mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Wu Fengguang <wfg@mail.ustc.edu.cn>
To: Andrew Morton <akpm@osdl.org>
Cc: linux-kernel@vger.kernel.org, christoph@lameter.com,
	riel@redhat.com, a.p.zijlstra@chello.nl, npiggin@suse.de,
	andrea@suse.de, marcelo.tosatti@cyclades.com,
	magnus.damm@gmail.com
Subject: Re: [PATCH 02/12] mm: supporting variables and functions for balanced zone aging
Date: Thu, 1 Dec 2005 20:11:01 +0800	[thread overview]
Message-ID: <20051201121101.GB6011@mail.ustc.edu.cn> (raw)
In-Reply-To: <20051201023714.612f0bbf.akpm@osdl.org>

On Thu, Dec 01, 2005 at 02:37:14AM -0800, Andrew Morton wrote:
> Wu Fengguang <wfg@mail.ustc.edu.cn> wrote:
> >
> >  The zone aging rates are currently imbalanced,
> 
> ZONE_DMA is out of whack.  It shouldn't be, and I'm not aware of anyone
> getting in and working out why.  I certainly wouldn't want to go and add
> all this stuff without having a good understanding of _why_ it's out of
> whack.  Perhaps it's just some silly bug, like the thing I pointed at in
> the previous email.

Yep, my rule is that if ever the DMA zone is reclaimed for watermark, it will
be running wild ;) So I leave it out by setting classzone_idx=0, and let the
age balancing code to catch it up. This scheme works fine: tested to be OK from
64M to 2G memory.

> > the gap can be as large as 3 times,
> 
> What's the testcase?
> 
> > which can severely damage read-ahead requests and shorten their
> >  effective life time.
> 
> Have you any performance numbers for this?

That's months ago, if I remember it right, the number of concurrent readers the
adaptive read-ahead code can handle without much thrashing was raised from ~100
to 800 with the balancing work.

This is my original announce back then:

    The page aging problem showed up when I was testing many slow reads with
    limited memory. Pages in the DMA zone were found out to be aged about 3
    times faster than that of Normal zone in systems with 64-512M memory.
    That is a BIG threat to the read-ahead pages. So I added some code to
    make the aging rates synchronized. You can see the effect by running:
    $ tar c / | cat > /dev/null &
    $ watch -n1 'grep "age " /proc/zoneinfo'
    There are still some extra DMA scans in the direct page reclaim path.
    It tend to happen in large memory system and is therefore not a big
    problem.

And here is some numbers collected by Magnus Damm:
http://lkml.org/lkml/2005/10/25/50

Good night!
Wu

  reply	other threads:[~2005-12-01 12:03 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-12-01 10:18 [PATCH 00/12] Balancing the scan rate of major caches Wu Fengguang
2005-12-01 10:18 ` [PATCH 01/12] vm: kswapd incmin Wu Fengguang
2005-12-01 10:33   ` Andrew Morton
2005-12-01 11:40     ` Wu Fengguang
2005-12-01 10:18 ` [PATCH 02/12] mm: supporting variables and functions for balanced zone aging Wu Fengguang
2005-12-01 10:37   ` Andrew Morton
2005-12-01 12:11     ` Wu Fengguang [this message]
2005-12-01 22:28     ` Marcelo Tosatti
2005-12-01 23:03       ` Andrew Morton
2005-12-02  1:19         ` Wu Fengguang
2005-12-02  1:30           ` Andrew Morton
2005-12-02  2:04             ` Wu Fengguang
2005-12-02  2:18               ` Andrea Arcangeli
2005-12-02  2:37                 ` Wu Fengguang
2005-12-02  2:52                   ` Andrea Arcangeli
2005-12-02  4:45                 ` Andrew Morton
2005-12-02  6:38                   ` Wu Fengguang
2005-12-02  2:27               ` Nick Piggin
2005-12-02  2:36                 ` Andrea Arcangeli
2005-12-02  2:43                 ` Wu Fengguang
2005-12-02  5:49           ` Andrew Morton
2005-12-02  7:18             ` Wu Fengguang
2005-12-02  7:27               ` Andrew Morton
2005-12-02 15:13             ` Marcelo Tosatti
2005-12-02 21:39               ` Andrew Morton
2005-12-03  0:26                 ` Marcelo Tosatti
2005-12-04  6:06                   ` Wu Fengguang
2005-12-02  1:26         ` Marcelo Tosatti
2005-12-02  3:40           ` Andrew Morton
2005-12-01 10:18 ` [PATCH 03/12] mm: balance zone aging in direct reclaim path Wu Fengguang
2005-12-01 10:18 ` [PATCH 04/12] mm: balance zone aging in kswapd " Wu Fengguang
2005-12-01 10:18 ` [PATCH 05/12] mm: balance slab aging Wu Fengguang
2005-12-01 10:18 ` [PATCH 06/12] mm: balance active/inactive list scan rates Wu Fengguang
2005-12-01 11:39   ` Peter Zijlstra
2005-12-01 10:18 ` [PATCH 07/12] mm: remove unnecessary variable and loop Wu Fengguang
2005-12-01 10:18 ` [PATCH 08/12] mm: remove swap_cluster_max from scan_control Wu Fengguang
2005-12-01 10:18 ` [PATCH 09/12] mm: accumulate sc.nr_scanned/sc.nr_reclaimed Wu Fengguang
2005-12-01 10:18 ` [PATCH 10/12] mm: merge sc.may_writepage and sc.may_swap into sc.flags Wu Fengguang
2005-12-01 10:18 ` [PATCH 11/12] mm: add page reclaim debug traces Wu Fengguang
2005-12-01 10:18 ` [PATCH 12/12] mm: fix minor scan count bugs Wu Fengguang

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=20051201121101.GB6011@mail.ustc.edu.cn \
    --to=wfg@mail.ustc.edu.cn \
    --cc=a.p.zijlstra@chello.nl \
    --cc=akpm@osdl.org \
    --cc=andrea@suse.de \
    --cc=christoph@lameter.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=magnus.damm@gmail.com \
    --cc=marcelo.tosatti@cyclades.com \
    --cc=npiggin@suse.de \
    --cc=riel@redhat.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

all inboxes | Powered by JetHome®