Re: Summary of resource management discussion

[Herbert Poetzl <herbert@13thfloor.at>]

On Tue, Mar 13, 2007 at 11:28:20PM +0530, Srivatsa Vaddagiri wrote:
> On Tue, Mar 13, 2007 at 05:24:59PM +0100, Herbert Poetzl wrote:
> > what about identifying different resource categories and
> > handling them according to the typical usage pattern?
> > 
> > like the following:
> > 
> >  - cpu and scheduler related accounting/limits
> >  - memory related accounting/limits
> >  - network related accounting/limits
> >  - generic/file system related accounting/limits
> > 
> > I don't worry too much about having the generic/file stuff
> > attached to the nsproxy, but the cpu/sched stuff might be
> > better off being directly reachable from the task
> 
> I think we should experiment with both combinations (a direct pointer
> to cpu_limit structure from task_struct and an indirect pointer), get
> some numbers and then decide. Or do you have results already with
> respect to that?

nope, no numbers for that, but I appreciate some testing
and probably can do some testing in this regard too
(although I want to get some testing done for the resource
 sharing between guests first)

> > > 3. How are cpusets related to vserver/containers?
> > > 
> > > 	Should it be possible to, lets say, create exclusive cpusets and
> > > 	attach containers to different cpusets?
> > 
> > that is what Linux-VServer does atm, i.e. you can put
> > an entire guest into a specific cpu set 
> 
> Interesting. What abt /dev/cpuset view? 

host only for now

best,
Herbert

> Is that same for all containers or do you restrict that view 
> to the containers cpuset only?
> 
> -- 
> Regards,
> vatsa
> _______________________________________________
> Containers mailing list
> Containers@lists.osdl.org
> https://lists.osdl.org/mailman/listinfo/containers