All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mike Snitzer <snitzer@redhat.com>
To: Spelic <spelic@shiftmail.org>
Cc: "Lukáš Czerner" <lczerner@redhat.com>,
	"device-mapper development" <dm-devel@redhat.com>,
	linux-ext4@vger.kernel.org, "Dave Chinner" <david@fromorbit.com>,
	xfs@oss.sgi.com
Subject: Re: Ext4 and xfs problems in dm-thin on allocation and discard
Date: Tue, 19 Jun 2012 14:48:59 -0400	[thread overview]
Message-ID: <20120619184858.GA8841@redhat.com> (raw)
In-Reply-To: <20120619144413.GA7225@redhat.com>

On Tue, Jun 19 2012 at 10:44am -0400,
Mike Snitzer <snitzer@redhat.com> wrote:

> On Tue, Jun 19 2012 at  9:52am -0400,
> Spelic <spelic@shiftmail.org> wrote:
>
> > I do not know what is the mechanism for which xfs cannot unmap
> > blocks from dm-thin, but it really can't.
> > If anyone has dm-thin installed he can try. This is 100%
> > reproducible for me.
> 
> I was initially surprised by this considering the thinp-test-suite does
> test a compilebench workload against xfs and ext4 using online discard
> (-o discard).
> 
> But I just modified that test to use a thin-pool with 'ignore_discard'
> and the test still passed on both ext4 and xfs.
> 
> So there is more work needed in the thinp-test-suite to use blktrace
> hooks to verify that discards are occuring when the compilebench
> generated files are removed.
> 
> I'll work through that and report back.

blktrace shows discards for both xfs and ext4.

But in general xfs is issuing discards with much smaller extents than
ext4 does, e.g.:

to the thin device:
+ 128 vs + 32

to the thin-pool's data device:
+ 120 vs + 16

WARNING: multiple messages have this Message-ID (diff)
From: Mike Snitzer <snitzer@redhat.com>
To: Spelic <spelic@shiftmail.org>
Cc: "Lukáš Czerner" <lczerner@redhat.com>,
	"device-mapper development" <dm-devel@redhat.com>,
	linux-ext4@vger.kernel.org, xfs@oss.sgi.com
Subject: Re: Ext4 and xfs problems in dm-thin on allocation and discard
Date: Tue, 19 Jun 2012 14:48:59 -0400	[thread overview]
Message-ID: <20120619184858.GA8841@redhat.com> (raw)
In-Reply-To: <20120619144413.GA7225@redhat.com>

On Tue, Jun 19 2012 at 10:44am -0400,
Mike Snitzer <snitzer@redhat.com> wrote:

> On Tue, Jun 19 2012 at  9:52am -0400,
> Spelic <spelic@shiftmail.org> wrote:
>
> > I do not know what is the mechanism for which xfs cannot unmap
> > blocks from dm-thin, but it really can't.
> > If anyone has dm-thin installed he can try. This is 100%
> > reproducible for me.
> 
> I was initially surprised by this considering the thinp-test-suite does
> test a compilebench workload against xfs and ext4 using online discard
> (-o discard).
> 
> But I just modified that test to use a thin-pool with 'ignore_discard'
> and the test still passed on both ext4 and xfs.
> 
> So there is more work needed in the thinp-test-suite to use blktrace
> hooks to verify that discards are occuring when the compilebench
> generated files are removed.
> 
> I'll work through that and report back.

blktrace shows discards for both xfs and ext4.

But in general xfs is issuing discards with much smaller extents than
ext4 does, e.g.:

to the thin device:
+ 128 vs + 32

to the thin-pool's data device:
+ 120 vs + 16

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

  reply	other threads:[~2012-06-19 18:49 UTC|newest]

