From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg Bradner Subject: Re: auotfs v4 Date: Fri, 14 Nov 2003 06:29:51 -0800 Sender: autofs-bounces@linux.kernel.org Message-ID: <3FB4E6DF.9020907@rhythm.com> References: <3FABB076.5080702@rhythm.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============19026586840762727==" Return-path: In-Reply-To: <3FABB076.5080702@rhythm.com> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: autofs-bounces@linux.kernel.org To: autofs@linux.kernel.org This is a multi-part message in MIME format. --===============19026586840762727== Content-Type: multipart/alternative; boundary="------------070804030801050209080403" This is a multi-part message in MIME format. --------------070804030801050209080403 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit This remains a major problem for me. I have changed the exports file on the NetApp, but am still having the same problem. Why would /nfs/hosts3/fs7/vol/vo0 get umount but not exported directories under that mount point? Why does /etc/mtab reflect an umount but /proc/mounts show them mounted? Any ideas? Greg Bradner wrote: > I am using RH 7.3 with the 2.4.20 kernel. It doen't have the patch. > But I don't need it at the moment. I need to figure out why my > /proc/mounts file is filling up with the same mount points. > > My problem is a NetApp. Not because it's a NetApp, rather I believe > it's the way volumes are exported. > > showmount -e fs7 > Export list for fs7: > /vol/vol0/systems (everyone) > /vol/vol0/muse (everyone) > /vol/vol0/avidcache (everyone) > /vol/vol0 (everyone) > > What I see in /proc/mounts: > fs7:/vol/vol0/muse /nfs/hosts3/fs7/vol/vol0/muse nfs > rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0 > fs7:/vol/vol0/avidcache /nfs/hosts3/fs7/vol/vol0/avidcache nfs > rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0 > fs7:/vol/vol0/systems /nfs/hosts3/fs7/vol/vol0/systems nfs > rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0 > > But I don't see: > /vol/vol0 > > I believe autofs is expiring and unmounting fs7:/vol/vol0 and getting > confused as what to do with the directories under that root vol. > > > Ian Kent wrote: > >>On Thu, 6 Nov 2003, Mike Waychison wrote: >> >> >> >>>Greg Bradner wrote: >>> >>> >>> >>>>Any ideas what 'mount table full' means? >>>>Nov 6 10:08:44 linn automount[27555]: >> NFSv3 not supported! >>>>Nov 6 10:08:44 linn automount[27555]: >> mount table full >>>>Nov 6 10:08:44 linn automount[27555]: mount(nfs): nfs: mount failure >>>>fs7:/vol/vol0/systems/celia/utils on /celia/utils/ >>>>Nov 6 10:09:15 linn automount[1899]: attempting to mount entry >>>>/misc/hsm_cache >>>>Nov 6 10:09:15 linn automount[27589]: >> NFSv3 not supported! >>>>Nov 6 10:09:15 linn automount[27589]: >> mount table full >>>>Nov 6 10:09:15 linn automount[27589]: mount(nfs): nfs: mount failure >>>>fs8:/vol/vol0/hsm_cache on /misc/hsm_cache/ >>>> >>>>Everything works fine after a reboot. >>>> >>>> >>>> >>>It means that you have more than 256 mounts that don't have a >>>block-device back-end. These include NFS, autofs, procfs, sysfs... >>> >>>This is a known limitation in linux 2.4. You may want to try a patch >>>like the following: >>> >>>http://people.redhat.com/zaitcev/linux/linux-2.4.19-pre7-unmaj.diff >>> >>> >>> >>> >> >>Most of the RedHat packaged kernels have this patch. >> >>What distro are you using? >>What version of autofs? >> >> >> > >-- > >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >Rhythm & Hues >5404 Jandy Place >Los Angeles, CA 90066 >Voice: 310 448-7763 >Fax: 310 448-7600 >gregb@rhythm.com > > -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Rhythm & Hues 5404 Jandy Place Los Angeles, CA 90066 Voice: 310 448-7763 Fax: 310 448-7600 gregb@rhythm.com --------------070804030801050209080403 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit This remains a major problem for me. I have changed the exports file on the NetApp, but am still having the same problem.
Why would /nfs/hosts3/fs7/vol/vo0 get umount but not exported directories under that mount point?  Why does /etc/mtab reflect an umount but /proc/mounts show them mounted?

Any ideas?

Greg Bradner wrote:
I am using RH 7.3 with the 2.4.20 kernel. It doen't have the patch. But I don't need it at the moment. I need to figure out why my /proc/mounts file is filling up with the same mount points.

My problem is a NetApp. Not because it's a NetApp, rather I believe it's the way volumes are exported.
> showmount -e fs7
Export list for fs7:
/vol/vol0/systems   (everyone)
/vol/vol0/muse      (everyone)
/vol/vol0/avidcache (everyone)
/vol/vol0           (everyone)

What I see in /proc/mounts:
fs7:/vol/vol0/muse /nfs/hosts3/fs7/vol/vol0/muse nfs rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0
fs7:/vol/vol0/avidcache /nfs/hosts3/fs7/vol/vol0/avidcache nfs rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0
fs7:/vol/vol0/systems /nfs/hosts3/fs7/vol/vol0/systems nfs rw,v2,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=fs7 0 0

But I don't see:
/vol/vol0

I believe autofs is expiring and unmounting fs7:/vol/vol0 and getting confused as what to do with the directories under that root vol.


Ian Kent wrote:
On Thu, 6 Nov 2003, Mike Waychison wrote:

  
Greg Bradner wrote:

    
Any ideas what 'mount table full' means?
Nov  6 10:08:44 linn automount[27555]: >> NFSv3 not supported!
Nov  6 10:08:44 linn automount[27555]: >> mount table full
Nov  6 10:08:44 linn automount[27555]: mount(nfs): nfs: mount failure 
fs7:/vol/vol0/systems/celia/utils on /celia/utils/
Nov  6 10:09:15 linn automount[1899]: attempting to mount entry 
/misc/hsm_cache
Nov  6 10:09:15 linn automount[27589]: >> NFSv3 not supported!
Nov  6 10:09:15 linn automount[27589]: >> mount table full
Nov  6 10:09:15 linn automount[27589]: mount(nfs): nfs: mount failure 
fs8:/vol/vol0/hsm_cache on /misc/hsm_cache/

Everything works fine after a reboot.

      
It means that you have more than 256 mounts that don't have a 
block-device back-end.  These include NFS, autofs, procfs, sysfs...

This is a known limitation in linux 2.4.  You may want to try a patch 
like the following:

http://people.redhat.com/zaitcev/linux/linux-2.4.19-pre7-unmaj.diff


    

Most of the RedHat packaged kernels have this patch.

What distro are you using?
What version of autofs?
 
  

-- 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rhythm & Hues
5404 Jandy Place
Los Angeles, CA 90066
Voice: 310 448-7763
Fax:   310 448-7600
gregb@rhythm.com
  

-- 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rhythm & Hues
5404 Jandy Place
Los Angeles, CA 90066
Voice: 310 448-7763
Fax:   310 448-7600
gregb@rhythm.com
--------------070804030801050209080403-- --===============19026586840762727== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ autofs mailing list autofs@linux.kernel.org http://linux.kernel.org/mailman/listinfo/autofs --===============19026586840762727==--