[RFC] [PATCH 0/3] Add group fairness to CFS

Ingo Molnar mingo at elte.hu
Wed May 23 11:38:24 PDT 2007


* Guillaume Chazarain <guichaz at yahoo.fr> wrote:

> Some observations:
> 
> o Doing an infinite loop as root seems to badly affect interactivity 
> much more than with a normal user. Note that this is subjective, so 
> maybe I'm smocking crack here.

hm, this shouldnt be the case. Can you see this with -v14?

> o Nice values are not reflected across users. From my test, if user1 
> has a single busy loop at nice 19, and user2 a single busy loop at 
> nice 0, both process will have a 50% CPU share, this looks wrong. Note 
> that I have no idea how to solve this one.

for containers it's exactly the right behavior: group scheduling is 
really a 'super' container concept that allows the allocation of CPU 
time regardless of how a group uses it. The only additional control we 
might want is to allocate different amount of CPU time to different 
groups. (i.e. a concept vaguely similar to "nice levels", but at the 
group level - using a different and saner API than nice levels.) Nice 
levels are really only meaningful at the lowest level.

for 'friendly users' it's perhaps not what we want - but those do not 
need to isolate themselves from each other anyway.

> Thanks for working in this very interesting direction.

seconded :)

	Ingo


More information about the Containers mailing list