linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "J. Bruce Fields" <bfields@fieldses.org>
To: Bryan Schumaker <bjschuma@netapp.com>
Cc: Trond.Myklebust@netapp.com, linux-nfs@vger.kernel.org
Subject: Re: [RFC 0/5] NFS Server Side Copy
Date: Mon, 22 Jul 2013 15:42:09 -0400	[thread overview]
Message-ID: <20130722194209.GE10109@fieldses.org> (raw)
In-Reply-To: <51ED8A47.7040506@netapp.com>

On Mon, Jul 22, 2013 at 03:38:47PM -0400, Bryan Schumaker wrote:
> On 07/22/2013 02:53 PM, J. Bruce Fields wrote:
> > On Fri, Jul 19, 2013 at 05:03:45PM -0400, bjschuma@netapp.com wrote:
> >> From: Bryan Schumaker <bjschuma@netapp.com>
> >>
> >> These patches build on Zach Brown's copyfile patches to add server side
> >> copy to both the NFS client and the NFS server.
> >>
> >> The first patch improves on the copyfile syscall to make it usable on my
> >> machine and also includes notes on other potential problems that I've
> >> found.  The remaining patches first implement a sync copy, then expand to
> >> async.
> >>
> >> My testing was done on a server exporting an ext4 filesystem exporting an
> >> ext4 filesystem.  I compared copying using the cp command to copying with
> >> the copyfile system call.
> > 
> > Were these tests using the full series of patches?  (So, using the
> > asynchronous mechanism?)
> 
> Yes.  Want me to re-run them without it?

I don't think it's urgent.

--b.

> 
> - Bryan
> 
> > 
> > --b.
> > 
> >>
> >>
> >> File size:  512 MB
> >> cp: 4.244 seconds
> >> copyfile: 0.961 seconds
> >>
> >> File size: 1024 MB
> >> cp: 9.091 seconds
> >> copyfile: 1.919 seconds
> >>
> >> File size: 1536 MB
> >> cp: 15.291 seconds
> >> copyfile: 6.016 seconds
> >>
> >>
> >> Repeating these tests on a btrfs exported filesystem supporting the copyfile
> >> system call drops the time for copyfile to about 0.01 seconds.
> >>
> >> Feel free to send me any questions, comments or other thoughts!
> >>
> >> - Bryan
> >>
> >> Bryan Schumaker (5):
> >>   Improve on the copyfile systemcall
> >>   NFSD: Implement the COPY call
> >>   NFS: Add COPY nfs operation
> >>   NFSD: Defer copying
> >>   NFS: Change copy to support async servers
> >>
> >>  fs/copy_range.c          |  10 +++-
> >>  fs/nfs/callback.h        |  13 ++++
> >>  fs/nfs/callback_proc.c   |   9 +++
> >>  fs/nfs/callback_xdr.c    |  54 ++++++++++++++++-
> >>  fs/nfs/inode.c           |   2 +
> >>  fs/nfs/nfs4_fs.h         |   7 +++
> >>  fs/nfs/nfs4file.c        | 101 +++++++++++++++++++++++++++++++
> >>  fs/nfs/nfs4proc.c        |  16 +++++
> >>  fs/nfs/nfs4xdr.c         | 150 +++++++++++++++++++++++++++++++++++++++++++++++
> >>  fs/nfsd/nfs4callback.c   | 136 ++++++++++++++++++++++++++++++++++++++++++
> >>  fs/nfsd/nfs4proc.c       | 104 ++++++++++++++++++++++++++++++--
> >>  fs/nfsd/nfs4state.c      |  15 ++++-
> >>  fs/nfsd/nfs4xdr.c        | 121 +++++++++++++++++++++++++++++++++++++-
> >>  fs/nfsd/state.h          |  23 +++++++-
> >>  fs/nfsd/vfs.c            |   9 +++
> >>  fs/nfsd/vfs.h            |   1 +
> >>  fs/nfsd/xdr4.h           |  24 ++++++++
> >>  fs/nfsd/xdr4cb.h         |   9 +++
> >>  include/linux/nfs4.h     |  14 ++++-
> >>  include/linux/nfs_xdr.h  |  33 +++++++++++
> >>  include/linux/syscalls.h |   1 +
> >>  21 files changed, 836 insertions(+), 16 deletions(-)
> >>
> >> -- 
> >> 1.8.3.3
> >>
> >> --
> >> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> >> the body of a message to majordomo@vger.kernel.org
> >> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

      reply	other threads:[~2013-07-22 19:42 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-19 21:03 [RFC 0/5] NFS Server Side Copy bjschuma
2013-07-19 21:03 ` [RFC 1/5] Improve on the copyfile systemcall bjschuma
2013-07-19 21:03 ` [RFC 2/5] NFSD: Implement the COPY call bjschuma
2013-07-22 18:05   ` J. Bruce Fields
2013-07-22 18:59     ` Bryan Schumaker
2013-07-19 21:03 ` [RFC 3/5] NFS: Add COPY nfs operation bjschuma
2013-07-24 14:21   ` Myklebust, Trond
2013-07-19 21:03 ` [RFC 4/5] NFSD: Defer copying bjschuma
2013-07-22 18:50   ` J. Bruce Fields
2013-07-22 19:17     ` Bryan Schumaker
2013-07-22 19:30       ` J. Bruce Fields
2013-07-22 19:37         ` Bryan Schumaker
2013-07-22 19:43           ` J. Bruce Fields
2013-07-22 19:54             ` Bryan Schumaker
2013-07-22 19:55               ` J. Bruce Fields
2013-08-05  8:38                 ` Ric Wheeler
2013-08-05 14:41                   ` J. Bruce Fields
2013-08-05 14:44                     ` Ric Wheeler
2013-08-05 14:56                       ` J. Bruce Fields
2013-08-05 14:44                     ` Bryan Schumaker
2013-08-05 14:50                     ` Myklebust, Trond
2013-08-05 18:11                       ` J. Bruce Fields
2013-08-05 18:17                         ` Chuck Lever
2013-08-05 18:24                           ` Myklebust, Trond
2013-08-05 18:30                         ` J. Bruce Fields
2013-07-19 21:03 ` [RFC 5/5] NFS: Change copy to support async servers bjschuma
2013-07-24 14:28   ` Myklebust, Trond
2013-07-22 18:53 ` [RFC 0/5] NFS Server Side Copy J. Bruce Fields
2013-07-22 19:38   ` Bryan Schumaker
2013-07-22 19:42     ` J. Bruce Fields [this message]

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=20130722194209.GE10109@fieldses.org \
    --to=bfields@fieldses.org \
    --cc=Trond.Myklebust@netapp.com \
    --cc=bjschuma@netapp.com \
    --cc=linux-nfs@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).