From owner-autofs@linux.kernel.org Sat Jan 2 08:33:15 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA06351 for ; Sat, 2 Jan 1999 08:33:14 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA27353 for autofs-list; Sat, 2 Jan 1999 06:58:05 -0800 Received: from cse.cuhk.edu.hk (cucs18.cse.cuhk.edu.hk [137.189.91.190]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA27350 for ; Sat, 2 Jan 1999 06:58:02 -0800 From: wongyb@cse.cuhk.edu.hk Received: from solar25 (solar25.cs.cuhk.hk [137.189.88.75]) by cse.cuhk.edu.hk with ESMTP id WAA28086; Sat, 2 Jan 1999 22:57:59 +0800 (HKT) Received: by solar25 (8.8.8+Sun/SMI-SVR4) id WAA21861; Sat, 2 Jan 1999 22:57:59 +0800 (HKT) Message-Id: <199901021457.WAA21861@solar25> Subject: Strange /etc/rc.d/init.d/autofs ?? To: autofs@linux.kernel.org Date: Sat, 2 Jan 1999 22:57:58 +0800 (HKT) X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Dear all, The rc script that comes with autofs touches /var/lock/subsys/automount on startup. On system shutdown, the /etc/rc.d/rc script will look the /var/lock/subsys/ to see what /etc/rc.d/rc?.d/K* that have to run to stop the processes. It is very unfortunate that the name of the K* script (which is K08autofs) is different from that in the /var dir (which is automount). The /usr directory of my system is NFS mounted from a remote server. Failure of shutting down processes that use /usr before umounting /usr generates a lot of NFS complaints. Should /var/lock/subsys/autofs be touched on startup? Thank you very much! -Terence From aoga@leopard.imedia.com Mon Jan 4 05:48:13 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id FAA08320 for ; Mon, 4 Jan 1999 05:48:13 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id FAA32656; Mon, 4 Jan 1999 05:47:24 -0800 From: Alvin Oga Message-Id: <199901041347.FAA32656@leopard.imedia.com> Subject: Re: Starting autofs To: mandrake@www2.netops.com (Isaac Ojeda) Date: Mon, 4 Jan 1999 05:47:24 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com In-Reply-To: from "Isaac Ojeda" at Dec 31, 98 10:03:49 am Content-Type: text Status: RO X-Status: hi ya isaac > Thanks. I got it running. I did not have the autofs active in the kernel > only as a module which still did not work after loading the module. I had > to re-compile the kernel with autofs on. good... have fun alvin > Thanks for the help. > > --------------------------------------------------------------------------- > Isaac Ojeda NetOps Corporation > Staff Engineer 485 Washington Ave > 914-747-7619 Direct Pleasantville, NY 10570 > 914-747-7627 Fax Main: 914-747-7600 > > > On Wed, 30 Dec 1998, Alvin Oga wrote: > > > > > hi isaac > > > > > I am still having problems getting autofs to run. I have setup the maps > > > but they could be wrong. I am running on a 2.0.34 kernel. > > > > > > Am I missing something? Can some email me a map file they use. > > > > try looking thru the examples at: > > http://www.linux-consulting.com/Amd_AutoFS/autofs-HOWTO.html > > > > please let me know if you find it helpful or if you have addtitional > > info to add > > > > have fun > > alvin > > > From owner-autofs@linux.kernel.org Thu Jan 7 15:03:06 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA11775 for ; Thu, 7 Jan 1999 15:03:06 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA23976 for autofs-list; Thu, 7 Jan 1999 14:11:07 -0800 Received: from mailhost3.lanl.gov (mailhost3.lanl.gov [128.165.3.9]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA23971 for ; Thu, 7 Jan 1999 14:11:04 -0800 Received: from nua.lampf.lanl.gov (nua.lampf.lanl.gov [128.165.53.157]) by mailhost3.lanl.gov (8.9.1a/8.9.1/(cic-5, 7/9/98)) with SMTP id PAA13647 for <@lanl.gov:autofs@linux.kernel.org>; Thu, 7 Jan 1999 15:10:59 -0700 Received: from nua.lampf.lanl.gov by nua.lampf.lanl.gov via ESMTP (940816.SGI.8.6.9/940406.SGI.AUTO) for id PAA29196; Thu, 7 Jan 