From: Jens Axboe <jens.axboe@oracle.com>
To: Matthew Wilcox <matthew@wil.cx>
Cc: David Woodhouse <dwmw2@infradead.org>,
Dave Chinner <david@fromorbit.com>,
Black_David@emc.com, martin.petersen@oracle.com,
chris.mason@oracle.com, James.Bottomley@hansenpartnership.com,
rwheeler@redhat.com, linux-scsi@vger.kernel.org,
linux-fsdevel@vger.kernel.org, coughlan@redhat.com
Subject: Re: Thin provisioning & arrays
Date: Mon, 10 Nov 2008 14:36:54 +0100 [thread overview]
Message-ID: <20081110133653.GH26778@kernel.dk> (raw)
In-Reply-To: <20081110133010.GN15439@parisc-linux.org>
On Mon, Nov 10 2008, Matthew Wilcox wrote:
> On Mon, Nov 10, 2008 at 10:59:49AM +0100, David Woodhouse wrote:
> > Storage devices are complex enough that they _already_ exhibit behaviour
> > which is fairly much non-deterministic in a number of ways. Especially
> > if we're talking about SSDs or large arrays, rather than just disks.
>
> If anything, SSDs are more deterministic than rotating storage.
> Variable numbers of sectors per track, unpredictable sector remapping,
> track re-reads due to errors during reads ... SSDs seem like a real
> improvement.
This was also discussed at LPC, and I would tend to disagree. On
rotation storage, slight skews here and there are seen, but rarely
anything major. On SSD's, when the GC kicks in it can be fairly
detrimental to performance. And one run doesn't necessarily compare to
the next run, even if you start with the same seed.
--
Jens Axboe
next prev parent reply other threads:[~2008-11-10 13:36 UTC|newest]
Thread overview: 105+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-11-06 14:43 thin provisioned LUN support Ric Wheeler
2008-11-06 15:17 ` James Bottomley
2008-11-06 15:24 ` David Woodhouse
2008-11-06 16:00 ` Ric Wheeler
2008-11-06 16:40 ` Martin K. Petersen
2008-11-06 17:04 ` Ric Wheeler
2008-11-06 17:15 ` Matthew Wilcox
2008-11-07 12:05 ` Jens Axboe
2008-11-07 12:14 ` Ric Wheeler
2008-11-07 12:17 ` David Woodhouse
2008-11-07 12:19 ` Jens Axboe
2008-11-07 14:26 ` thin provisioned LUN support & file system allocation policy Ric Wheeler
2008-11-07 14:34 ` Matthew Wilcox
2008-11-07 14:45 ` Jörn Engel
2008-11-07 14:43 ` Theodore Tso
2008-11-07 14:54 ` Ric Wheeler
2008-11-07 15:26 ` jim owens
2008-11-07 15:31 ` David Woodhouse
2008-11-07 15:35 ` jim owens
2008-11-07 15:46 ` Theodore Tso
2008-11-07 15:51 ` Martin K. Petersen
2008-11-07 16:06 ` Ric Wheeler
2008-11-07 15:56 ` James Bottomley
2008-11-07 15:36 ` James Bottomley
2008-11-07 15:48 ` David Woodhouse
2008-11-07 15:36 ` Theodore Tso
2008-11-07 15:45 ` Matthew Wilcox
2008-11-07 16:07 ` jim owens
2008-11-07 16:12 ` James Bottomley
2008-11-07 16:23 ` jim owens
2008-11-07 16:02 ` Ric Wheeler
2008-11-07 14:55 ` Matthew Wilcox
2008-11-07 15:20 ` thin provisioned LUN support James Bottomley
2008-11-09 23:08 ` Dave Chinner
2008-11-09 23:37 ` James Bottomley
2008-11-10 0:33 ` Dave Chinner
2008-11-10 14:31 ` James Bottomley
2008-11-07 15:49 ` Chris Mason
2008-11-07 16:00 ` Martin K. Petersen
