From: Mel Gorman <mgorman@suse.de>
To: dormando <dormando@rydia.net>
Cc: Christoph Lameter <cl@linux.com>,
Andrew Morton <akpm@linux-foundation.org>,
Jiri Slaby <jslaby@suse.cz>,
Valdis Kletnieks <Valdis.Kletnieks@vt.edu>,
Rik van Riel <riel@redhat.com>,
Zlatko Calusic <zcalusic@bitsync.net>,
Johannes Weiner <hannes@cmpxchg.org>,
Satoru Moriya <satoru.moriya@hds.com>,
Michal Hocko <mhocko@suse.cz>, Linux-MM <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 0/10] Reduce system disruption due to kswapd V2
Date: Thu, 11 Apr 2013 10:10:44 +0100 [thread overview]
Message-ID: <20130411091044.GG3710@suse.de> (raw)
In-Reply-To: <alpine.DEB.2.02.1304101524120.7738@dtop>
On Wed, Apr 10, 2013 at 03:28:32PM -0700, dormando wrote:
> > On Tue, Apr 09, 2013 at 05:27:18PM +0000, Christoph Lameter wrote:
> > > One additional measure that may be useful is to make kswapd prefer one
> > > specific processor on a socket. Two benefits arise from that:
> > >
> > > 1. Better use of cpu caches and therefore higher speed, less
> > > serialization.
> > >
> >
> > Considering the volume of pages that kswapd can scan when it's active
> > I would expect that it trashes its cache anyway. The L1 cache would be
> > flushed after scanning struct pages for just a few MB of memory.
> >
> > > 2. Reduction of the disturbances to one processor.
> > >
> >
> > I've never checked it but I would have expected kswapd to stay on the
> > same processor for significant periods of time. Have you experienced
> > problems where kswapd bounces around on CPUs within a node causing
> > workload disruption?
>
> When kswapd shares the same CPU as our main process it causes a measurable
> drop in response time (graphs show tiny spikes at the same time memory is
> freed). Would be nice to be able to ensure it runs on a different core
> than our latency sensitive processes at least. We can pin processes to
> subsets of cores but I don't think there's a way to keep kswapd from
> waking up on any of them?
I've never tried it myself but does the following work?
taskset -p MASK `pidof kswapd`
where MASK is a cpumask describing what CPUs kswapd can run on?
Obviously care should be taken to ensure that you bind kswapd to a CPU
running on the node kswapd cares about.
--
Mel Gorman
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2013-04-11 9:10 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-09 11:06 [PATCH 0/10] Reduce system disruption due to kswapd V2 Mel Gorman
2013-04-09 11:06 ` [PATCH 01/10] mm: vmscan: Limit the number of pages kswapd reclaims at each priority Mel Gorman
2013-04-09 13:27 ` Michal Hocko
2013-04-10 6:47 ` Kamezawa Hiroyuki
2013-04-09 11:06 ` [PATCH 02/10] mm: vmscan: Obey proportional scanning requirements for kswapd Mel Gorman
2013-04-10 7:16 ` Kamezawa Hiroyuki
2013-04-10 14:08 ` Mel Gorman
2013-04-11 0:14 ` Kamezawa Hiroyuki
2013-04-11 9:09 ` Mel Gorman
2013-04-09 11:06 ` [PATCH 03/10] mm: vmscan: Flatten kswapd priority loop Mel Gorman
2013-04-10 7:47 ` Kamezawa Hiroyuki
2013-04-10 13:29 ` Mel Gorman
2013-04-12 2:45 ` Rik van Riel
2013-04-09 11:06 ` [PATCH 04/10] mm: vmscan: Decide whether to compact the pgdat based on reclaim progress Mel Gorman
2013-04-10 8:05 ` Kamezawa Hiroyuki
2013-04-10 13:57 ` Mel Gorman
2013-04-12 2:46 ` Rik van Riel
2013-04-09 11:07 ` [PATCH 05/10] mm: vmscan: Do not allow kswapd to scan at maximum priority Mel Gorman
2013-04-09 11:07 ` [PATCH 06/10] mm: vmscan: Have kswapd writeback pages based on dirty pages encountered, not priority Mel Gorman
2013-04-12 2:51 ` Rik van Riel
2013-04-09 11:07 ` [PATCH 07/10] mm: vmscan: Block kswapd if it is encountering pages under writeback Mel Gorman
2013-04-12 2:54 ` Rik van Riel
2013-04-09 11:07 ` [PATCH 08/10] mm: vmscan: Have kswapd shrink slab only once per priority Mel Gorman
2013-04-09 11:07 ` [PATCH 09/10] mm: vmscan: Check if kswapd should writepage once per pgdat scan Mel Gorman
2013-04-09 11:07 ` [PATCH 10/10] mm: vmscan: Move logic from balance_pgdat() to kswapd_shrink_zone() Mel Gorman
2013-04-12 2:56 ` Rik van Riel
2013-04-09 17:27 ` [PATCH 0/10] Reduce system disruption due to kswapd V2 Christoph Lameter
2013-04-10 14:14 ` Mel Gorman
2013-04-10 22:28 ` dormando
2013-04-10 23:46 ` KOSAKI Motohiro
2013-04-11 9:10 ` Mel Gorman [this message]
2013-04-11 20:13 ` Michal Hocko
2013-04-11 20:55 ` Zlatko Calusic
2013-04-12 19:40 ` Mel Gorman
2013-04-12 19:52 ` Mel Gorman
2013-04-12 20:07 ` Zlatko Calusic
2013-04-12 20:41 ` Mel Gorman
2013-04-12 21:14 ` Zlatko Calusic
2013-04-22 6:37 ` Zlatko Calusic
2013-04-22 6:43 ` Simon Jeons
2013-04-22 6:54 ` Zlatko Calusic
2013-04-22 7:12 ` Simon Jeons
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=20130411091044.GG3710@suse.de \
--to=mgorman@suse.de \
--cc=Valdis.Kletnieks@vt.edu \
--cc=akpm@linux-foundation.org \
--cc=cl@linux.com \
--cc=dormando@rydia.net \
--cc=hannes@cmpxchg.org \
--cc=jslaby@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.cz \
--cc=riel@redhat.com \
--cc=satoru.moriya@hds.com \
--cc=zcalusic@bitsync.net \
/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).