From: Christoph Hellwig <hch@infradead.org>
To: R?diger Meier <sweet_f_a@gmx.de>
Cc: Christoph Hellwig <hch@infradead.org>,
Bryan Schumaker <bjschuma@netapp.com>,
Justin Piszcz <jpiszcz@lucidpixels.com>,
"J. Bruce Fields" <bfields@fieldses.org>,
linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org,
xfs@oss.sgi.com
Subject: Re: 2.6.xx: NFS: directory motion/cam2 contains a readdir loop
Date: Wed, 27 Jul 2011 16:47:24 -0400 [thread overview]
Message-ID: <20110727204724.GA21314@infradead.org> (raw)
In-Reply-To: <201107272227.03265.sweet_f_a@gmx.de>
On Wed, Jul 27, 2011 at 10:26:55PM +0200, R?diger Meier wrote:
> At the time I've started this thread
> http://comments.gmane.org/gmane.linux.nfs/40863
> I had the feeling that the readdir cache changings in 2.6.37 have
> something to do with these loop problems.
>
> After that thread I've accepted that's a general problem with
> ext4/dirindex and nfs but seeing it again on xfs with just 5000 files
> I'm in doubt again.
Two separate issues. For one thing the nfs code simply doesn't seem
to handle changing directories very well, and one and a half the Linux
NFS server might even send incoherent readdir output in a single
protocol reply.
Issue two is that the ext3/4 hashed directory format is too simply (not
to say dumb) to provide a proper 32-bit linear value for the dirent
d_off field. It's not a complex task, and the first relatively simple
generation of xfs btree directories couldn't handle it either. The
v2 directory format handles it fine, but at the cost of a much more
complex codebase.
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org>
To: R?diger Meier <sweet_f_a@gmx.de>
Cc: "J. Bruce Fields" <bfields@fieldses.org>,
linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org,
Bryan Schumaker <bjschuma@netapp.com>,
Christoph Hellwig <hch@infradead.org>,
Justin Piszcz <jpiszcz@lucidpixels.com>,
xfs@oss.sgi.com
Subject: Re: 2.6.xx: NFS: directory motion/cam2 contains a readdir loop
Date: Wed, 27 Jul 2011 16:47:24 -0400 [thread overview]
Message-ID: <20110727204724.GA21314@infradead.org> (raw)
In-Reply-To: <201107272227.03265.sweet_f_a@gmx.de>
On Wed, Jul 27, 2011 at 10:26:55PM +0200, R?diger Meier wrote:
> At the time I've started this thread
> http://comments.gmane.org/gmane.linux.nfs/40863
> I had the feeling that the readdir cache changings in 2.6.37 have
> something to do with these loop problems.
>
> After that thread I've accepted that's a general problem with
> ext4/dirindex and nfs but seeing it again on xfs with just 5000 files
> I'm in doubt again.
Two separate issues. For one thing the nfs code simply doesn't seem
to handle changing directories very well, and one and a half the Linux
NFS server might even send incoherent readdir output in a single
protocol reply.
Issue two is that the ext3/4 hashed directory format is too simply (not
to say dumb) to provide a proper 32-bit linear value for the dirent
d_off field. It's not a complex task, and the first relatively simple
generation of xfs btree directories couldn't handle it either. The
v2 directory format handles it fine, but at the cost of a much more
complex codebase.
