From: Christoph Hellwig <hch@infradead.org>
To: Dominik Brodowski <linux@dominikbrodowski.net>,
Michael Monnerie <michael.monnerie@is.it-management.at>,
Christoph Hellwig <hch@infradead.org>,
linux-raid@vger.kernel.org, xfs@oss.sgi
Subject: Re: How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
Date: Wed, 4 Aug 2010 07:18:03 -0400 [thread overview]
Message-ID: <20100804111803.GA32643@infradead.org> (raw)
In-Reply-To: <20100804102526.GB13766@isilmar-3.linta.de>
On Wed, Aug 04, 2010 at 12:25:26PM +0200, Dominik Brodowski wrote:
> Hey,
>
> many thanks for your feedback. It seems the crypto step is the culprit:
>
> Reading 1.1 GB with dd, iflag=direct, bs=8k:
>
> /dev/sd* 35.3 MB/s ( 90 %)
> /dev/md* 39.1 MB/s (100 %)
> /dev/mapper/md*_crypt 3.9 MB/s ( 10 %)
> /dev/mapper/vg1-* 3.9 MB/s ( 10 %)
>
> The "good" news: it also happens on my notebook, even though it has a
> different setup (no raid, disk -> lv/vg -> crypt). On my notebook, I'm
> more than happy to test out different kernel versions, patches etc.
>
> /dev/sd* 17.7 MB/s (100 %)
> /dev/mapper/vg1-* 16.2 MB/s ( 92 %)
> /dev/mapper/*_crypt 3.1 MB/s ( 18 %)
The good news is that you have it tracked down, the bad news is that
I know very little about dm-crypt. Maybe the issue is the single
threaded decryption in dm-crypt? Can you check how much CPU time
the dm crypt kernel thread uses?
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org>
To: Dominik Brodowski <linux@dominikbrodowski.net>,
Michael Monnerie <michael.monnerie@is.it-management.at>,
Christoph Hellwig <hch@infradead.org>,
linux-raid@vger.kernel.org, xfs@oss.sgi.com,
linux-kernel@vger.kernel.org, dm-devel@redhat.com
Subject: Re: How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
Date: Wed, 4 Aug 2010 07:18:03 -0400 [thread overview]
Message-ID: <20100804111803.GA32643@infradead.org> (raw)
In-Reply-To: <20100804102526.GB13766@isilmar-3.linta.de>
On Wed, Aug 04, 2010 at 12:25:26PM +0200, Dominik Brodowski wrote:
> Hey,
>
> many thanks for your feedback. It seems the crypto step is the culprit:
>
> Reading 1.1 GB with dd, iflag=direct, bs=8k:
>
> /dev/sd* 35.3 MB/s ( 90 %)
> /dev/md* 39.1 MB/s (100 %)
> /dev/mapper/md*_crypt 3.9 MB/s ( 10 %)
> /dev/mapper/vg1-* 3.9 MB/s ( 10 %)
>
> The "good" news: it also happens on my notebook, even though it has a
> different setup (no raid, disk -> lv/vg -> crypt). On my notebook, I'm
> more than happy to test out different kernel versions, patches etc.
>
> /dev/sd* 17.7 MB/s (100 %)
> /dev/mapper/vg1-* 16.2 MB/s ( 92 %)
> /dev/mapper/*_crypt 3.1 MB/s ( 18 %)
The good news is that you have it tracked down, the bad news is that
I know very little about dm-crypt. Maybe the issue is the single
threaded decryption in dm-crypt? Can you check how much CPU time
the dm crypt kernel thread uses?
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org>
To: Dominik Brodowski <linux@dominikbrodowski.net>,
Michael Monnerie <michael.monnerie@is.it-management.at>,
Christoph Hellwig <hch@infradead.org>,
linux-raid@vger.kernel.org, xfs@oss.sgi.com,
linux-kernel@vger.kernel.org, dm-devel@redhat.com
Subject: Re: How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs
Date: Wed, 4 Aug 2010 07:18:03 -0400 [thread overview]
Message-ID: <20100804111803.GA32643@infradead.org> (raw)
In-Reply-To: <20100804102526.GB13766@isilmar-3.linta.de>
On Wed, Aug 04, 2010 at 12:25:26PM +0200, Dominik Brodowski wrote:
> Hey,
>
> many thanks for your feedback. It seems the crypto step is the culprit:
>
> Reading 1.1 GB with dd, iflag=direct, bs=8k:
>
> /dev/sd* 35.3 MB/s ( 90 %)
> /dev/md* 39.1 MB/s (100 %)
> /dev/mapper/md*_crypt 3.9 MB/s ( 10 %)
> /dev/mapper/vg1-* 3.9 MB/s ( 10 %)
>
> The "good" news: it also happens on my notebook, even though it has a
> different setup (no raid, disk -> lv/vg -> crypt). On my notebook, I'm
> more than happy to test out different kernel versions, patches etc.
