Re: Missing recalculation of scheduler tunables in case of cpu hot add/remove

[Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>]

Peter Zijlstra wrote:
> On Thu, 2009-11-26 at 17:10 +0100, Christian Ehrhardt wrote:
>
>   
>> What I consider more important at the moment is that there is no hook to 
>> recalculate these values in case cpu hot add/remove takes place.
>> As an example someone could boot a machine with one online cpu and get 
>> the low non scaled defaults, later on driven by load the system 
>> activates more and more processors. Therefore the system could end up 
>> having  a large amount of cpus with non recalculated scheduler tunables.
>>     
>
> This is virt junk that's playing dumb games with hotplug isn't it?
>   
Some sort of, its on s390 which does that all the time. By default there 
is a daemon that activates/deactivates cpus according to load to cover 
load peaks but also save virtualization overhead.
> Normal machines simply don't change their numbers of cpus, if they
> hotplug its usually for things like suspend or actual replacement of a
> faulty piece of kit, in which case there's little point in adjusting
> things.
>
>   
What is still "normal" today, you cant get s390 without virt so I would 
consider it normal and a real use case for us :-)
> Aside from that, we probably should put an upper limit in place, as I
> guess large cpu count machines get silly large values
I agree to that, but in the code is already an upper limit of 
200.000.000 - well we might discuss if that is too low/high.


-- 

Grüsse / regards, Christian Ehrhardt
IBM Linux Technology Center, Open Virtualization