Bug#568288: mount.nfs: mount(2): Invalid argument

2011-03-21 Thread ael
On Sun, Mar 20, 2011 at 11:32:32PM +0100, Luk Claes wrote:
  As noted above, I only had version 4 configured in one of them:
  might that have upset the negotiation/default somehow?
 
 It might if the export was not configured to be for NFSv4 (with an
 fsid=0 entry etc). Another possibility is a bug in an older version of
 the Debian package or using an older kernel at the other end AFAICS.
 
 Can you still reproduce this issue or are you fine with me closing this bug?

I will have to find time to experiment. I still think the error message 
could be improved, although I guess if mount is talking to a kernel
which is not configured for the right version, it is a tall order for
it to diagnose that that is the problem.

ael




-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/2011032844.GA2341@elf.conquest



Bug#568288: mount.nfs: mount(2): Invalid argument

2011-03-20 Thread Luk Claes
Hi

 Mounting an nfs file system suddenly stopped working with errors
 like this:
 --
 # mount.nfs  exact:/ /mnt/exact -v
 mount.nfs: timeout set for Wed Feb  3 16:09:00 2010
 mount.nfs: trying text-based options
'addr=192.168.0.5,vers=4,clientaddr=192.168.0.3'
 mount.nfs: mount(2): Invalid argument
 mount.nfs: an incorrect mount option was specified
 --

 Both systems (that on which the mount was issued and the server)
 were running testing. They were, however using custom kernels, but these
 had worked without problems until today. Only the server was set up for
 nfs version 4 which seems to be the heart of the problem from the message
 about the text based options above - see below as well.

I guess it's an old kernel?

 The problem was resolved by adding either -o nfsvers=2 or
 -o nfsvers-3 to the mount command.

 Thus there seems to be some sort of problem negotiating the nfs version
 between the client  server? And a totally misleading error message?

The default changed to version 4, but that does not work with old
kernels, so I guess that's the problem?

That would also explain the error message.

Cheers

Luk



-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4d864a2f@debian.org



Bug#568288: mount.nfs: mount(2): Invalid argument

2011-03-20 Thread ael
On Sun, Mar 20, 2011 at 07:40:47PM +0100, Luk Claes wrote:
 Hi
 
  Mounting an nfs file system suddenly stopped working with errors
  like this:
  --
  # mount.nfs  exact:/ /mnt/exact -v
  mount.nfs: timeout set for Wed Feb  3 16:09:00 2010
  mount.nfs: trying text-based options
 'addr=192.168.0.5,vers=4,clientaddr=192.168.0.3'
  mount.nfs: mount(2): Invalid argument
  mount.nfs: an incorrect mount option was specified
  --
 
  Both systems (that on which the mount was issued and the server)
  were running testing. They were, however using custom kernels, but these
  had worked without problems until today. Only the server was set up for
  nfs version 4 which seems to be the heart of the problem from the message
  about the text based options above - see below as well.
 
 I guess it's an old kernel?

Actually, no. I either run the current Debian kernel, or the latest
from stable-git, compiled locally, of course.


  The problem was resolved by adding either -o nfsvers=2 or
  -o nfsvers-3 to the mount command.
 
  Thus there seems to be some sort of problem negotiating the nfs version
  between the client  server? And a totally misleading error message?
 
 The default changed to version 4, but that does not work with old
 kernels, so I guess that's the problem?

I don't think it can be: I was running kernel.org kernels, later then
Debian testing. I can't remember what version I was at when I reported
the bug: I confess that I had forgotten about the report.
Maybe I had something wrong in the kernel configs, but I doubt it.
As noted above, I only had version 4 configured in one of them:
might that have upset the negotiation/default somehow?

ael



-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20110320221117.GA3806@precise.conquest



Bug#568288: mount.nfs: mount(2): Invalid argument

2011-03-20 Thread Luk Claes
On 03/20/2011 11:11 PM, ael wrote:
 On Sun, Mar 20, 2011 at 07:40:47PM +0100, Luk Claes wrote:
 Hi

 Mounting an nfs file system suddenly stopped working with errors
 like this:
 --
 # mount.nfs  exact:/ /mnt/exact -v
 mount.nfs: timeout set for Wed Feb  3 16:09:00 2010
 mount.nfs: trying text-based options
 'addr=192.168.0.5,vers=4,clientaddr=192.168.0.3'
 mount.nfs: mount(2): Invalid argument
 mount.nfs: an incorrect mount option was specified
 --

 Both systems (that on which the mount was issued and the server)
 were running testing. They were, however using custom kernels, but these
 had worked without problems until today. Only the server was set up for
 nfs version 4 which seems to be the heart of the problem from the message
 about the text based options above - see below as well.

 I guess it's an old kernel?
 
 Actually, no. I either run the current Debian kernel, or the latest
 from stable-git, compiled locally, of course.
 
 
 The problem was resolved by adding either -o nfsvers=2 or
 -o nfsvers-3 to the mount command.

 Thus there seems to be some sort of problem negotiating the nfs version
 between the client  server? And a totally misleading error message?

 The default changed to version 4, but that does not work with old
 kernels, so I guess that's the problem?
 
 I don't think it can be: I was running kernel.org kernels, later then
 Debian testing. I can't remember what version I was at when I reported
 the bug: I confess that I had forgotten about the report.
 Maybe I had something wrong in the kernel configs, but I doubt it.
 As noted above, I only had version 4 configured in one of them:
 might that have upset the negotiation/default somehow?

It might if the export was not configured to be for NFSv4 (with an
fsid=0 entry etc). Another possibility is a bug in an older version of
the Debian package or using an older kernel at the other end AFAICS.

Can you still reproduce this issue or are you fine with me closing this bug?

Cheers

Luk



-- 
To UNSUBSCRIBE, email to debian-kernel-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/4d868080.2060...@debian.org