1999 15:11:02 -0700 Message-ID: <369530F6.5119FF1C@nua.lampf.lanl.gov> Date: Thu, 07 Jan 1999 15:11:02 -0700 From: Ben Sapp Reply-To: bsapp@lanl.gov Organization: Los Alamos National Laboratory X-Mailer: Mozilla 4.08 [en] (X11; I; IRIX 5.3 IP22) MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: /net Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello, I saw in TODO list for Version four that it is planned to add support for /net directory like on Sun and SGI workstations. I would like to do what ever I can to help get this done. I have about 1 years worth of experience with C and only nine months of experience with Linux. But, I am eager to help. If some one on this list could tell me how I should procede to help I would be appreciative. Thanks -- Ben Sapp Los Alamos National Laboratory email: Phone: (505)667-3277 Fax: (505)665-7920 URL: http://www.neutrino.lanl.gov/ -- From owner-autofs@linux.kernel.org Thu Jan 7 15:48:48 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA11813 for ; Thu, 7 Jan 1999 15:48:48 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA26250 for autofs-list; Thu, 7 Jan 1999 15:03:55 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA26247 for ; Thu, 7 Jan 1999 15:03:51 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA11828; Thu, 7 Jan 1999 15:03:35 -0800 From: Alvin Oga Message-Id: <199901072303.PAA11828@leopard.imedia.com> Subject: Re: /net To: bsapp@lanl.gov Date: Thu, 7 Jan 1999 15:03:35 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org, alvin@linux-consulting.com In-Reply-To: <369530F6.5119FF1C@nua.lampf.lanl.gov> from "Ben Sapp" at Jan 7, 99 03:11:02 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi ben... > I saw in TODO list for Version four that it is planned to add support > for /net directory like on Sun and SGI workstations. I would like to do > what ever I can to help get this done. I have about 1 years worth of > experience with C and only nine months of experience with Linux. But, I > am eager to help. If some one on this list could tell me how I should > procede to help I would be appreciative. I cheated... to implement /net/hosts/dirs... # # # Repeat these "set of commands" for each machine # # ( assumes you've defined what is exportable on the various hosts ) # m1: vi /etc/auto.master /.autofs /etc/auto.misc --timeout 60 m1: vi /etc/auto.misc m2 -fstype=nfs,soft,intr m2:/ m3 -fstype=nfs,soft,intr m3:/ m4 -fstype=nfs,soft,intr m4:/ m1: mkdir /net m1: cd /net ; ln -s /.autofs/m2 . m1: cd /net ; ln -s /.autofs/m3 . m1: cd /net ; ln -s /.autofs/m4 . m1: cd /net ; ln -s / m1 ( only for the local machine ) Restart autofs # # from any m1, m2, m3, m4 machines, show list of all /etc/auto.misc files # m1: ls -l /net/*/etc/auto.misc m1: ls -l /net/*/etc/passwd ( I expect all passwd file to be identical ) # # diff m2 against m3 # m1: diff /net/m2/etc/auto.misc /net/m3/etc/auto.misc next trick is to run this off a secure nfs network... have fun linuxing alvin http://www.linux-consulting.com/Amd_AutoFS/autofs-HOWTO.html From owner-autofs@linux.kernel.org Thu Jan 7 16:03:10 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA11823 for ; Thu, 7 Jan 1999 16:03:10 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA27470 for autofs-list; Thu, 7 Jan 1999 15:38:53 -0800 Received: from timbuk-e1.cray.com (timbuk-e1.cray.com [128.162.1.30]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA27466 for ; Thu, 7 Jan 1999 15:38:50 -0800 Received: from ledzep.cray.com (ledzep.cray.com [137.38.226.97]) by timbuk-e1.cray.com (8.8.8/CRI-gate-news-1.3) with ESMTP id RAA22492; Thu, 7 Jan 1999 17:38:51 -0600 (CST) Received: from teal.cray.com (teal.cray.com [128.162.188.144]) by ledzep.cray.com (8.8.8/craymail-smart) with ESMTP id RAA8875279; Thu, 7 Jan 1999 17:38:47 -0600 (CST) Received: from sgi.com by teal.cray.com (980427.SGI.8.8.8/SGI-client.1.2) via