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
next prev parent reply other threads:[~2011-07-27 20:47 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-07-27 13:54 2.6.xx: NFS: directory motion/cam2 contains a readdir loop Justin Piszcz
2011-07-27 16:07 ` J. Bruce Fields
2011-07-27 16:28 ` Justin Piszcz
2011-07-27 16:28 ` Justin Piszcz
2011-07-27 16:40 ` Bryan Schumaker
2011-07-27 16:40 ` Bryan Schumaker
2011-07-27 17:00 ` Ruediger Meier
2011-07-27 17:00 ` Ruediger Meier
2011-07-27 17:09 ` Bryan Schumaker
2011-07-27 17:09 ` Bryan Schumaker
2011-07-27 17:17 ` Justin Piszcz
2011-07-27 17:17 ` Justin Piszcz
2011-07-27 17:45 ` J. Bruce Fields
2011-07-27 17:45 ` J. Bruce Fields
2011-07-27 18:28 ` Bryan Schumaker
2011-07-27 18:28 ` Bryan Schumaker
2011-07-27 17:15 ` Justin Piszcz
2011-07-27 17:15 ` Justin Piszcz
2011-07-27 18:11 ` Christoph Hellwig
2011-07-27 18:11 ` Christoph Hellwig
2011-07-27 19:35 ` Justin Piszcz
2011-07-27 19:35 ` Justin Piszcz
2011-07-27 19:39 ` Christoph Hellwig
2011-07-27 19:39 ` Christoph Hellwig
2011-07-27 19:44 ` Justin Piszcz
2011-07-27 19:44 ` Justin Piszcz
2011-07-27 19:47 ` Christoph Hellwig
2011-07-27 19:47 ` Christoph Hellwig
2011-07-27 19:54 ` Bryan Schumaker
2011-07-27 19:54 ` Bryan Schumaker
2011-07-27 20:02 ` Christoph Hellwig
2011-07-27 20:02 ` Christoph Hellwig
2011-07-27 20:05 ` Christoph Hellwig
2011-07-27 20:05 ` Christoph Hellwig
2011-07-27 20:26 ` Rüdiger Meier
2011-07-27 20:26 ` Rüdiger Meier
2011-07-27 20:47 ` Christoph Hellwig [this message]
2011-07-27 20:47 ` Christoph Hellwig
2011-07-27 21:21 ` Rüdiger Meier
2011-07-27 21:21 ` Rüdiger Meier
2011-07-27 19:57 ` Justin Piszcz
2011-07-27 19:57 ` Justin Piszcz
2011-07-27 20:37 ` Trond Myklebust
2011-07-27 20:37 ` Trond Myklebust
2011-07-27 20:54 ` Trond Myklebust
2011-07-27 20:54 ` Trond Myklebust
2011-07-27 20:54 ` Trond Myklebust
2011-07-27 20:56 ` Trond Myklebust
2011-07-27 20:56 ` Trond Myklebust
2011-07-27 20:56 ` Trond Myklebust
[not found] ` <1311800195.25645.45.camel-SyLVLa/KEI9HwK5hSS5vWB2eb7JE58TQ@public.gmane.org>
2011-07-27 21:24 ` Justin Piszcz
2011-07-27 21:24 ` Justin Piszcz
2011-07-27 21:24 ` Justin Piszcz
[not found] ` <alpine.DEB.2.02.1107271723500.25432-0qmrozcXWo8bm2hyYBkBBg@public.gmane.org>
2011-07-27 22:44 ` Justin Piszcz
2011-07-27 22:44 ` Justin Piszcz
2011-07-27 22:44 ` Justin Piszcz
2011-07-28 20:48 ` Trond Myklebust
2011-07-28 20:48 ` Trond Myklebust
2011-07-28 20:48 ` Trond Myklebust
2011-07-29 20:52 ` Bryan Schumaker
2011-07-29 20:52 ` Bryan Schumaker
2011-07-29 20:59 ` Justin Piszcz
2011-07-29 20:59 ` Justin Piszcz
2011-07-29 22:03 ` Trond Myklebust
2011-07-29 22:03 ` Trond Myklebust
2011-07-29 22:23 ` Justin Piszcz
2011-07-29 22:23 ` Justin Piszcz
2011-07-30 9:58 ` Justin Piszcz
2011-07-30 9:58 ` Justin Piszcz
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=20110727204724.GA21314@infradead.org \
--to=hch@infradead.org \
--cc=bfields@fieldses.org \
--cc=bjschuma@netapp.com \
--cc=jpiszcz@lucidpixels.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=sweet_f_a@gmx.de \
--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.