From: Roman Gushchin <guro@fb.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: <linux-mm@kvack.org>, Vladimir Davydov <vdavydov.dev@gmail.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>,
David Rientjes <rientjes@google.com>, Tejun Heo <tj@kernel.org>,
<kernel-team@fb.com>, <cgroups@vger.kernel.org>,
<linux-doc@vger.kernel.org>, <linux-kernel@vger.kernel.org>
Subject: Re: [v6 3/4] mm, oom: introduce oom_priority for memory cgroups
Date: Thu, 24 Aug 2017 15:11:08 +0100 [thread overview]
Message-ID: <20170824141108.GB21167@castle.DHCP.thefacebook.com> (raw)
In-Reply-To: <20170824134859.GO5943@dhcp22.suse.cz>
On Thu, Aug 24, 2017 at 03:48:59PM +0200, Michal Hocko wrote:
> On Thu 24-08-17 13:51:13, Roman Gushchin wrote:
> > On Thu, Aug 24, 2017 at 02:10:54PM +0200, Michal Hocko wrote:
> > > On Wed 23-08-17 17:52:00, Roman Gushchin wrote:
> > > > Introduce a per-memory-cgroup oom_priority setting: an integer number
> > > > within the [-10000, 10000] range, which defines the order in which
> > > > the OOM killer selects victim memory cgroups.
> > >
> > > Why do we need a range here?
> >
> > No specific reason, both [INT_MIN, INT_MAX] and [-10000, 10000] will
> > work equally.
>
> Then do not enforce a range because this just reduces possible usecases
> (e.g. timestamp one...).
I agree.
>
> > We should be able to predefine an OOM killing order for
> > any reasonable amount of cgroups.
> >
> > >
> > > > OOM killer prefers memory cgroups with larger priority if they are
> > > > populated with eligible tasks.
> > >
> > > So this is basically orthogonal to the score based selection and the
> > > real size is only the tiebreaker for same priorities? Could you describe
> > > the usecase? Becasuse to me this sounds like a separate oom killer
> > > strategy. I can imagine somebody might be interested (e.g. always kill
> > > the oldest memcgs...) but an explicit range wouldn't fly with such a
> > > usecase very well.
> >
> > The usecase: you have a machine with several containerized workloads
> > of different importance, and some system-level stuff, also in (memory)
> > cgroups.
> > In case of global memory shortage, some workloads should be killed in
> > a first order, others should be killed only if there is no other option.
> > Several workloads can have equal importance. Size-based tiebreaking
> > is very useful to catch memory leakers amongst them.
>
> OK, please document that in the changelog.
Sure.
>
> > > That brings me back to my original suggestion. Wouldn't a "register an
> > > oom strategy" approach much better than blending things together and
> > > then have to wrap heads around different combinations of tunables?
> >
> > Well, I believe that 90% of this patchset is still relevant;
>
> agreed and didn't say otherwise.
>
> > the only
> > thing you might want to customize/replace size-based tiebreaking with
> > something else (like timestamp-based tiebreaking, mentioned by David earlier).
>
> > What about tunables, there are two, and they are completely orthogonal:
> > 1) oom_priority allows to define an order, in which cgroups will be OOMed
> > 2) oom_kill_all defines if all or just one task should be killed
> >
> > So, I don't think it's a too complex interface.
> >
> > Again, I'm not against oom strategy approach, it just looks as a much bigger
> > project, and I do not see a big need.
>
> Well, I was thinking that our current oom victim selection code is
> quite extensible already. Your patches will teach it kill the whole
> group semantic which is already very useful. Now we can talk about the
> selection criteria and this is something to be replaceable. Because even
> the current discussion has shown that different people might and will
> have different requirements. Can we structure the code in such a way
> that new comparison algorithm would be simple to add without reworking
> the whole selection logic?
I'd say that extended oom_priority range and potentially customizable
memcg_oom_badness() should do the job for memcgroups.
We can extract a part of the oom_badness() into something like this:
unsigned long task_oom_badness(struct task_struct *p)
{
/*
* The baseline for the badness score is the proportion of RAM that each
* task's rss, pagetable and swap space use.
*/
return get_mm_rss(p->mm) + get_mm_counter(p->mm, MM_SWAPENTS) +
atomic_long_read(&p->mm->nr_ptes) + mm_nr_pmds(p->mm);
}
Also, it would be nice to introduce an oom_priority for tasks, as David
suggested.
>
> > Do you have an example, which can't be effectively handled by an approach
> > I'm suggesting?
>
> No, I do not have any which would be _explicitly_ requested but I do
> envision new requirements will emerge. The most probable one would be
> kill the youngest container because that would imply the least amount of
> work wasted.
I agree, this a nice feature. It can be implemented in userspace
by setting oom_priority.
Thanks!
Roman
next prev parent reply other threads:[~2017-08-24 14:11 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-23 16:51 [v6 1/4] mm, oom: refactor the oom_kill_process() function Roman Gushchin
2017-08-23 16:51 ` [v6 0/4] cgroup-aware OOM killer Roman Gushchin
2017-08-23 16:51 ` [v6 2/4] mm, oom: " Roman Gushchin
2017-08-23 23:19 ` David Rientjes
2017-08-25 10:57 ` Roman Gushchin
2017-08-24 11:47 ` Michal Hocko
2017-08-24 12:28 ` Roman Gushchin
2017-08-24 12:58 ` Michal Hocko
2017-08-24 13:58 ` Roman Gushchin
2017-08-24 14:13 ` Michal Hocko
2017-08-24 14:58 ` Roman Gushchin
2017-08-25 8:14 ` Michal Hocko
2017-08-25 10:39 ` Roman Gushchin
2017-08-25 10:58 ` Michal Hocko
2017-08-30 11:22 ` Roman Gushchin
2017-08-30 20:56 ` David Rientjes
2017-08-31 13:34 ` Roman Gushchin
2017-08-31 20:01 ` David Rientjes
2017-08-23 16:52 ` [v6 3/4] mm, oom: introduce oom_priority for memory cgroups Roman Gushchin
2017-08-24 12:10 ` Michal Hocko
2017-08-24 12:51 ` Roman Gushchin
2017-08-24 13:48 ` Michal Hocko
2017-08-24 14:11 ` Roman Gushchin [this message]
2017-08-28 20:54 ` David Rientjes
2017-08-23 16:52 ` [v6 4/4] mm, oom, docs: describe the cgroup-aware OOM killer Roman Gushchin
2017-08-24 11:15 ` [v6 1/4] mm, oom: refactor the oom_kill_process() function Michal Hocko
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170824141108.GB21167@castle.DHCP.thefacebook.com \
--to=guro@fb.com \
--cc=cgroups@vger.kernel.org \
--cc=hannes@cmpxchg.org \
--cc=kernel-team@fb.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=penguin-kernel@I-love.SAKURA.ne.jp \
--cc=rientjes@google.com \
--cc=tj@kernel.org \
--cc=vdavydov.dev@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).