[linux-cifs-client] cifs mount failing

[BJ Taylor]

kernel version for the server is: 2.6.22.19-0.2-bigsmp
kernel version for the client is: 2.6.27.19-3.2-default

The servers were each updated about 4 months ago.  So not extremely current,
but not bad either.

I didn't see any messages in /var/log/messages that were related.

I realized that when I said that these mounted drive saw moderate use, that
may not be quite accurate.  Compared to how a lot of systems use mounted
drives, these use them fairly heavily.  These are web servers that server a
fair number of clients, and each request to the web server makes at least
one file read over the cifs mounted folder.  These systems never use
hibernate or anything like that, but they do use the drive folder fairly
heavily.

I will update the systems again this week, and if we can still reproduce the
issue, I will report it in the bugzilla.

Is there any way to increase the verbosity of the client logging?  I don't
even know which service is the most directly related to cifs folders, or is
this just part of the kernel itself?  I wish I had more concrete information
to put in bugzilla so that it won't just be dismissed as 'not
reproducable'.  Any suggestions on that?

BJ Taylor
Property Solutions International, Inc.
btaylor at propertysolutions.com



On Thu, Jul 9, 2009 at 10:13 PM, Suresh Jayaraman <sjayaraman at suse.de>wrote:

> BJ Taylor wrote:
> > I have 2 opensuse 11.1 machines that have a mounted folder that's
> > connected to an opensuse 10.3 samba server.� The folder is fairly
> > consistently.� However, every few days, the folder will stop
> > responding.� Doing a 'ls' on the folder (not inside the folder) I get an
> > error saying that the host server is not responding.� However, I have at
> > least 4 other servers that are connected to the same share and do not
>
> What is the kernel version being used (uname -r) and the Samba server
> version? Are you running up-to-date openSUSE 10.3 and 11.1 boxes?
>
> > ever get this problem.� Once the machine gets this problem, the only way
> > I know to fix it is to reboot the machine.� umount -l and umount -f both
> > crash the machine (ie, it is unresponsive to all network traffic and
> > even from the monitor and keyboard plugged into it, and will stay that
> > way for days).� Searching the logs has been futile.� I either do not
> > know where to look or nothing is being recorded.� Using netstat I can
>
> Do you find any CIFS related/Oops message in /var/log/messages?
>
> > Are there any known bugs that would cause this?� Would updating the
> > samba server to opensuse 11.1 fix the problem by any chance?� Any advice
> > on how I can further debug this and get some meaningful output?
> >
>
> Doesn't sound like a known issue to me. But, there is no harm in
> updating systems (both client and server) and see whether the problem is
> still reproducible or not. Do you suspend/hibernate the box with cifs
> mounted?
>
> Also, please report the bug here:
>
> https://bugzilla.novell.com/enter_bug.cgi?product=openSUSE+11.1&format=guided
>
>
> Thanks,
>
> --
> Suresh Jayaraman
>
-------------- next part --------------
HTML attachment scrubbed and removed