>
> /dev/sd* 17.7 MB/s (100 %)
> /dev/mapper/vg1-* 16.2 MB/s ( 92 %)
> /dev/mapper/*_crypt 3.1 MB/s ( 18 %)
The good news is that you have it tracked down, the bad news is that
I know very little about dm-crypt. Maybe the issue is the single
threaded decryption in dm-crypt? Can you check how much CPU time
the dm crypt kernel thread uses?
next prev parent reply other threads:[~2010-08-04 11:18 UTC|newest]
Thread overview: 63+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-04 7:35 How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs Dominik Brodowski
2010-08-04 7:35 ` Dominik Brodowski
2010-08-04 8:50 ` Christoph Hellwig
2010-08-04 8:50 ` Christoph Hellwig
2010-08-04 9:13 ` Dominik Brodowski
2010-08-04 9:13 ` Dominik Brodowski
2010-08-04 9:21 ` Christoph Hellwig
2010-08-04 9:21 ` Christoph Hellwig
2010-08-04 9:16 ` Michael Monnerie
2010-08-04 9:16 ` Michael Monnerie
2010-08-04 10:25 ` Dominik Brodowski
2010-08-04 10:25 ` Dominik Brodowski
2010-08-04 11:18 ` Christoph Hellwig [this message]
2010-08-04 11:18 ` Christoph Hellwig
2010-08-04 11:18 ` Christoph Hellwig
2010-08-04 11:24 ` Dominik Brodowski
2010-08-04 11:24 ` Dominik Brodowski
2010-08-04 11:53 ` Mikael Abrahamsson
2010-08-04 11:53 ` Mikael Abrahamsson
2010-08-04 12:56 ` Mike Snitzer
2010-08-04 12:56 ` Mike Snitzer
2010-08-04 22:24 ` Neil Brown
2010-08-04 22:24 ` Neil Brown
2010-08-05 8:33 ` Stan Hoeppner
2010-08-07 10:13 ` Dave Chinner
2010-08-07 23:43 ` Stan Hoeppner
2010-08-08 7:46 ` Michael Monnerie
2010-08-04 11:18 ` Christoph Hellwig
[not found] ` <20100804111803.GA32643__39273.3621680692$1280923964$gmane$org@infradead.org>
2010-08-04 17:43 ` Andi Kleen
2010-08-04 20:33 ` Valdis.Kletnieks
2010-08-04 20:33 ` Valdis.Kletnieks
2010-08-04 20:33 ` Valdis.Kletnieks
2010-08-05 9:31 ` direct-io regression [Was: How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs] Dominik Brodowski
2010-08-05 9:31 ` Dominik Brodowski
2010-08-05 11:32 ` Chris Mason
2010-08-05 11:32 ` Chris Mason
2010-08-05 11:32 ` Chris Mason
2010-08-05 11:32 ` Chris Mason
2010-08-05 12:36 ` Josef Bacik
2010-08-05 12:36 ` Josef Bacik
2010-08-05 12:36 ` Josef Bacik
2010-08-05 12:36 ` Josef Bacik
2010-08-05 15:35 ` Dominik Brodowski
2010-08-05 15:35 ` Dominik Brodowski
2010-08-05 15:39 ` Chris Mason
2010-08-05 15:39 ` Chris Mason
2010-08-05 15:39 ` Chris Mason
2010-08-05 15:39 ` Chris Mason
2010-08-05 15:53 ` Dominik Brodowski
2010-08-05 15:53 ` Dominik Brodowski
2010-08-05 15:53 ` Dominik Brodowski
2010-08-05 16:35 ` Dominik Brodowski
2010-08-05 16:35 ` Dominik Brodowski
2010-08-05 16:35 ` Dominik Brodowski
2010-08-05 20:47 ` Performance impact of CONFIG_DEBUG? direct-io test case Dominik Brodowski
2010-08-05 20:47 ` Dominik Brodowski
2010-08-05 20:54 ` Performance impact of CONFIG_SCHED_MC? " Dominik Brodowski
2010-08-05 20:54 ` Dominik Brodowski
2010-08-05 16:35 ` direct-io regression [Was: How to track down abysmal performance ata - raid1 - crypto - vg/lv - xfs] Dominik Brodowski
2010-08-05 18:58 ` Jeff Moyer
2010-08-05 18:58 ` Jeff Moyer
2010-08-05 19:01 ` Chris Mason
2010-08-05 19:01 ` Chris Mason
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=20100804111803.GA32643@infradead.org \
--to=hch@infradead.org \
--cc=linux-raid@vger.kernel.org \
--cc=linux@dominikbrodowski.net \
--cc=michael.monnerie@is.it-management.at \
--cc=xfs@oss.sgi \
/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.