linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Brendan Conoboy <blc@redhat.com>
To: Andrew Clayton <andrew@pccl.info>
Cc: linux-raid@vger.kernel.org
Subject: Re: RAID 5 performance issue.
Date: Fri, 05 Oct 2007 14:25:29 -0600	[thread overview]
Message-ID: <47069DB9.2050608@redhat.com> (raw)
In-Reply-To: <20071003105321.06943824@zeus.pccl.info>

Andrew Clayton wrote:
> If anyone has any idea's I'm all ears.

Hi Andrew,

Are you sure your drives are healthy?  Try benchmarking each drive 
individually and see if there is a dramatic performance difference 
between any of them.  One failing drive can slow down an entire array. 
  Only after you have determined that your drives are healthy when 
accessed individually are combined results particularly meaningful.  For 
a generic SATA 1 drive you should expect a sustained raw read or write 
in excess of 45 MB/s.  Check both read and write (this will destroy 
data) and make sure your cache is clear prior to the read test and after 
the write test.  If each drive is working at a reasonable rate 
individually, you're ready to move on.

The next question is: What happens when you access more than one device 
at the same time?  You should either get nearly full combined 
performance, max out CPU, or get throttled by bus bandwidth (An actual 
kernel bug could also come into play here, but I tend to doubt it).  Is 
the onboard SATA controller real SATA or just an ATA-SATA converter?  If 
the latter, you're going to have trouble getting faster performance than 
any one disk can give you at a time.  The output of 'lspci' should tell 
you if the onboard SATA controller is on its own bus or sharing space 
with some other device.  Pasting the output here would be useful.

Assuming you get good performance out of all 3 drives at the same time, 
it's time to create a RAID 5 md device with the three, make sure your 
parity is done building, then benchmark that.  It's going to be slower 
to write and a bit slower to read (especially if your CPU is maxed out), 
but that is normal.

Assuming you get good performance out of your md device, it's time to 
put your filesystem on the md device and benchmark that.  If you use 
ext3, remember to set the stride parameter per the raid howto.  I am 
unfamiliar with other fs/md interactions, so be sure to check.

If you're actually maxing out your bus bandwidth and the onboard sata 
controller is on a different bus than the pci sata controller, try 
balancing the drives between the two to get a larger combined pipe.

Good luck,

-- 
Brendan Conoboy / Red Hat, Inc. / blc@redhat.com

  parent reply	other threads:[~2007-10-05 20:25 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-10-03  9:53 RAID 5 performance issue Andrew Clayton
2007-10-03 16:43 ` Justin Piszcz
2007-10-03 16:48   ` Justin Piszcz
2007-10-03 20:10     ` Andrew Clayton
2007-10-03 20:16       ` Justin Piszcz
2007-10-06 12:30       ` Justin Piszcz
2007-10-06 16:05         ` Justin Piszcz
2007-10-03 20:19   ` Andrew Clayton
2007-10-03 20:35     ` Justin Piszcz
2007-10-03 20:46       ` Andrew Clayton
2007-10-03 20:36     ` David Rees
2007-10-03 20:48       ` Andrew Clayton
2007-10-04 14:08       ` Andrew Clayton
2007-10-04 14:09         ` Justin Piszcz
2007-10-04 14:10           ` Justin Piszcz
2007-10-04 14:44             ` Andrew Clayton
2007-10-04 16:20               ` Justin Piszcz
2007-10-04 18:26                 ` Andrew Clayton
2007-10-05 10:25                   ` Justin Piszcz
2007-10-05 10:57                     ` Andrew Clayton
2007-10-05 11:08                       ` Justin Piszcz
2007-10-05 12:53                         ` Andrew Clayton
2007-10-05 13:18                           ` Justin Piszcz
2007-10-05 13:30                           ` Andrew Clayton
2007-10-05 14:07                             ` Justin Piszcz
2007-10-05 14:32                               ` Andrew Clayton
2007-10-05 16:10                               ` Andrew Clayton
2007-10-05 16:16                                 ` Justin Piszcz
2007-10-05 19:33                                   ` Andrew Clayton
2007-10-05 18:58                                 ` Richard Scobie
2007-10-05 19:02                                   ` Justin Piszcz
2007-10-05 19:02               ` John Stoffel
2007-10-05 19:42                 ` Andrew Clayton
2007-10-05 20:56                   ` John Stoffel
2007-10-07 17:22                     ` Andrew Clayton
2007-10-11 17:06                       ` Bill Davidsen
2007-10-11 18:07                         ` Andrew Clayton
2007-10-11 23:43                           ` Justin Piszcz
2007-10-04 14:36           ` Andrew Clayton
2007-10-04 14:39             ` Justin Piszcz
2007-10-04 15:03               ` Andrew Clayton
2007-10-04 16:19                 ` Justin Piszcz
2007-10-04 19:01                   ` Andrew Clayton
2007-10-04 16:46                 ` Steve Cousins
2007-10-04 17:06                   ` Steve Cousins
2007-10-04 19:06                   ` Andrew Clayton
2007-10-05 10:20                     ` Justin Piszcz
2007-10-04 14:39             ` Justin Piszcz
2007-10-03 17:53 ` Goswin von Brederlow
2007-10-03 20:20   ` Andrew Clayton
2007-10-03 20:48     ` Richard Scobie
2007-10-05 20:25 ` Brendan Conoboy [this message]
2007-10-06  0:38   ` Dean S. Messing
2007-10-06  8:18     ` Justin Piszcz
2007-10-08  1:40       ` Dean S. Messing
2007-10-08  8:44         ` Justin Piszcz

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=47069DB9.2050608@redhat.com \
    --to=blc@redhat.com \
    --cc=andrew@pccl.info \
    --cc=linux-raid@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;
as well as URLs for NNTP newsgroup(s).