From: Ian Kent Subject: Re: more than 1024 mounts ==>hangs Date: Tue, 09 Dec 2008 16:19:05 +0900 Message-ID: <1228807145.3044.63.camel@zeus.themaw.net> References: <4937D794.6040908@itwm.fraunhofer.de> <1228403765.3038.118.camel@zeus.themaw.net> <1228493730.14030.77.camel@zeus.themaw.net> <4939563C.3030405@itwm.fraunhofer.de> <493CF2E0.8000903@itwm.fraunhofer.de> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <493CF2E0.8000903@itwm.fraunhofer.de> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: autofs-bounces@linux.kernel.org Errors-To: autofs-bounces@linux.kernel.org To: Martin Vogt Cc: autofs@linux.kernel.org On Mon, 2008-12-08 at 11:11 +0100, Martin Vogt wrote: > Martin Vogt wrote: > > Ian Kent wrote: > > > >> You would need to enable debug logging. > >> autofs and syslog setup instructions for this can be seen at > >> http://people.redhat.com/jmoyer. > >> > > > > I think I can simply start it with "-d" ?. > > As from the page this logging is meant to be used when > > the daemon ist started without debugging, but I can start > > it directly and can directly hang autofs. > > > > Hm, I think I will write a small script for you, which > > lets you simply reproduce this behaviour. > > > > (A script with 2000 direct bind mount :) > > > > Hello list, > > the following script: > > #!/bin/sh > NET=/net/pxe1/var/tmp > for i in `seq 1 2025` ; do > echo "/homes/vogt${i} -fstype=bind,ro :${NET}" > done > > creates a direct mount map with 2025 direct mount. > (I assume you know how to configure autofs how to use this map :) > (Please really use a remote host for the mounts when you trigger > it, for "localhost" you will find another bug, but currently I'm focused > on the remote part, so really use a remote machine here) What do you mean? Are you saying that entries like "localhost:/net/pxe1/var/tmp", when /net/pxe1/var/tmp is local path are a problem or when the colon escaped path, /net/pxe1/var/tmp refers to another map for the host pxe1 that happens to be the localhost? > > Then run a find . -maxdepth 2 to really trigger the mounts. > These are the bugs: > > - SuSE 11.1 rc1 hangs on the 1024th mount(autofs-5.0.3+patches) > - RHEL52 works (autofs-5.0.1-rc1) on mounting > - RHEL52 has a kernel panic on unmounting (see screenshot attachment) > > (started with: automount -d -f -t 600 ) > > For the kernel panic I started automount in the forground and hit CTRL-C > when it was done with mounting, then after a long time, when it tries > to umount the mounts it gives the kernel panic. > > regards, > > Martin > > > > _______________________________________________ > autofs mailing list > autofs@linux.kernel.org > http://linux.kernel.org/mailman/listinfo/autofs