ESMTP id RAA10404; Thu, 7 Jan 1999 17:38:46 -0600 (CST) Message-ID: <36954586.F7B11FD1@sgi.com> Date: Thu, 07 Jan 1999 17:38:46 -0600 From: Dan Mack Organization: Silicon Graphics, Inc. X-Mailer: Mozilla 4.51C-SGI [en] (X11; I; IRIX64 6.5-ALPHA-1275265720 IP30) X-Accept-Language: en MIME-Version: 1.0 To: bsapp@lanl.gov CC: autofs@linux.kernel.org Subject: Re: /net (or implementing -hosts) References: <369530F6.5119FF1C@nua.lampf.lanl.gov> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Ben Sapp wrote: > > Hello, > > I saw in TODO list for Version four that it is planned to add support > for /net directory like on Sun and SGI workstations. I would like to do > what ever I can to help get this done. I have about 1 years worth of > experience with C and only nine months of experience with Linux. But, I > am eager to help. If some one on this list could tell me how I should > procede to help I would be appreciative. Great! I would like this feature as well and would be willing to help. Note that the top level directory does not have to be /net. The top level hierarchy is arbitrary and is activated by using the -hosts entry in your auto_master file, eg.: /net -hosts -nosuid or /hosts -hosts -nosuid Implementation should not be too hard but it will involve querying the remote machine to get the list of filesystems to mount. Dan -- - -- -- --- --- --- ---- ---- ----- ----- ----- ----- ------ Dan Mack (mack@sgi.com) Information Services 651-683-7288 Silicon Graphics / Eagan, MN - -- --- --- ----- From owner-autofs@linux.kernel.org Thu Jan 7 16:26:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA11858 for ; Thu, 7 Jan 1999 16:26:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA27592 for autofs-list; Thu, 7 Jan 1999 15:41:46 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA27589 for ; Thu, 7 Jan 1999 15:41:45 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA01547; Thu, 7 Jan 1999 15:41:42 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA10361; Thu, 7 Jan 1999 15:41:41 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA07028; Thu, 7 Jan 1999 15:41:41 -0800 Message-Id: <199901072341.PAA07028@cesium.transmeta.com> Subject: Re: /net (or implementing -hosts) In-Reply-To: <36954586.F7B11FD1@sgi.com> from Dan Mack at "Jan 7, 99 05:38:46 pm" To: mack@sgi.com (Dan Mack) Date: Thu, 7 Jan 1999 15:41:40 -0800 (PST) Cc: bsapp@lanl.gov, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: > > Great! I would like this feature as well and would be willing to help. > Note that the top level directory does not have to be /net. The top > level hierarchy is arbitrary and is activated by using the -hosts entry > in your auto_master file, eg.: > > /net -hosts -nosuid > > or > /hosts -hosts -nosuid > > Implementation should not be too hard but it will involve querying the > remote machine to get the list of filesystems to mount. > This is an FAQ. It requires a complete overhaul of the autofs kernel *and* user space code -- it is highly nontrivial, and definitely 2.3 material. -hpa From owner-autofs@linux.kernel.org Fri Jan 15 19:20:21 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA20120 for ; Fri, 15 Jan 1999 19:20:20 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id SAA09731 for autofs-list; Fri, 15 Jan 1999 18:38:13 -0800 Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id SAA09726 for ; Fri, 15 Jan 1999 18:38:07 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id VAA17721; Fri, 15 Jan 1999 21:46:01 -0500 (EST) Message-Id: <199901160246.VAA17721@jubilee.bbn.com> To: autofs@linux.kernel.org Cc: hpa@transmeta.com, Brent.Callaghan@eng.Sun.COM Subject: Sun style YP/NIS maps and autofs Reply-to: frichard@bbn.com Date: Fri, 15 Jan 1999 21:38:02 -0500 From: Fred Richardson Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi- I was directed here to pose the following question