Home » Mailing lists » Devel » [PATCH] BC: resource beancounters (v2)
| Re: BC: resource beancounters (v2) [message #5661 is a reply to message #5655] |
Fri, 25 August 2006 16:30   |
Andrey Savochkin
Messages: 47
Registered: December 2005
|
Member |
|
|
On Fri, Aug 25, 2006 at 07:30:03AM -0700, Andrew Morton wrote:
> On Fri, 25 Aug 2006 15:49:15 +0400
> Kirill Korotaev <dev@sw.ru> wrote:
>
> > Andrey Savochkin wrote already a brief summary on vm resource management:
> >
> > ------------- cut ----------------
> > The task of limiting a container to 4.5GB of memory bottles down to the
> > question: what to do when the container starts to use more than assigned
> > 4.5GB of memory?
> >
> > At this moment there are only 3 viable alternatives.
> >
> > A) Have separate memory management for each container,
> > with separate buddy allocator, lru lists, page replacement mechanism.
> > That implies a considerable overhead, and the main challenge there
> > is sharing of pages between these separate memory managers.
> >
> > B) Return errors on extension of mappings, but not on page faults, where
> > memory is actually consumed.
> > In this case it makes sense to take into account not only the size of used
> > memory, but the size of created mappings as well.
> > This is approximately what "privvmpages" accounting/limiting provides in
> > UBC.
> >
> > C) Rely on OOM killer.
> > This is a fall-back method in UBC, for the case "privvmpages" limits
> > still leave the possibility to overload the system.
> >
>
> D) Virtual scan of mm's in the over-limit container
>
> E) Modify existing physical scanner to be able to skip pages which
> belong to not-over-limit containers.
I've actually tried (E), but it didn't work as I wished.
It didn't handle well shared pages.
Then, in my experiments such modified scanner was unable to regulate
quality-of-service. When I ran 2 over-the-limit containers, they worked
equally slow regardless of their limits and work set size.
That is, I didn't observe a smooth transition "under limit, maximum
performance" to "slightly over limit, a bit reduced performance" to
"significantly over limit, poor performance". Neither did I see any fairness
in how containers got penalized for exceeding their limits.
My explanation of what I observed is that
- since filesystem caches play a huge role in performance, page scanner will
be very limited in controlling container's performance if caches
stay shared between containers,
- in the absence of decent disk I/O manager, stalls due to swapin/swapout
are more influenced by disk subsystem than by page scanner policy.
So in fact modified page scanner provides control over memory usage only as
"stay under limits or die", and doesn't show many advantages over (B) or (C).
At the same time, skipping pages visibly penalizes "good citizens", not only
in disk bandwidth but in CPU overhead as well.
So I settled for (A)-(C) for now.
But it certainly would be interesting to hear if someone else makes such
experiments.
Best regards
Andrey
|
|
|
|
 |
|
[PATCH] BC: resource beancounters (v2)
By: dev on Wed, 23 August 2006 10:44 |
|
|
[PATCH 1/6] BC: kconfig
By: dev on Wed, 23 August 2006 10:59 |
|
|
Re: [PATCH 1/6] BC: kconfig
|
|
|
Re: [PATCH 1/6] BC: kconfig
|
|
|
Re: [PATCH 1/6] BC: kconfig
|
|
|
Re: [PATCH 1/6] BC: kconfig
By: dev on Fri, 25 August 2006 11:27 |
|
|
Re: [PATCH 1/6] BC: kconfig
By: dev on Fri, 25 August 2006 11:31 |
|
|
Re: [PATCH 1/6] BC: kconfig
By: rdunlap on Wed, 23 August 2006 22:29 |
|
|
Re: [PATCH 1/6] BC: kconfig
|
|
|
Re: [PATCH 1/6] BC: kconfig
|
|
|
Re: [ckrm-tech] [PATCH 1/6] BC: kconfig
By: dev on Thu, 24 August 2006 11:47 |
|
|
Re: [ckrm-tech] [PATCH 1/6] BC: kconfig
|
|
|
[PATCH 2/6] BC: beancounters core (API)
By: dev on Wed, 23 August 2006 11:00 |
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
By: dev on Wed, 23 August 2006 13:25 |
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
By: dev on Wed, 23 August 2006 13:46 |
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
By: dev on Thu, 24 August 2006 12:03 |
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
By: dev on Fri, 25 August 2006 10:51 |
|
|
Re: [PATCH 2/6] BC: beancounters core (API)
|
|
|
[PATCH 3/6] BC: context inheriting and changing
By: dev on Wed, 23 August 2006 11:02 |
|
|
[PATCH 4/6] BC: user interface (syscalls)
By: dev on Wed, 23 August 2006 11:03 |
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
By: dev on Wed, 23 August 2006 13:40 |
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
|
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
By: Alan Cox on Wed, 23 August 2006 17:08 |
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
|
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
By: Alan Cox on Thu, 24 August 2006 10:42 |
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
|
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
By: dev on Fri, 25 August 2006 10:54 |
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
|
|
|
Re: [PATCH 4/6] BC: user interface (syscalls)
|
|
|
[PATCH 5/6] BC: kernel memory accounting (core)
By: dev on Wed, 23 August 2006 11:04 |
|
|
Re: [PATCH 5/6] BC: kernel memory accounting (core)
|
|
|
Re: [PATCH 5/6] BC: kernel memory accounting (core)
|
|
|
Re: [PATCH 5/6] BC: kernel memory accounting (core)
By: dev on Fri, 25 August 2006 10:06 |
|
|
[PATCH 6/6] BC: kernel memory accounting (marks)
By: dev on Wed, 23 August 2006 11:05 |
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
|
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
By: dev on Tue, 29 August 2006 09:52 |
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
|
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
|
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
|
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
|
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
By: dev on Tue, 29 August 2006 14:34 |
|
|
Re: [PATCH 6/6] BC: kernel memory accounting (marks)
By: dev on Tue, 29 August 2006 15:53 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
By: dev on Fri, 25 August 2006 11:47 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
By: dev on Mon, 28 August 2006 08:27 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: BC: resource beancounters (v2)
|
|
|
Re: BC: resource beancounters (v2)
|
|
|
Re: BC: resource beancounters (v2)
|
|
|
Re: BC: resource beancounters (v2)
|
|
|
Re: BC: resource beancounters (v2)
By: Alan Cox on Sat, 26 August 2006 16:16 |
|
|
Re: BC: resource beancounters (v2)
|
|
|
Re: Re: BC: resource beancounters (v2)
By: kir on Mon, 28 August 2006 17:40 |
|
|
Re: Re: BC: resource beancounters (v2)
|
|
|
Re: Re: BC: resource beancounters (v2)
By: Alan Cox on Tue, 29 August 2006 10:15 |
|
|
Re: Re: BC: resource beancounters (v2)
|
|
|
Re: Re: BC: resource beancounters (v2)
By: Alan Cox on Tue, 29 August 2006 18:46 |
|
|
Re: Re: BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
By: dev on Tue, 29 August 2006 15:33 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
|
|
Re: [PATCH] BC: resource beancounters (v2)
By: Alan Cox on Thu, 24 August 2006 10:38 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
By: Alan Cox on Fri, 25 August 2006 15:36 |
|
|
Re: [PATCH] BC: resource beancounters (v2)
|
Goto Forum:
Current Time: Mon Jul 14 20:16:14 GMT 2025
Total time taken to generate the page: 0.01537 seconds
|
OpenVZ Forum
Members
Search
Help
Register
Login
Home
