From: Daniel Phillips <phillips@istop.com>
To: Andrew Morton <akpm@osdl.org>
Cc: Joel.Becker@oracle.com, linux-cluster@redhat.com,
wim.coekaerts@oracle.com, linux-fsdevel@vger.kernel.org,
ak@suse.de, linux-kernel@vger.kernel.org
Subject: Re: [Linux-cluster] Re: GFS, what's remaining
Date: Sun, 4 Sep 2005 02:40:08 -0400 [thread overview]
Message-ID: <200509040240.08467.phillips@istop.com> (raw)
In-Reply-To: <20050903214653.1b8a8cb7.akpm@osdl.org>
On Sunday 04 September 2005 00:46, Andrew Morton wrote:
> Daniel Phillips <phillips@istop.com> wrote:
> > The model you came up with for dlmfs is beyond cute, it's downright
> > clever.
>
> Actually I think it's rather sick. Taking O_NONBLOCK and making it a
> lock-manager trylock because they're kinda-sorta-similar-sounding? Spare
> me. O_NONBLOCK means "open this file in nonblocking mode", not "attempt to
> acquire a clustered filesystem lock". Not even close.
Now, I see the ocfs2 guys are all ready to back down on this one, but I will
at least argue weakly in favor.
Sick is a nice word for it, but it is actually not that far off. Normally,
this fs will acquire a lock whenever the user creates a virtual file and the
create will block until the global lock arrives. With O_NONBLOCK, it will
return, erm... ETXTBSY (!) immediately. Is that not what O_NONBLOCK is
supposed to accomplish?
> It would be much better to do something which explicitly and directly
> expresses what you're trying to do rather than this strange "lets do this
> because the names sound the same" thing.
>
> What happens when we want to add some new primitive which has no posix-file
> analog?
>
> Waaaay too cute. Oh well, whatever.
The explicit way is syscalls or a set of ioctls, which he already has the
makings of. If there is going to be a userspace api, I would hope it looks
more like the contents of userdlm.c than the traditional Vaxcluster API,
which sucks beyond belief.
Another explicit way is to do it with a whole set of virtual attributes
instead of just a single file trying to capture the whole model. That is
really unappealing, but I am afraid that is exactly what a whole lot of
sysfs/configfs usage is going to end up looking like.
But more to the point: we have no urgent need for a userspace dlm api at the
moment. Nothing will break if we just put that issue off for a few months,
quite the contrary.
If the only user is their tools I would say let it go ahead and be cute, even
sickeningly so. It is not supposed to be a general dlm api, at least that is
my understanding. It is just supposed to be an interface for their tools.
Of course it would help to know exactly how those tools use it. Too sleepy
to find out tonight...
Regards,
Daniel
next prev parent reply other threads:[~2005-09-04 6:37 UTC|newest]
Thread overview: 106+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-09-01 10:46 GFS, what's remaining David Teigland
2005-09-01 10:42 ` Arjan van de Ven
2005-09-01 10:59 ` Andrew Morton
2005-09-01 14:49 ` Alan Cox
2005-09-01 14:27 ` Christoph Hellwig
2005-09-01 15:28 ` Alan Cox
2005-09-01 15:11 ` Lars Marowsky-Bree
2005-09-01 17:56 ` Christoph Hellwig
2005-09-02 7:04 ` David Teigland
2005-09-01 17:23 ` Daniel Phillips
2005-09-01 20:21 ` Andrew Morton
2005-09-02 21:17 ` Andi Kleen
2005-09-02 23:03 ` Bryan Henderson
2005-09-03 0:16 ` Mark Fasheh
2005-09-03 6:42 ` Daniel Phillips
2005-09-03 6:46 ` Wim Coekaerts
2005-09-03 22:21 ` Daniel Phillips
2005-09-04 1:09 ` [Linux-cluster] " Joel Becker
2005-09-04 1:32 ` Andrew Morton
2005-09-04 3:06 ` Joel Becker
