LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Nick Piggin <nickpiggin@yahoo.com.au>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@elte.hu>, "LKML," <linux-kernel@vger.kernel.org>
Subject: Re: Poor PostgreSQL scaling on Linux 2.6.25-rc5 (vs 2.6.22)
Date: Mon, 17 Mar 2008 11:44:35 +1100	[thread overview]
Message-ID: <200803171144.35479.nickpiggin@yahoo.com.au> (raw)
In-Reply-To: <1205308704.8514.197.camel@twins>

On Wednesday 12 March 2008 18:58, Peter Zijlstra wrote:
> On Wed, 2008-03-12 at 12:21 +1100, Nick Piggin wrote:
> > (Back onto lkml)
> >
> > On Tuesday 11 March 2008 23:02, Ingo Molnar wrote:
> > > another thing to try would be to increase:
> > >
> > >   /proc/sys/kernel/sched_migration_cost
> > >
> > > from its 500 usecs default to a few msecs ?
> >
> > This doesn't really help either (at 10ms).
> >
> > (For the record, I've tried turning SD_WAKE_IDLE, SD_WAKE_AFFINE
> > on and off for each domain and that hasn't helped either).
> >
> > I've also tried increasing sched_latency_ns as far as it can go.
> > BTW. this is a pretty nasty behaviour if you ask my opinion. It
> > starts *increasing* the number of involuntary context switches
> > as resources get oversubscribed. That's completely unintuitive as
> > far as I can see -- when we get overloaded, the obvious thing to
> > do is try to increase efficiency, or at least try as hard as
> > possible not to lose it. So context switches should be steady or
> > decreasing as I add more processes to a runqueue.
> >
> > It seems to max out at nearly 100 context switches per second,
> > and this has actually shown to be too frequent for modern CPUs
> > and big caches.
> >
> > Increasing the tunable didn't help for this workload, but it really
> > needs to be fixed so it doesn't decrease timeslices as the number
> > of processes increases.
>
> /proc/sys/kernel/sched_min_granularity_ns
> /proc/sys/kernel/sched_latency_ns
>
> period := max(latency, nr_running * min_granularity)
> slice := period * w_{i} / W
> W := \Sum_{i} w_{i}
>
> So if you want to increase the slice length for loaded systems, up
> min_granularity.

OK, but the very concept of reducing efficiency when load increases
is nasty, and leads to nasty feedback loops. It's just a very bad
behaviour to have out of the box, and as a general observation, 10ms
is too short a default timeslice IMO.

I don't see how it is really helpful for interactive processes either.
By definition, if they are not CPU bound, then they should be run
quite soon after waking up; if they are CPU bound, then reducing
efficiency by increasing context switches is effectively going to
increase their latency anyway.

Can this be changed by default, please?


  reply	other threads:[~2008-03-17  1:00 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-03-11  6:49 Nick Piggin
2008-03-11  7:58 ` Ingo Molnar
2008-03-11  8:28   ` Nick Piggin
2008-03-11  9:59     ` Ingo Molnar
2008-03-11 21:07 ` Nicholas Miell
2008-03-11 21:34   ` Cyrus Massoumi
2008-03-11 23:12     ` Nicholas Miell
2008-03-11 23:42       ` Nick Piggin
2008-03-11 23:47       ` Stephen Hemminger
2008-03-12  9:00       ` Zhang, Yanmin
     [not found] ` <20080311102538.GA30551@elte.hu>
     [not found]   ` <20080311120230.GA5386@elte.hu>
2008-03-12  1:21     ` Nick Piggin
2008-03-12  7:58       ` Peter Zijlstra
2008-03-17  0:44         ` Nick Piggin [this message]
2008-03-17  5:16           ` Ray Lee
2008-03-17  5:21             ` Willy Tarreau
2008-03-17  7:19               ` Nick Piggin
2008-03-17  8:26                 ` Willy Tarreau
2008-03-17  8:54                   ` Nick Piggin
2008-03-17  9:28                     ` Peter Zijlstra
2008-03-17  9:56                       ` Nick Piggin
2008-03-17 10:16                       ` Ingo Molnar
2008-03-17  5:34             ` Nick Piggin
2008-03-14 15:42 Xose Vazquez Perez

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=200803171144.35479.nickpiggin@yahoo.com.au \
    --to=nickpiggin@yahoo.com.au \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=peterz@infradead.org \
    --subject='Re: Poor PostgreSQL scaling on Linux 2.6.25-rc5 (vs 2.6.22)' \
    /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

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).