All of lore.kernel.org
 help / color / mirror / Atom feed
From: Lutz Vieweg <lvml@5t9.de>
To: linux-raid@vger.kernel.org
Subject: Re: Software RAID and TRIM
Date: Wed, 20 Jul 2011 14:20:08 +0200	[thread overview]
Message-ID: <j06h5o$ebm$1@dough.gmane.org> (raw)
In-Reply-To: <j0613g$aoe$1@dough.gmane.org>

On 07/20/2011 09:42 AM, David Brown wrote:
>> http://tdm27.wordpress.com/2011/07/19/some-solid-state-drive-benchmarks/
>
> The telling figure here, though, is that TRIM made the single drive significantly slower.

More precisely, online-TRIM of ext4 on Intel SSDs seems to be a bad combination.

I think it's clear you cannot gain much from TRIM if you're willing to spend
the money for 2 times overprovisioning, anyway. You can lose significantly from
online-trim when the filesystem issues a lot of TRIM commands all the time and
when the SSD is slow to process them.

TRIM gains you an advantage with less over-provisioning, and is better
done in batches after significant amounts of data have been written/deleted.

When you try with different levels of over-provisioning, also try with
batched discards (fstrim) between runs of your benchmark.

Regards,

Lutz Vieweg



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

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-06-28 15:31 Software RAID and TRIM Tom De Mulder
2011-06-28 16:11 ` Mathias Burén
2011-06-29 10:32   ` Tom De Mulder
2011-06-29 10:45     ` NeilBrown
2011-06-29 11:10       ` Tom De Mulder
2011-06-29 11:48         ` Scott E. Armitage
2011-06-29 12:46           ` Roberto Spadim
2011-06-29 12:46       ` David Brown
2011-06-30  0:28         ` NeilBrown
2011-06-30  7:50           ` David Brown
2011-06-29 13:39       ` Namhyung Kim
2011-06-30  0:27         ` NeilBrown
2011-07-17 22:11       ` Lutz Vieweg
2011-07-17 21:57     ` Lutz Vieweg
2011-06-29 10:33   ` Tom De Mulder
2011-06-29 12:42     ` David Brown
2011-06-29 12:55       ` Tom De Mulder
2011-06-29 13:02         ` Roberto Spadim
2011-06-29 13:10         ` David Brown
2011-06-30  5:51         ` Mikael Abrahamsson
2011-07-04  9:13           ` Tom De Mulder
2011-07-04 16:26             ` Werner Fischer
2011-07-17 22:31               ` Lutz Vieweg
2011-07-17 22:16         ` Lutz Vieweg
2011-07-17 22:00     ` Lutz Vieweg
2011-06-28 16:17 ` Johannes Truschnigg
2011-06-28 16:40 ` David Brown
2011-07-17 21:52   ` Lutz Vieweg
2011-07-18  5:14     ` Mikael Abrahamsson
2011-07-18 10:35     ` David Brown
2011-07-18 10:48       ` Tom De Mulder
2011-07-18 18:09       ` Lutz Vieweg
2011-07-18 20:18         ` David Brown
2011-07-19  9:29           ` Lutz Vieweg
2011-07-19 10:22             ` David Brown
2011-07-19 13:41               ` Lutz Vieweg
2011-07-19 15:06                 ` David Brown
2011-07-20 10:39                   ` Lutz Vieweg
2011-07-19 14:19               ` Tom De Mulder
2011-07-20  7:42                 ` David Brown
2011-07-20 12:20                   ` Lutz Vieweg [this message]
2011-07-20 12:13                 ` Werner Fischer
2011-07-20 12:25                   ` Lutz Vieweg
2011-07-18 10:53     ` Tom De Mulder
2011-07-18 12:13       ` Werner Fischer

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='j06h5o$ebm$1@dough.gmane.org' \
    --to=lvml@5t9.de \
    --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 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.