about using Sun automounter maps with Linux. I'm currently using RH5.2 with 2.2.0-pre7-ac4 and the "auto.master" maps that we use have the following format: /net/ auto. -ro,hard,intr,noquota,grpid,retrans=4 and then we have a map for each "" called "auto." that looks like this: * .:/nfs//& Now I'm having the problem that the RH5.2 init script spawns an automount job for each line in the auto.master file. That's 117 automount jobs in all... Do you know if I can convert this into something autofs can use? Any help would be greatly appreciated! -Fred From aoga@leopard.imedia.com Fri Jan 15 19:20:15 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA20116 for ; Fri, 15 Jan 1999 19:20:14 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id TAA13784; Fri, 15 Jan 1999 19:19:35 -0800 From: Alvin Oga Message-Id: <199901160319.TAA13784@leopard.imedia.com> Subject: Re: Sun style YP/NIS maps and autofs To: frichard@bbn.com Date: Fri, 15 Jan 1999 19:19:35 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, autofs@linux.kernel.org, Brent.Callaghan@eng.Sun.COM In-Reply-To: <199901160246.VAA17721@jubilee.bbn.com> from "Fred Richardson" at Jan 15, 99 09:38:02 pm Content-Type: text Status: RO X-Status: hi fred assuming that you want to acess from any_linux:/Net//* - I gonna suggest this but donno if it does what you want... - nor do I remember if the options syntax are right or supported ( we can fix it ( the file ) later? ) linux: vi /etc/auto.master /.autofs /etc/auto.misc --timeout 60 - or break up into different auto.misc corresponding files linux: vi /etc/auto.misc host1 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host1:/ host2 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host2:/ host3 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host3:/ hostxx can be host.domain.name... linux: mkdir /.autofs /net linux: cd /net ; ln -s /.autofs/host1 . linux: cd /net ; ln -s /.autofs/host2 . linux: cd /net ; ln -s /.autofs/host3 . linux: /etc/rc.d/init.d/autofs stop linux: /etc/rc.d/init.d/autofs start - or whatever to restart autofs linux: ls -l /net/host*/etc/auto.misc am I way off base ??? at least the above works for me...and is untested on 2.2 kernels... have fun alvin http://www.linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > I was directed here to pose the following question about using Sun > automounter maps with Linux. I'm currently using RH5.2 with > 2.2.0-pre7-ac4 and the "auto.master" maps that we use have the > following format: > > /net/ auto. -ro,hard,intr,noquota,grpid,retrans=4 > > and then we have a map for each "" called "auto." that > looks like this: > > * .:/nfs//& > > Now I'm having the problem that the RH5.2 init script spawns an > automount job for each line in the auto.master file. That's 117 > automount jobs in all... > > Do you know if I can convert this into something autofs can use? Any > help would be greatly appreciated! > > -Fred > From owner-autofs@linux.kernel.org Fri Jan 15 04:27:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id EAA19534 for ; Fri, 15 Jan 1999 04:27:26 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id DAA01228 for autofs-list; Fri, 15 Jan 1999 03:06:15 -0800 Received: from gatekeeper.atml.co.uk (gatekeeper.virata.com [194.129.40.2]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id DAA01210 for ; Fri, 15 Jan 1999 03:06:08 -0800 Received: from boletus.atml.co.uk (root@boletus.atml.co.uk [10.1.3.2]) by gatekeeper.atml.co.uk (8.8.5/8.8.5) with ESMTP id LAA19783 for ; Fri, 15 Jan 1999 11:06:05 GMT Received: from anawara.atml.co.uk (anawara.atml.co.uk [192.168.219.8]) by boletus.atml.co.uk (8.8.5/8.8.5) with ESMTP id LAA19128 for ; Fri, 15 Jan 1999 11:06:05 GMT Received: from inkcap.atml.co.uk (root@inkcap.atml.co.uk [192.168.219.79]) by anawara.atml.co.uk (8.8.8/8.8.5) with ESMTP id LAA18446 for ; Fri, 15 Jan 1999 