2008-11-07 16:06 ` James Bottomley
2008-11-07 16:11 ` Chris Mason
2008-11-07 16:18 ` James Bottomley
2008-11-07 16:22 ` Ric Wheeler
2008-11-07 16:27 ` James Bottomley
2008-11-07 16:28 ` David Woodhouse
2008-11-07 17:22 ` Chris Mason
2008-11-07 18:09 ` Ric Wheeler
2008-11-07 18:36 ` Theodore Tso
2008-11-07 18:41 ` Ric Wheeler
[not found] ` <49148BDF.9050707@redhat.com>
2008-11-07 19:35 ` Theodore Tso
2008-11-07 19:55 ` Martin K. Petersen
2008-11-07 20:19 ` Theodore Tso
2008-11-07 20:21 ` Matthew Wilcox
[not found] ` <20081107202149.GJ15439@parisc-linux.org>
2008-11-07 20:26 ` Ric Wheeler
2008-11-07 20:48 ` Chris Mason
2008-11-07 21:04 ` Ric Wheeler
2008-11-07 21:13 ` Theodore Tso
2008-11-07 20:42 ` Theodore Tso
2008-11-07 21:06 ` Martin K. Petersen
2008-11-07 20:37 ` Ric Wheeler
2008-11-10 2:44 ` Black_David
2008-11-10 2:36 ` Black_David
2008-11-07 19:44 ` jim owens
2008-11-07 19:48 ` Matthew Wilcox
2008-11-07 19:50 ` Ric Wheeler
2008-11-09 23:36 ` Dave Chinner
2008-11-10 3:40 ` Thin provisioning & arrays Black_David
2008-11-10 8:31 ` Dave Chinner
2008-11-10 9:59 ` David Woodhouse
2008-11-10 13:30 ` Matthew Wilcox
2008-11-10 13:36 ` Jens Axboe [this message]
2008-11-10 17:05 ` UNMAP is a hint Black_David
2008-11-10 17:30 ` Matthew Wilcox
2008-11-10 17:56 ` Ric Wheeler
2008-11-10 22:18 ` Thin provisioning & arrays Dave Chinner
2008-11-11 1:23 ` Black_David
2008-11-11 2:09 ` Keith Owens
2008-11-11 13:59 ` Ric Wheeler
2008-11-11 14:55 ` jim owens
2008-11-11 15:38 ` Ric Wheeler
2008-11-11 15:59 ` jim owens
2008-11-11 16:25 ` Ric Wheeler
2008-11-11 16:53 ` jim owens
2008-11-11 23:08 ` Dave Chinner
2008-11-11 23:52 ` jim owens
2008-11-11 22:49 ` Dave Chinner
2008-11-06 15:27 ` thin provisioned LUN support jim owens
2008-11-06 15:57 ` jim owens
2008-11-06 16:21 ` James Bottomley
[not found] ` <yq1d4h8nao5.fsf@sermon.lab.mkp.net>
2008-11-06 15:42 ` Ric Wheeler
2008-11-06 15:57 ` David Woodhouse
2008-11-06 22:36 ` Dave Chinner
2008-11-06 22:55 ` Ric Wheeler
[not found] ` <491375E9.7020707@redhat.com>
2008-11-06 23:06 ` James Bottomley
2008-11-06 23:10 ` Ric Wheeler
2008-11-06 23:26 ` James Bottomley
2008-11-06 23:32 ` thin provisioned LUN support - T10 activity Black_David
2008-11-07 11:59 ` thin provisioned LUN support Artem Bityutskiy
2008-11-10 20:39 ` Aggregating discard requests in the filesystem Matthew Wilcox
2008-11-10 20:44 ` Chris Mason
2008-11-11 0:12 ` Brad Boyer
2008-11-11 15:25 ` jim owens
2008-11-11 16:40 ` thin provisioned LUN support Christoph Hellwig
2008-11-11 17:07 ` jim owens
2008-11-11 17:33 ` Christoph Hellwig
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=20081110133653.GH26778@kernel.dk \
--to=jens.axboe@oracle.com \
--cc=Black_David@emc.com \
--cc=James.Bottomley@hansenpartnership.com \
--cc=chris.mason@oracle.com \
--cc=coughlan@redhat.com \
--cc=david@fromorbit.com \
--cc=dwmw2@infradead.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=matthew@wil.cx \
--cc=rwheeler@redhat.com \
/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).