From: "J. Bruce Fields" <bfields@fieldses.org>
To: Matthew Wilcox <matthew@wil.cx>
Cc: Peter Staubach <staubach@redhat.com>,
linux-fsdevel@vger.kernel.org,
David Teigland <teigland@redhat.com>,
nfs@lists.sourceforge.net, Marc Eshel <eshel@almaden.ibm.com>,
David Richter <richterd@umich.edu>
Subject: Re: [NFS] [PATCH] locks: provide a file lease method enabling cluster-coherent leases
Date: Fri, 1 Jun 2007 14:08:42 -0400 [thread overview]
Message-ID: <20070601180842.GG10492@fieldses.org> (raw)
In-Reply-To: <20070601174123.GS23968@parisc-linux.org>
On Fri, Jun 01, 2007 at 11:41:23AM -0600, Matthew Wilcox wrote:
> Samba internally prohibits renaming or deleting an open file, to match
> Windows semantics. So it won't notice the difference. At least, that's
> what I remember from a discussion with Tridge when we were implementing
> leases back in 2000.
NFSv4 should be doing the same thing (err, for some reason I thought it
already did--I remember writing patches to do it--but I must not have
finished them). But there's still the problem of renames done by local
users, other file servers in the multiprotocol case, etc.
> I think it's an acceptable change in Linux semantics to break leases on
> rename/delete/link. Though I'm not quite sure why you need to -- nobody
> else is touching the contents of the file, so it's not like you need to
> write the data back to it, or discard your cached copy of it in the case
> of a read-only lease.
NFSv4 delegations are defined to cover the file's metadata (including
the name it was opened under) as well as its data. Otherwise you
wouldn't be able to perform opens locally, because you wouldn't know
whether the open path still works, or whether the file permissions still
allow the same access.
And other parts of the protocol make use of this assumption (e.g. OPEN
with CLAIM_DELEGATE_CUR or CLAIM_DELEGATE_PREV, which expect to be able
to re-open a file under its original name).
--b.
next prev parent reply other threads:[~2007-06-01 18:08 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-05-31 21:40 cluster-coherent leases J. Bruce Fields
2007-05-31 21:40 ` [PATCH] locks: share more common lease code J. Bruce Fields
2007-05-31 21:40 ` [PATCH] locks: provide a file lease method enabling cluster-coherent leases J. Bruce Fields
2007-05-31 21:40 ` [PATCH] gfs2: stop giving out non-cluster-coherent leases J. Bruce Fields
2007-05-31 22:34 ` [PATCH] locks: provide a file lease method enabling cluster-coherent leases Trond Myklebust
2007-06-01 16:53 ` [NFS] " J. Bruce Fields
2007-06-02 17:39 ` Trond Myklebust
2007-06-02 18:09 ` Marc Eshel
2007-06-02 18:21 ` Trond Myklebust
2007-06-04 13:59 ` J. Bruce Fields
2007-06-05 22:04 ` Robert Rappaport
2007-06-05 22:53 ` [NFS] " Trond Myklebust
2007-06-05 22:56 ` Marc Eshel
2007-06-05 23:40 ` Trond Myklebust
2007-06-06 18:43 ` Robert Rappaport
2007-06-07 14:43 ` [NFS] " Robert Rappaport
2007-06-07 17:05 ` J. Bruce Fields
2007-06-08 22:14 ` (no subject) J. Bruce Fields
2007-06-08 22:14 ` (unknown), J. Bruce Fields
[not found] ` <11813408953536-git-send-email->
2007-06-08 22:14 ` [PATCH 1/5] locks: share more common lease code J. Bruce Fields
2007-06-08 22:14 ` J. Bruce Fields
[not found] ` <11813408952518-git-send-email->
2007-06-08 22:14 ` [PATCH 2/5] locks: provide a file lease method enabling cluster-coherent leases J. Bruce Fields
2007-06-08 22:14 ` J. Bruce Fields
[not found] ` <11813408951909-git-send-email->
2007-06-08 22:14 ` [PATCH 3/5] locks: rename lease functions to reflect locks.c conventions J. Bruce Fields
2007-06-08 22:14 ` J. Bruce Fields
[not found] ` <11813408954053-git-send-email->
2007-06-08 22:14 ` [PATCH 4/5] gfs2: stop giving out non-cluster-coherent leases J. Bruce Fields
2007-06-08 22:14 ` J. Bruce Fields
[not found] ` <11813408951694-git-send-email->
2007-06-08 22:14 ` [PATCH 5/5] nfs: disable leases over NFS J. Bruce Fields
2007-06-08 22:14 ` J. Bruce Fields
2007-06-09 14:18 ` [PATCH 4/5] gfs2: stop giving out non-cluster-coherent leases Steven Whitehouse
2007-06-09 16:35 ` Marc Eshel
2007-06-11 9:38 ` Steven Whitehouse
2007-06-11 17:07 ` J. Bruce Fields
2007-06-09 16:56 ` [PATCH 3/5] locks: rename lease functions to reflect locks.c conventions Marc Eshel
2007-06-11 16:53 ` J. Bruce Fields
2007-06-02 18:23 ` [NFS] [PATCH] locks: provide a file lease method enabling cluster-coherent leases Trond Myklebust
2007-06-01 13:14 ` Peter Staubach
2007-06-01 16:44 ` [NFS] " J. Bruce Fields
2007-06-01 17:41 ` Matthew Wilcox
2007-06-01 18:08 ` J. Bruce Fields [this message]
2007-05-31 21:51 ` [PATCH] locks: share more common lease code Trond Myklebust
2007-06-01 16:30 ` [NFS] " J. Bruce Fields
2007-06-01 16:36 ` Trond Myklebust
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=20070601180842.GG10492@fieldses.org \
--to=bfields@fieldses.org \
--cc=eshel@almaden.ibm.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=matthew@wil.cx \
--cc=nfs@lists.sourceforge.net \
--cc=richterd@umich.edu \
--cc=staubach@redhat.com \
--cc=teigland@redhat.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.