public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Jon Burgess <lkml@jburgess.uklinux.net>
To: Andrew Morton <akpm@osdl.org>
Cc: Jon Burgess <lkml@jburgess.uklinux.net>, linux-kernel@vger.kernel.org
Subject: Re: ext2/3 performance regression in 2.6 vs 2.4 for small interleaved writes
Date: Sat, 14 Feb 2004 15:00:27 +0000	[thread overview]
Message-ID: <402E380B.8070806@jburgess.uklinux.net> (raw)
In-Reply-To: <20040212015626.48631555.akpm@osdl.org>

Andrew Morton wrote:

>So you'll see that instead of a single full-bandwidth
>write, we do two half-bandwidth writes.  If it weren't for disk writeback
>caching, it would be as much as 4x slower.
>
Write caching does indeed make a big difference. Here is a test run on a 
drive with and without write caching (hdparm -W 0/1). The test was done 
on 2.6.2 with ext3 and shows the write speed in MB/s:

Write Cache   1 Stream   2 Streams
Enabled       21.54      3.66
Disabled      18.11      0.46  

The two stream case is almost 10x slower without write caching.
I don't think this explains the difference between 2.4 and 2.6 unless 
one of them changes the write cache mode of the drive.

    Jon


      parent reply	other threads:[~2004-02-14 15:00 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-02-11 19:04 ext2/3 performance regression in 2.6 vs 2.4 for small interleaved writes Jon Burgess
2004-02-11 20:28 ` Rik van Riel
2004-02-11 21:02   ` Michael Frank
2004-02-11 21:18     ` Diego Calleja
2004-02-12  2:00       ` Dave Olien
2004-02-12  2:23         ` Andrea Arcangeli
2004-02-12  9:42           ` ext2/3 performance regression in 2.6 vs 2.4 for small interl Giuliano Pochini
2004-02-12 10:15             ` John Bradford
2004-02-12 10:27             ` Nick Piggin
2004-02-12 17:05               ` Michael Frank
2004-02-12 17:18                 ` Valdis.Kletnieks
2004-02-12 20:55                   ` Helge Hafting
2004-02-13  1:57                     ` Jamie Lokier
2004-02-13  2:05                       ` Nick Piggin
2004-02-12 14:59             ` Andrea Arcangeli
2004-02-13 12:15     ` ext2/3 performance regression in 2.6 vs 2.4 for small interleaved writes Jon Burgess
2004-02-12 10:40   ` Jon Burgess
2004-02-12 20:17     ` Hans Reiser
2004-02-12  9:56 ` Andrew Morton
2004-02-12 20:20   ` Jon Burgess
2004-02-13  8:28     ` Juan Piernas Canovas
2004-02-16 17:51     ` Alex Zarochentsev
2004-02-16 20:03       ` Jon Burgess
2004-02-13 12:35   ` Jon Burgess
2004-02-14 15:00   ` Jon Burgess [this message]

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=402E380B.8070806@jburgess.uklinux.net \
    --to=lkml@jburgess.uklinux.net \
    --cc=akpm@osdl.org \
    --cc=linux-kernel@vger.kernel.org \
    /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