mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Arjan van de Ven <arjan@infradead.org>
To: "Martin A. Fink" <fink@mpe.mpg.de>
Cc: Matthias Schniedermeyer <ms@citd.de>, linux-kernel@vger.kernel.org
Subject: Re: SATA-performance: Linux vs. FreeBSD
Date: Tue, 13 Feb 2007 11:08:10 +0100	[thread overview]
Message-ID: <1171361290.12771.77.camel@laptopd505.fenrus.org> (raw)
In-Reply-To: <200702131025.15964.fink@mpe.mpg.de>

> > 
> The problem is: FreeBSD is fast, but lacks of some special drivers. Linux has 
> all drivers but access to harddisk is unpredictable and thus unreliable!
> What can I do??


there's several tunables you can do;
1) increase /sys/block/<device>/queue/nr_requests
   the linux default is on the low side
2) investigate other elevators; cfq is great for interactive use but not
so great for max throughput. you can do this by echo'ing "deadline"
into /sys/block/<device>/scheduler
3) make sure ext3 is set to "data=writeback"; the default journalling
mode is very strict, fine for smallish files but for multi-gigabyte
it'll start to hurt
4) try to use iostat -x /dev/<foo>  1 to see what values avg-rq and
avg-qu are.. avg-rq should be at least several hundred if not more.
5) echo a larger value into /sys/block/<device>/queue/max_sectors_kb
   the default seems to be 512 which is... really low. The hw max is in
   another file in that directory; if you want max throughput set the
   max_sectors_kb value to the hw max. (you pay in terms of fairness for
   this; it's the eternal fairness/latency versus throughput tradeoff)




  reply	other threads:[~2007-02-13 10:08 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-02-12 14:02 Martin A. Fink
2007-02-12 17:04 ` Andi Kleen
2007-02-12 16:27   ` Martin A. Fink
2007-02-12 18:41     ` Andi Kleen
2007-02-12 17:56       ` Martin A. Fink
2007-02-12 18:17         ` Ray Lee
2007-02-12 19:08         ` Alan
2007-02-12 20:34           ` Nigel Cunningham
2007-02-13  9:34           ` Martin A. Fink
2007-02-13 11:25             ` Alan
2007-02-13 12:32               ` Martin A. Fink
2007-02-13 14:47                 ` Theodore Tso
2007-02-13 15:03                   ` Alan
2007-02-13 17:12               ` Jeff Garzik
2007-02-12 23:31         ` Matthias Schniedermeyer
2007-02-13  9:25           ` Martin A. Fink
2007-02-13 10:08             ` Arjan van de Ven [this message]
2007-02-13 11:18               ` Andi Kleen
2007-02-13 10:25                 ` Arjan van de Ven
2007-02-13 11:27               ` Alan
2007-02-13 11:59                 ` Jörn Engel
2007-02-13 19:54               ` Jeffrey Hundstad
2007-02-13 10:16             ` Matthias Schniedermeyer
2007-02-13 10:29               ` Martin A. Fink
2007-02-13 12:04                 ` Jörn Engel
2007-02-13 12:24                 ` Matthias Schniedermeyer
2007-02-13 12:49                   ` Martin A. Fink
2007-02-13 13:53                     ` Matthias Schniedermeyer
2007-02-12 16:37   ` Martin A. Fink
2007-02-12 18:19     ` Stefan Richter
2007-02-13 19:09     ` Jeff Carr
2007-02-12 17:42   ` Martin A. Fink
2007-02-15  5:48 ` Tejun Heo

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=1171361290.12771.77.camel@laptopd505.fenrus.org \
    --to=arjan@infradead.org \
    --cc=fink@mpe.mpg.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=ms@citd.de \
    /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

Powered by JetHome