11:06:04 GMT Received: from virata.com (lgd@localhost [127.0.0.1]) by inkcap.atml.co.uk (8.8.8/8.8.8/Debian/GNU) with ESMTP id LAA01494 for ; Fri, 15 Jan 1999 11:06:03 GMT Message-ID: <369F211B.922E9762@virata.com> Date: Fri, 15 Jan 1999 11:06:03 +0000 From: Luke Diamand Reply-To: Luke Diamand X-Mailer: Mozilla 4.5 [en] (X11; I; Linux 2.2.0-pre7 i686) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Direct mount points References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi! I apologise if this is a FAQ, but can anyone tell if it's possible to get autofs to do direct maps? There seems to be something about it in the HOWTO, but it didn't seem to work for me. Regards! Luke Diamand -- Virata http://www.virata.com Cambridge Tel: +44 1223 566919 Fax: +44 1223 566915 From frichard@cyan.bbn.com Fri Jan 15 19:38:52 1999 Return-Path: Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA20150 for ; Fri, 15 Jan 1999 19:38:51 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id WAA18220; Fri, 15 Jan 1999 22:44:55 -0500 (EST) Message-Id: <199901160344.WAA18220@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com, aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, autofs@linux.kernel.org, Brent.Callaghan@eng.Sun.COM Subject: Re: Sun style YP/NIS maps and autofs In-reply-to: Your message of "Fri, 15 Jan 1999 19:19:35 PST." <199901160319.TAA13784@leopard.imedia.com> Reply-to: frichard@bbn.com Date: Fri, 15 Jan 1999 22:36:56 -0500 From: Fred Richardson Status: RO X-Status: Hi Alvin- Thanks for the quick response! Actually, this isn't quite it. I don't want to mount (or export) the root from each machine. Instead, by some ancient ad-hoc convention, each file system is exported as: /nfs//u?? So the mount entery looks like: /nfs//u?? .:/nfs//u?? I think the Sun/Solaris/SGI boxes have been accessing these using the `multiple mounts' feature of the Sun and SGI automounters (I just figured this out by reading the Solaris automounter man page more carefully). I'm pretty sure autofs-3.1.1 doesn't have this feature yet. The TODO file indicates that they are slated for autofs-4.x. -Fred From aoga@leopard.imedia.com Fri Jan 15 19:49:40 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA20156 for ; Fri, 15 Jan 1999 19:49:40 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id TAA13992; Fri, 15 Jan 1999 19:49:03 -0800 From: Alvin Oga Message-Id: <199901160349.TAA13992@leopard.imedia.com> Subject: Re: Sun style YP/NIS maps and autofs (fwd) To: frichard@bbn.com Date: Fri, 15 Jan 1999 19:49:03 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, Brent.Callaghan@eng.Sun.COM Content-Type: text Status: RO X-Status: hi fred merging two emails together... fred> Thanks for the quick response! Actually, this isn't quite it. I fred> don't want to mount (or export) the root from each machine. Instead, fred> by some ancient ad-hoc convention, each file system is exported as: fred> fred> /nfs//u?? fred> fred> So the mount entery looks like: fred> fred> /nfs//u?? .:/nfs//u?? fred> changed my example to show uxx insted... > linux: vi /etc/auto.master > /.autofs /etc/auto.misc --timeout 60 > > - or break up into different auto.misc corresponding files > > linux: vi /etc/auto.misc host1.uxx -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host1:/uxx host2.uyy -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host2:/uyy host3.uzz -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host3:/uzz > > hostxx can be host.domain.name... > > linux: mkdir /.autofs /net linux: cd /net ; ln -s /.autofs/host1.uxx . linux: cd /net ; ln -s /.autofs/host2.uyy . linux: cd /net ; ln -s /.autofs/host3.uzz . > > linux: /etc/rc.d/init.d/autofs stop > linux: /etc/rc.d/init.d/autofs start - or whatever to restart autofs > > linux: ls -l /net/host*/stuff_in_u?? have fun alvin > > > > /net/ auto. -ro,hard,intr,noquota,grpid,retrans=4 > > > > and then we have a map for each "" called "auto." that > > looks like this: > > > > * .:/nfs//& > > > > Now I'm having the problem that the RH5.2 init script spawns an > > automount job for each line in the auto.master file. That's 117 > > automount jobs in all... > > > > Do you know if I can convert this into something autofs can use? Any > > help would be greatly appreciated! > > > > -Fred From owner-autofs@linux.kernel.org Fri Jan 15 20:01:50 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id UAA20180 for ; Fri, 15 Jan 1999 20:01:48 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id TAA11551 for autofs-list; Fri, 15 Jan 1999 19:19:48 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id TAA11548 for ; Fri, 15 Jan 1999 19:19:44 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id TAA13784; Fri, 15 Jan 1999 19:19:35 -0800 From: Alvin Oga Message-Id: <199901160319.TAA13784@leopard.imedia.com> Subject: Re: Sun style YP/NIS maps and autofs To: frichard@bbn.com Date: Fri, 15 Jan 1999 19:19:35 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, autofs@linux.kernel.org, Brent.Callaghan@eng.Sun.COM In-Reply-To: <199901160246.VAA17721@jubilee.bbn.com> from "Fred Richardson" at Jan 15, 99 09:38:02 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi fred assuming that you want to acess from any_linux:/Net//* - I gonna suggest this but donno if it does what you want... - nor do I remember if the options syntax are right or supported ( we can fix it ( the file ) later? ) linux: vi /etc/auto.master /.autofs /etc/auto.misc --timeout 60 - or break up into different auto.misc corresponding files linux: vi /etc/auto.misc host1 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host1:/ host2 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host2:/ host3 -fstype=nfs,ro,hard,intr,noquota,grpid,retrans=4 host3:/ hostxx can be host.domain.name... linux: mkdir /.autofs /net linux: cd /net ; ln -s /.autofs/host1 . linux: cd /net ; ln -s /.autofs/host2 . linux: cd /net ; ln -s /.autofs/host3 . linux: /etc/rc.d/init.d/autofs stop linux: /etc/rc.d/init.d/autofs start - or whatever to restart autofs linux: ls -l /net/host*/etc/auto.misc am I way off base ??? at least the above works for me...and is untested on 2.2 kernels... have fun alvin http://www.linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > I was directed here to pose the following question about using Sun > automounter maps with Linux. I'm currently using RH5.2 with > 2.2.0-pre7-ac4 and the "auto.master" maps that we use have the > following format: > > /net/ auto. -ro,hard,intr,noquota,grpid,retrans=4 > > and then we have a map for each "" called "auto." that > looks like this: > > * .:/nfs//& > > Now I'm having the problem that the RH5.2 init script spawns an > automount job for each line in the auto.master file. That's 117 > automount jobs in all... > > Do you know if I can convert this into something autofs can use? Any > help would be greatly appreciated! > > -Fred > From frichard@cyan.bbn.com Fri Jan 15 20:08:27 1999 Return-Path: Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id UAA20191 for ; Fri, 15 Jan 1999 20:08:27 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id XAA18377; Fri, 15 Jan 1999 23:14:31 -0500 (EST) Message-Id: <199901160414.XAA18377@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com, aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, Brent.Callaghan@eng.Sun.COM Subject: Re: Sun style YP/NIS maps and autofs (fwd) In-reply-to: Your message of "Fri, 15 Jan 1999 19:49:03 PST." <199901160349.TAA13992@leopard.imedia.com> Reply-to: frichard@bbn.com Date: Fri, 15 Jan 1999 23:06:33 -0500 From: Fred Richardson Status: RO X-Status: Alvin- Does there have to be an entery for each partition? Currently we use a wild card for that, so the expanded `multiple mounts' map entery would be