Re: [RFD] Isolated memory cgroups again

[Michal Hocko <mhocko@suse.cz>]

On Fri 21-10-11 12:39:22, Glauber Costa wrote:
> On 10/21/2011 03:41 AM, Ying Han wrote:
> >On Wed, Oct 19, 2011 at 6:33 PM, Michal Hocko<mhocko@suse.cz>  wrote:
[...]
> >>TODO
[...]
> >>- is bool sufficient. Don't we rather want something like priority
> >>  instead?
[...]
> >Hi Michal:
> >
> >I didn't read through the patch itself but only the description. If we
> >wanna protect a memcg being reclaimed from under global memory
> >pressure, I think we can approach it by making change on soft_limit
> >reclaim.
> >
> >I have a soft_limit change built on top of Johannes's patchset, which
> >does basically soft_limit aware reclaim under global memory pressure.
> >The implementation is simple, and I am looking forward to discuss more
> >with you guys in the conference.
> >
> >--Ying
> I don't think soft limits will help his case, if I know understand
> it correctly. Global reclaim can be triggered regardless of any soft
> limits we may set.
> 
> Now, there are two things I still don't like about it:
> * The definition of a "main workload", "main cgroup", or anything
> like that.

This was just because I wanted to point out the particular case that I
am interested in. You can of course setup more cgroups to be isolated
and balance them by the soft limit.

> I'd prefer to rank them according to some parameter,
> something akin to swapiness. This would allow for other people to
> use it in a different way, while still making you capable of
> reaching your goals through parameter settings (i.e. one cgroup has
> a high value of reclaim, all others, a much lower one)

Yes, this has been mentioned in the patch TODO section (above). I wanted
the first post to be as easy as possible for the discussion starter. I
guess that we really need something like priority in fact.

> 
> * The fact that you seem to want to *skip* reclaim altogether for a
> cgroup. That's a dangerous condition, IMHO. What I think we should
> try to achieve, is "skip it for practical purposes on sane
> workloads". 

Yes the feature might be dangerous (we provide many ways to shoot self
toes already ;)) but that is what you get if you want to guarantee
something.
But I agree, I guess we can be more clever and if it is priority based
we can map isolation priorities to the reclaim priorities somehow.

> Again, a parameter that when set to a very high mark, has the effect
> of disallowing reclaim for a cgroup under most sane circumstances.
> 
> What do you think of the above, Michal ?

Yes I guess that priority based isolation is the way to go. We should,
however, start with a consensus in this regard (should we do something
like that at all?).

Thanks
-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic