From: Linus Torvalds <torvalds@osdl.org>
To: Daniel Jacobowitz <dan@debian.org>
Cc: Andries Brouwer <aebr@win.tue.nl>, Rob Love <rml@ximian.com>,
rob@landley.net, Pascal Schmidt <der.eremit@email.de>,
linux-kernel@vger.kernel.org, Greg KH <greg@kroah.com>
Subject: Re: udev and devfs - The final word
Date: Sun, 4 Jan 2004 19:33:16 -0800 (PST) [thread overview]
Message-ID: <Pine.LNX.4.58.0401041918260.2162@home.osdl.org> (raw)
In-Reply-To: <20040105030737.GA29964@nevyn.them.org>
On Sun, 4 Jan 2004, Daniel Jacobowitz wrote:
>
> I think you two are talking straight past each other. Andries is
> talking about the fsid, which is determined by the NFS server, based on
> its idea of the device number of the filesystem underlying the exported
> directory. Right now, I can reboot my host system, and when it comes
> up then the NFS directories it exports to clients will have the same
> fsid. With random device numbers it won't work; after rebooting the
> NFS server all clients will be forced to explicitly unmount and
> remount.
Ahh. I'll buy into that, and yes, this is an example of something that
needs to be fixed.
It shouldn't be fixed by saying "device numbers have to be stable across
reboots", because the fact is, we're most likely going to have storage
that is really really hard to enumerate in a repeatable fashion.
So the _proper_ thing to do is to have the NFS server not use the device
number as part of fsid. It should use the _stable_ UUID of the filesystem
or some similar label.
And it should do that exactly because the device number isn't as stable as
NFS exporting would like it to be. Exactly because things like network-
attached disks etc. How would you otherwise export a disk that perhaps
gets its address from DHCP?
[ I incredulously asked a NetApp person why you'd ever want to expose the
_disk_ over ethenet, rather than just have the NAS device export a
filesystem of its own. It turns out that some people really want to just
see a block device, either because Windows sucks at network filesystems
or because they want to do things like databases onto them. The point
being that once you do that, you'll likely want to export the thing as
an SMB share from the thing that "owns" the disk.
So you would literally have a _disk_ whose IP address changed depending
on what other machines were booted on the same network. ]
Issues like this is also why Linux vendors have already started doing
things like "mount by label" - because disks have a nasty tendency to move
around, and specifying the fstab contents (or "root=xxx" on the kernel
command line) with physical location or similar just doesn't work all
that well. It happens today with things like USB2 or firewire disks. They
get moved around, and they get a new device number.
It's still not _common_, but it's slowly getting there.
> Now, it seems to me that this isn't much of an argument against random
> device numbers. Have userspace set a UUID for the device if you want,
> and use that in the fsid instead. But that's the argument; it has
> nothing to do with the NFS server exporting its /dev.
I buy into that, and I agree 100% with you that this is just a case where
you should use a UUID.
Linus
next prev parent reply other threads:[~2004-01-05 3:33 UTC|newest]
Thread overview: 169+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <18Cz7-7Ep-7@gated-at.bofh.it>
2003-12-31 3:05 ` udev and devfs - The final word Pascal Schmidt
2003-12-31 19:23 ` Greg KH
2003-12-31 20:19 ` Rob Love
2003-12-31 22:01 ` Nathan Conrad
2003-12-31 22:20 ` Rob Love
2003-12-31 21:45 ` Tommi Virtanen
2003-12-31 23:10 ` Rob Love
2003-12-31 21:52 ` Tommi Virtanen
2004-01-02 0:17 ` Hollis Blanchard
2004-01-02 0:36 ` viro
2004-01-03 6:04 ` Greg KH
2003-12-31 22:55 ` viro
2003-12-31 23:05 ` Rob Love
2003-12-31 23:48 ` Andreas Dilger
2004-01-07 10:15 ` Olaf Hering
2004-01-07 11:18 ` viro
2004-01-07 13:00 ` Olaf Hering
2004-01-07 13:26 ` viro
2004-01-07 13:27 ` Olaf Hering
2004-01-01 0:15 ` Andries Brouwer
2004-01-01 0:31 ` Rob Love
2004-01-01 12:34 ` Rob Landley
2004-01-01 15:22 ` Rob Love
2004-01-01 15:48 ` Andries Brouwer
2004-01-01 15:54 ` Rob Love
2004-01-02 20:42 ` Linus Torvalds
2004-01-03 3:00 ` Andries Brouwer
2004-01-03 4:46 ` Linus Torvalds
2004-01-03 13:10 ` Andries Brouwer
2004-01-03 22:27 ` Linus Torvalds
2004-01-03 23:08 ` Andries Brouwer
2004-01-04 1:16 ` Mark Mielke
2004-01-04 1:54 ` Valdis.Kletnieks
2004-01-04 18:44 ` Mark Mielke
2004-01-04 2:09 ` Linus Torvalds
2004-01-04 2:49 ` Andries Brouwer
2004-01-04 3:04 ` Linus Torvalds
2004-01-04 4:36 ` Pentium 4 HT SMP Ananda Bhattacharya
2004-01-04 5:55 ` Martin J. Bligh
2004-01-04 13:21 ` udev and devfs - The final word Andries Brouwer
2004-01-04 21:05 ` Linus Torvalds
2004-01-04 22:01 ` Andries Brouwer
2004-01-04 22:37 ` viro
2004-01-05 1:02 ` Mark Mielke
2004-01-05 2:24 ` Valdis.Kletnieks
2004-01-05 2:29 ` Andries Brouwer
2004-01-05 3:42 ` viro
2004-01-04 22:37 ` Helge Hafting
2004-01-04 23:35 ` Valdis.Kletnieks
2004-01-05 1:43 ` Jeremy Maitin-Shepard
2004-01-05 1:47 ` st_dev:st_ino (was: Re: udev and devfs - The final word) Mark Mielke
2004-01-05 2:02 ` st_dev:st_ino Jeremy Maitin-Shepard
2004-01-05 3:14 ` st_dev:st_ino viro
2004-01-05 1:58 ` udev and devfs - The final word viro
2004-01-05 2:12 ` Jeremy Maitin-Shepard
2004-01-05 2:52 ` Linus Torvalds
2004-01-05 3:06 ` David Lang
2004-01-05 3:48 ` Rob Landley
2004-01-05 4:52 ` Trond Myklebust
2004-01-05 7:03 ` [offtopic] " Rob Landley
2004-01-05 12:07 ` Trond Myklebust
2004-01-05 15:13 ` Mark Mielke
2004-01-05 16:36 ` Andreas Schwab
2004-01-05 22:18 ` Mark Mielke
2004-01-05 3:07 ` Daniel Jacobowitz
2004-01-05 3:33 ` Linus Torvalds [this message]
2004-01-05 3:50 ` viro
2004-01-05 4:02 ` Linus Torvalds
2004-01-05 4:38 ` viro
2004-01-05 4:52 ` Linus Torvalds
2004-01-05 6:11 ` viro
2004-01-05 7:47 ` Greg KH
2004-01-05 11:15 ` Vojtech Pavlik
2004-01-05 20:11 ` Theodore Ts'o
2004-01-05 21:06 ` Vojtech Pavlik
2004-01-05 22:22 ` Theodore Ts'o
2004-01-06 0:14 ` Rob Landley
2004-01-06 17:28 ` [OT] " Disconnect
2004-01-11 22:12 ` Ed L Cashin
2004-01-05 5:26 ` Eric W. Biederman
2004-01-05 7:39 ` Greg KH
2004-01-07 9:57 ` Pavel Machek
2004-01-05 12:27 ` Andries Brouwer
2004-01-05 16:13 ` Linus Torvalds
2004-01-05 17:29 ` Vojtech Pavlik
2004-01-05 17:33 ` Linus Torvalds
2004-01-05 17:52 ` Davide Libenzi
2004-01-05 18:03 ` Linus Torvalds
2004-01-05 18:09 ` Hugo Mills
2004-01-05 19:10 ` Paul Rolland
2004-01-05 19:52 ` Andries Brouwer
2004-01-05 20:38 ` Linus Torvalds
2004-01-05 22:17 ` Shawn
2004-01-05 22:25 ` Mark Mielke
2004-01-05 23:05 ` Shawn
2004-01-05 23:23 ` Shawn
2004-01-06 0:43 ` Greg KH
2004-01-06 0:53 ` Shawn
2004-01-05 23:13 ` Andries Brouwer
2004-01-05 23:32 ` Linus Torvalds
2004-01-06 0:59 ` viro
2004-01-06 1:17 ` Linus Torvalds
2004-01-06 4:28 ` viro
2004-01-06 5:07 ` Linus Torvalds
2004-01-06 1:06 ` Andries Brouwer
2004-01-06 15:00 ` Mark Mielke
2004-01-06 0:00 ` Greg KH
2004-01-06 1:41 ` Andries Brouwer
2004-01-07 17:14 ` Greg KH
2004-01-06 0:31 ` Rob Landley
2004-01-06 7:14 ` Vojtech Pavlik
2004-01-06 0:36 ` Silly udev script [was Re: udev and devfs - The final word] Greg KH
2004-01-05 7:44 ` udev and devfs - The final word James H. Cloos Jr.
2004-01-05 7:45 ` Nigel Cunningham
2004-01-05 11:01 ` Robin Rosenberg
2004-01-05 12:39 ` Nigel Cunningham
2004-01-05 14:31 ` IRQ disabled on linux 2.6.1-rc1-mm1 Mainak Mandal _00007001_
2004-01-07 13:39 ` udev and devfs - The final word Robin Rosenberg
2004-01-07 17:16 ` Nigel Cunningham
2004-01-05 9:06 ` Valdis.Kletnieks
2004-01-05 4:15 ` Peter Chubb
2004-01-05 4:42 ` Linus Torvalds
2004-01-03 18:34 ` Wrapping jiffies [was Re: udev and devfs - The final word] Pavel Machek
2004-01-01 19:43 ` udev and devfs - The final word Kai Henningsen
2004-01-02 7:26 ` Rob Landley
2004-01-04 8:57 ` Greg KH
2004-01-04 9:43 ` Rob Landley
2004-01-02 0:17 ` Maciej Zenczykowski
[not found] ` <20040102103104.GA28168@mark.mielke.cc>
2004-01-03 6:07 ` Greg KH
2004-01-03 6:51 ` Valdis.Kletnieks
2004-01-03 11:57 ` Ian Kent
2004-01-03 22:08 ` Greg KH
2004-01-07 10:23 ` Olaf Hering
2004-01-01 23:14 ` Rob
2004-01-02 3:53 ` Tyler Hall
2004-01-01 16:17 ` Pascal Schmidt
2004-01-01 20:03 ` Greg KH
2004-01-08 13:53 "Andrey Borzenkov"
2004-01-08 15:40 ` Ian Kent
2004-01-08 17:26 ` Diego Calleja
2004-01-08 19:25 ` Andrey Borzenkov
2004-01-08 22:40 ` Alex Goddard
2004-01-09 7:03 ` "Andrey Borzenkov"
2004-01-08 18:14 ` Alex Goddard
2004-01-08 18:35 ` Alex Goddard
2004-01-08 19:22 ` Andrey Borzenkov
2004-01-09 8:51 ` Helge Hafting
-- strict thread matches above, loose matches on Subject: below --
2004-01-08 13:05 "Andrey Borzenkov"
2004-01-06 1:20 Paul Zimmerman
[not found] <fa.flhsork.uka2hg@ifi.uio.no>
[not found] ` <fa.hv9hpq7.1l1q9p3@ifi.uio.no>
2004-01-01 19:53 ` walt
2004-01-01 21:53 ` Martin Schlemmer
2004-01-01 16:59 Shaheed
[not found] <fa.af64864.ugabhg@ifi.uio.no>
[not found] ` <fa.de7jae9.1jk0pjt@ifi.uio.no>
2003-12-31 22:17 ` walt
2004-01-01 2:03 ` Martin Schlemmer
2004-01-01 2:05 ` Martin Schlemmer
2003-12-31 0:29 Greg KH
2003-12-31 0:53 ` Prakash K. Cheemplavam
2003-12-31 19:17 ` Greg KH
2004-01-02 16:45 ` Shawn
2003-12-31 12:43 ` Paulo Marques
2004-01-01 1:18 ` Helge Hafting
2004-01-03 5:59 ` Greg KH
2004-01-03 15:22 ` Helge Hafting
2004-01-03 21:18 ` viro
2004-01-03 22:11 ` Greg KH
[not found] ` <20040103140140.3b848e9f.witukind@nsbm.kicks-ass.org>
2004-01-03 22:16 ` Greg KH
2004-01-03 22:33 ` Christoph Hellwig
2004-01-02 17:54 ` Andreas Jellinghaus
2004-01-02 18:19 ` Shawn
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=Pine.LNX.4.58.0401041918260.2162@home.osdl.org \
--to=torvalds@osdl.org \
--cc=aebr@win.tue.nl \
--cc=dan@debian.org \
--cc=der.eremit@email.de \
--cc=greg@kroah.com \
--cc=linux-kernel@vger.kernel.org \
--cc=rml@ximian.com \
--cc=rob@landley.net \
/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).