2005-09-04 4:22 ` [Linux-cluster] " Daniel Phillips
2005-09-04 4:30 ` Joel Becker
2005-09-04 4:51 ` Daniel Phillips
2005-09-04 5:00 ` Joel Becker
2005-09-04 5:52 ` [Linux-cluster] " Daniel Phillips
2005-09-04 5:56 ` Joel Becker
2005-09-04 4:46 ` Andrew Morton
2005-09-04 4:58 ` Joel Becker
2005-09-04 5:41 ` Andrew Morton
2005-09-04 5:49 ` Joel Becker
2005-09-05 4:30 ` David Teigland
2005-09-05 8:54 ` [Linux-cluster] " Andrew Morton
2005-09-05 9:24 ` David Teigland
2005-09-05 9:19 ` [Linux-cluster] " Andrew Morton
2005-09-05 9:30 ` Daniel Phillips
2005-09-05 9:48 ` David Teigland
2005-09-05 12:21 ` Alan Cox
2005-09-05 19:53 ` [Linux-cluster] " Andrew Morton
2005-09-05 23:20 ` Alan Cox
2005-09-05 23:06 ` Andrew Morton
2005-09-14 9:01 ` [Linux-cluster] " Patrick Caulfield
2005-09-05 19:11 ` kurt.hackel
2005-09-04 6:10 ` Mark Fasheh
2005-09-04 7:23 ` Andrew Morton
2005-09-04 8:17 ` Mark Fasheh
2005-09-04 8:37 ` Andrew Morton
2005-09-04 6:40 ` Daniel Phillips [this message]
2005-09-04 7:28 ` Andrew Morton
2005-09-04 8:01 ` [Linux-cluster] " Joel Becker
2005-09-04 8:18 ` Andrew Morton
2005-09-04 9:11 ` Joel Becker
2005-09-04 9:18 ` [Linux-cluster] " Andrew Morton
2005-09-04 9:39 ` Joel Becker
2005-09-04 18:03 ` [Linux-cluster] " Hua Zhong
2005-09-04 19:51 ` Daniel Phillips
2005-09-04 7:12 ` Hua Zhong
2005-09-04 8:37 ` Alan Cox
2005-09-05 23:32 ` Joel Becker
2005-09-03 5:57 ` Daniel Phillips
2005-09-05 14:14 ` Lars Marowsky-Bree
2005-09-05 15:49 ` Daniel Phillips
2005-09-05 16:18 ` Dmitry Torokhov
2005-09-06 0:57 ` Daniel Phillips
2005-09-06 2:03 ` Dmitry Torokhov
2005-09-06 4:02 ` Daniel Phillips
2005-09-06 4:07 ` GFS, what's remainingh Dmitry Torokhov
2005-09-06 4:58 ` Daniel Phillips
2005-09-06 5:05 ` Dmitry Torokhov
2005-09-06 6:48 ` Daniel Phillips
2005-09-06 6:55 ` Dmitry Torokhov
2005-09-06 7:18 ` Daniel Phillips
2005-09-06 14:31 ` Dmitry Torokhov
2005-09-06 13:42 ` Alan Cox
2005-09-03 7:06 ` GFS, what's remaining Wim Coekaerts
2005-09-06 12:55 ` Suparna Bhattacharya
2005-09-03 5:18 ` David Teigland
2005-09-03 6:14 ` Arjan van de Ven
2005-09-03 6:42 ` D. Hazelton
2005-09-03 10:35 ` David Teigland
2005-09-03 20:56 ` Daniel Phillips
2005-09-04 20:33 ` Pavel Machek
2005-09-04 22:18 ` Joel Becker
2005-09-05 5:54 ` Theodore Ts'o
2005-09-05 7:09 ` Mark Fasheh
2005-09-05 14:07 ` Theodore Ts'o
2005-09-05 8:27 ` real read-only [was Re: GFS, what's remaining] Pavel Machek
2005-09-05 14:03 ` Theodore Ts'o
2005-09-05 10:44 ` Re: GFS, what's remaining Stephen C. Tweedie
2005-09-05 16:41 ` Greg Freemyer
2005-09-01 11:35 ` Arjan van de Ven
2005-09-02 9:44 ` David Teigland
2005-09-02 11:46 ` Jörn Engel
2005-09-03 5:28 ` Greg KH
2005-09-05 3:47 ` David Teigland
2005-09-05 8:58 ` Jörn Engel
2005-09-05 9:18 ` David Teigland
2005-09-05 5:43 ` David Teigland
2005-09-05 6:32 ` Pekka Enberg
2005-09-05 7:55 ` David Teigland
2005-09-05 8:00 ` Pekka Enberg
2005-09-10 10:11 ` Arjan van de Ven
2005-09-05 6:29 ` David Teigland
2005-09-08 5:41 ` David Teigland
2005-09-01 12:33 ` Pekka Enberg
2005-09-01 17:27 ` Daniel Phillips
-- strict thread matches above, loose matches on Subject: below --
2005-09-01 18:47 [Linux-cluster] " Hua Zhong (hzhong)
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=200509040240.08467.phillips@istop.com \
--to=phillips@istop.com \
--cc=Joel.Becker@oracle.com \
--cc=ak@suse.de \
--cc=akpm@osdl.org \
--cc=linux-cluster@redhat.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=wim.coekaerts@oracle.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).