linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stan Hoeppner <stan@hardwarefreak.com>
To: Pol Hallen <raid1@fuckaround.org>
Cc: linux-raid@vger.kernel.org
Subject: Re: standard performance (write speed 20Mb/s)
Date: Wed, 20 Jul 2011 07:31:35 -0500	[thread overview]
Message-ID: <4E26CAA7.8030201@hardwarefreak.com> (raw)
In-Reply-To: <201107181403.49563.raid1@fuckaround.org>

On 7/18/2011 7:03 AM, Pol Hallen wrote:
>> Then you have multiple problems causing poor performance.  Now is the
>> time for you to post your  hardware configuration, drive model
>> number(s), mdadm config, and filesystem.  You should have done all of
>> this in your initial post.
> 
> hello and sorry for the late :-(
> 
> now is a resync in progress, I can't stop it. Try /usr/share/mdadm/checkarray 
> -x md0 but nothing..
> 
> fs is ext3 (done less 10 days ago)
> 
>> You've been on this list long enough to have known of the WD Green
>> sector alignment issue and avoided it, and to know better than to run
>> performance tests while an array is still in its initial sync.  Thus I
>> initially assumed your problem was merely the insane dd block size.
> 
> when I wrote a post there was not resync in progress..
> 
> tell me if you need known other things, thanks

I can't see how.  You dumped the entire system.  Which is probably why
no one has responded until now.  Never provide all this garbage.  It
takes too long to sift through and grab the relevant bits, which is only
this:


> Device Model:     ST31500341AS
> User Capacity:    1,500,301,910,016 bytes

> Device Model:     ST31500341AS
> User Capacity:    1,500,301,910,016 bytes

> Device Model:     ST3500418AS
> User Capacity:    500,107,862,016 bytes

> Device Model:     WDC WD20EARS-00MVWB0
> User Capacity:    2,000,398,934,016 bytes

> Device Model:     WDC WD20EARS-00MVWB0
> User Capacity:    2,000,398,934,016 bytes

This mishmash combo of drives, including two WD Green Advanced Format
drives, is likely the cause of the horrible write performance.  You're
assembling the array from partitions.  The partitions on the WD20EARS
drives are undoubtedly not 4KB sector aligned, causing excessive
read-modify-write cycles.  See:

http://en.wikipedia.org/wiki/Advanced_Format

Repartition the Green drives with proper alignment.  If you still have
issues we'll go from there.  Note:  Having 3 dissimilar drives in an
array is never going to be an optimally performing solution.

-- 
Stan

  reply	other threads:[~2011-07-20 12:31 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-07-18 12:03 standard performance (write speed 20Mb/s) Pol Hallen
2011-07-20 12:31 ` Stan Hoeppner [this message]
2011-07-20 20:02   ` Pol Hallen
     [not found]     ` <CADNH=7HR7euaWem0tpLxJfRe0hYnRP5fwxJ6MC5vJeNf=T=PzA@mail.gmail.com>
2011-07-21  0:02       ` Stan Hoeppner
2011-07-21  6:57         ` Mathias Burén
2011-07-21  8:55     ` Erwan Leroux
2011-07-21 15:06       ` Paweł Brodacki
2011-07-21 15:30         ` Erwan Leroux
2011-07-21 15:42           ` Erwan Leroux
2011-07-25  6:39             ` Paweł Brodacki
2011-07-21 17:57       ` standard performance (write speed 65Mb/s) Pol Hallen
2011-07-22  0:24         ` Stan Hoeppner
2011-07-22 11:34           ` Pol Hallen
  -- strict thread matches above, loose matches on Subject: below --
2011-07-16 19:40 standard performance (write speed 20Mb/s) Pol Hallen
2011-07-17  4:26 ` Stan Hoeppner
2011-07-17  8:12   ` Pol Hallen
2011-07-17 12:11     ` John Robinson
2011-07-17 12:22       ` Iustin Pop
2011-07-17 12:51         ` John Robinson
2011-07-17 13:28           ` Iustin Pop
2011-07-18  9:04             ` John Robinson
2011-07-17 22:05         ` Stan Hoeppner
2011-07-27  5:42       ` Simon Matthews
2011-07-27  5:46         ` Roman Mamedov
2011-07-27 10:22         ` Stan Hoeppner
2011-07-27 10:26           ` John Robinson
2011-07-27 12:35             ` Joe Landman
2011-07-27 13:54             ` Stan Hoeppner
2011-07-17 23:03     ` Stan Hoeppner
2011-07-18 11:52       ` Pol Hallen
2011-07-17 16:48 ` Gordon Henderson

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=4E26CAA7.8030201@hardwarefreak.com \
    --to=stan@hardwarefreak.com \
    --cc=linux-raid@vger.kernel.org \
    --cc=raid1@fuckaround.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;
as well as URLs for NNTP newsgroup(s).