Thread overview: 72+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-06-18 21:33 Ext4 and xfs problems in dm-thin on allocation and discard Spelic
2012-06-18 21:33 ` Spelic
2012-06-19  1:57 ` Dave Chinner
2012-06-19  1:57   ` Dave Chinner
2012-06-19  3:12   ` Mike Snitzer
2012-06-19  3:12     ` Mike Snitzer
2012-06-19  6:32     ` Lukáš Czerner
2012-06-19  6:32       ` Lukáš Czerner
2012-06-19 11:29       ` Spelic
2012-06-19 11:29         ` Spelic
2012-06-19 12:20         ` Lukáš Czerner
2012-06-19 12:20           ` Lukáš Czerner
2012-06-19 13:34         ` Mike Snitzer
2012-06-19 13:34           ` Mike Snitzer
2012-06-19 13:16       ` Mike Snitzer
2012-06-19 13:16         ` Mike Snitzer
2012-06-19 13:25         ` Lukáš Czerner
2012-06-19 13:25           ` Lukáš Czerner
2012-06-19 13:30           ` Mike Snitzer
2012-06-19 13:30             ` Mike Snitzer
2012-06-19 13:52             ` Spelic
2012-06-19 13:52               ` Spelic
2012-06-19 14:05               ` Eric Sandeen
2012-06-19 14:05                 ` Eric Sandeen
2012-06-19 14:44               ` Mike Snitzer
2012-06-19 14:44                 ` Mike Snitzer
2012-06-19 18:48                 ` Mike Snitzer [this message]
2012-06-19 18:48                   ` Mike Snitzer
2012-06-19 20:06                   ` Dave Chinner
2012-06-19 20:06                     ` Dave Chinner
2012-06-19 20:21                     ` Ted Ts'o
2012-06-19 20:21                       ` Ted Ts'o
2012-06-19 20:39                       ` Dave Chinner
2012-06-19 20:39                         ` Dave Chinner
2012-06-20  9:01                         ` Christoph Hellwig
2012-06-20  9:01                           ` Christoph Hellwig
2012-06-19 21:37                     ` Spelic
2012-06-19 21:37                       ` Spelic
2012-06-19 23:12                       ` Dave Chinner
2012-06-19 23:12                         ` Dave Chinner
2012-06-20 12:11   ` Spelic
2012-06-20 12:11     ` Spelic
2012-06-20 22:53     ` Dave Chinner
2012-06-20 22:53       ` Dave Chinner
2012-06-21 17:47       ` Mike Snitzer
2012-06-21 17:47         ` Mike Snitzer
2012-06-21 23:29         ` Dave Chinner
2012-06-21 23:29           ` Dave Chinner
2012-07-01 14:53         ` Paolo Bonzini
2012-07-01 14:53           ` Paolo Bonzini
2012-07-02 13:00           ` Mike Snitzer
2012-07-02 13:00             ` Mike Snitzer
2012-07-02 13:15             ` Paolo Bonzini
2012-07-02 13:15               ` Paolo Bonzini
2012-06-19 14:09 ` Lukáš Czerner
2012-06-19 14:09   ` Lukáš Czerner
2012-06-19 14:19   ` Ted Ts'o
2012-06-19 14:19     ` Ted Ts'o
2012-06-19 14:23     ` Eric Sandeen
2012-06-19 14:23       ` Eric Sandeen
2012-06-19 14:37     ` Lukáš Czerner
2012-06-19 14:37       ` Lukáš Czerner
2012-06-19 14:43     ` [dm-devel] " Alasdair G Kergon
2012-06-19 14:43       ` Alasdair G Kergon
2012-06-19 15:28       ` Mike Snitzer
2012-06-19 15:28         ` Mike Snitzer
2012-06-19 16:03         ` [dm-devel] " Alasdair G Kergon
2012-06-19 16:03           ` Alasdair G Kergon
2012-06-19 19:58         ` Ted Ts'o
2012-06-19 19:58           ` Ted Ts'o
2012-06-19 20:44           ` Mike Snitzer
2012-06-19 20:44             ` Mike Snitzer

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=20120619184858.GA8841@redhat.com \
    --to=snitzer@redhat.com \
    --cc=david@fromorbit.com \
    --cc=dm-devel@redhat.com \
    --cc=lczerner@redhat.com \
    --cc=linux-ext4@vger.kernel.org \
    --cc=spelic@shiftmail.org \
    --cc=xfs@oss.sgi.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 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.