 16) values, and you specify, say, 3 (or, say, 10) all
>> > the unspecified ones will have the same default value (e.g., 10). How do
>> > you like this?
>>
>> Yes, two have both ways is better.
>>
> Ok, then.
>
>> I would even add the third way - if the number of values is power of
>> number of nodes,
>> take them as it is and expand by rows (as it is right now).
>>
> Sounds fine, although, I'm not sure I'm getting 100% of what you're
> saying. What do you mean with "the number of values is power of number
> of nodes"? Do you mind giving an example?

Apologies, used wrong words :)
I meant to say, third way is the one as before - user specifies all
values and we use them
as is.

>
>> so we have distances:
>>
>> 1) [10, 20] - only two values => same node 10, othes - 20;
>>
> Yep.
>
>> 2) did you mean, that if vnodes = 4, vdistance = [10, 10, 10]; the
>> rest of it should be 10?
>> Then it means all distances are the same?
>>
> As said on IRC, I mean that we should have a default value for the
> distances, in case the user does not say anything about them.
>
> If it says something, but not all of it, we can try doing something wise
> with what he gives us (which is exactly what we're doing above if we
> have only 2 values).
>
> However, if it's not immediate to translate what he says into something
> that we need, we either exit with error or use as much info as we can,
> and fill the rest with a default value (and print a warning).
>

Ok, let me see if I can come up with something quick :)

> Personally, I'm fine with both.
>
>> 3) if vdistance number of elements = vnodes * vnodes, take as it is.
>>
> Sure.
>
> Regards,
> Dario
>
> --
> <<This happens because I choose it to happen!>> (Raistlin Majere)
> -----------------------------------------------------------------
> Dario Faggioli, Ph.D, http://about.me/dario.faggioli
> Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)
>



-- 
Elena

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.