All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Dragan Krnic" <dkrnic@lycos.com>
To: linux-kernel@vger.kernel.org
Subject: Re: extreme performance degradation - ext3/sata
Date: Thu, 15 Jan 2004 09:33:05 +0100	[thread overview]
Message-ID: <PDNDJHLOKGHIMBAA@mailcity.com> (raw)

Hi,

I've encountered a problem with sata drives and ext3/reiser,
which I documented in the correspondence with Neil Brown
below. He referred me to your forum. Since I'm not subscribed
please cc my return address.

Regards
Dragan

>> I have extreme performance degradation when writing to
>> ext3fs on SATA drives, about 7 times slower than reading
>> from the same volume - 17 MB/s writing, 120 MB/s reading. 
>> 
>> The problem also affects reiserfs in the same way, however
>> xfs appears to be immune to this problem. It reads AND writes
>> to the same volume at about 100 MB/s, which is the same
>> speed at which null-bytes from /dev/zero can be poured
>> into the raw volume.
>> 
>> My setup is: SuSE 8.2 (2.4.20), Siemens Scenic W600 (i865g),
>> P4/3GHz, 2GB PC2100 RAM, a SCSI boot/swap/root disk, 
>> 6 Maxtor 250 GB SATA drives, 2 on on-board connectors,
>> 4 on 2 Promise SATA 150 Tx2+ PCI 32/33 controllers
>> (pdc-ultra.o).
>> 
>> The volumes have been formatted with commands like:
>> 
>>    mke2fs   -J device=/dev/sda2 /dev/md0
>>    mkreiserfs    -j   /dev/sda2 /dev/md0
>>    mkfs.xfs -l logdev=/dev/sda2 /dev/md0
>> 
>> whereby "/dev/md0" may also be "/dev/vg01/lvol1".
>> 
>> The performance degradation, as well as the steady performance
>> of xfs, is totaly independent of the disk clustering mode, 
>> e.g. 5-way RAID5, 6-way RAID5, 3-way RAID0 of 2-way RAID1,
>> 2-way RAID0 (for higher priority on-board controller),
>> 4-way RAID0 (for slightly slower PCI controllers), using
>> either software raid or LVM.
>> 
>> I wish I could give you more info to help you pin-point the
>> performance killer. Please let me know, what more I can do 
>> to help.

                 reply	other threads:[~2004-01-15  8:33 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=PDNDJHLOKGHIMBAA@mailcity.com \
    --to=dkrnic@lycos.com \
    --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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.