something like this: /net/sycamore * -port=2049 sycamore.bbn.com:/nfs/sycamore/& -rw,hard,intr,noquota,grpid,timeo=40,retrans=0 This allows me to mount /nfs/sycamore/u10 or /nfs/sycamore/u20 without actually putting those partitions explicitly in the mount map. Of course, it's easy to specify a partition that doesn't exist. Can the wild card be used like this? /etc/auto.master: /net /etc/auto.misc --timeout 60 /etc/auto.misc: sycamore/* sycamore.bbn.com:/nfs/sycamore/& -fstype=nfs,port=2049,rw,hard,intr,noquota,grpid,timeo=40,retrans=0 From aoga@leopard.imedia.com Fri Jan 15 20:14:53 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id UAA20195 for ; Fri, 15 Jan 1999 20:14:53 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id UAA14097; Fri, 15 Jan 1999 20:14:17 -0800 From: Alvin Oga Message-Id: <199901160414.UAA14097@leopard.imedia.com> Subject: Re: Sun style YP/NIS maps and autofs (fwd) To: frichard@bbn.com Date: Fri, 15 Jan 1999 20:14:17 -0800 (PST) Cc: alvin@linux-consulting.com, aoga@leopard.imedia.com (Alvin Oga), Brent.Callaghan@eng.Sun.COM In-Reply-To: <199901160414.XAA18377@jubilee.bbn.com> from "Fred Richardson" at Jan 15, 99 11:06:33 pm Content-Type: text Status: RO X-Status: hi fred > Does there have to be an entery for each partition? Currently we use > a wild card for that, so the expanded `multiple mounts' map entery > would be something like this: > > /net/sycamore * -port=2049 sycamore.bbn.com:/nfs/sycamore/& -rw,hard,intr,noquota,grpid,timeo=40,retrans=0 > > This allows me to mount /nfs/sycamore/u10 or /nfs/sycamore/u20 without > actually putting those partitions explicitly in the mount map. Of > course, it's easy to specify a partition that doesn't exist. > > Can the wild card be used like this? > > /etc/auto.master: > /net /etc/auto.misc --timeout 60 > > /etc/auto.misc: > sycamore/* sycamore.bbn.com:/nfs/sycamore/& -fstype=nfs,port=2049,rw,hard,intr,noquota,grpid,timeo=40,retrans=0 ^^^^^^^^^^^^^^^ the & is NOT needed to see all the "directories" not sure that you can use * in "sycamore/*" and is not needed for autofs if that * represents mounts & directories on sycamore the above using autofs will create a mountpoint in /net/sycamore that will be dynamically managed/created/deleted by autofs you cannot use an autofs mount point as your "access" - or at least you're not supposed to... REMEMBER....tht /net is your autofs mountpoint ( key )... so I use /.autofs ( due to autofs ) like /.amd for the amd automounter and yes... I do use: ls -l /.autofs/sycamore/* ls -l /Net/sycamore/* where sycamore is a symlink to /.autofs in /Net have fun alvin From aoga@leopard.imedia.com Fri Jan 15 22:28:27 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id WAA20313 for ; Fri, 15 Jan 1999 22:28:27 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id WAA16485; Fri, 15 Jan 1999 22:27:51 -0800 From: Alvin Oga Message-Id: <199901160627.WAA16485@leopard.imedia.com> Subject: Re: Sun style YP/NIS maps and autofs (fwd) To: frichard@bbn.com Date: Fri, 15 Jan 1999 22:27:50 -0800 (PST) Cc: alvin@linux-consulting.com, aoga@leopard.imedia.com (Alvin Oga) In-Reply-To: <199901160535.AAA18903@jubilee.bbn.com> from "Fred Richardson" at Jan 16, 99 00:27:54 am Content-Type: text Status: RO X-Status: hi ya fred > Thanks for all your help, but it's still no good... > > I tried what you suggested: > > mkdir /.autofs > mkdir /nfs > cd /nfs; ln -s /.autofs/sycamore sycamore > > /etc/auto.master: > /.autofs /etc/auto.misc --timeout 60 > > /etc/auto.misc: > sycamore sycamore.bbn.com:/nfs/sycamore -fstype=nfs,port=2049,rw,hard,intr,noquota,grpid,timeo=40,retrans=0 That should be reversed ----------------------- sycamore -fstype=nfs,port=2049,rw,hard,intr,noquota,grpid,timeo=40,retrans=0 sycamore.bbn.com:/nfs/sycamore > The problem is that sycamore doesn't export /nfs/sycamore. It exports > /nfs/sycamore/u??. So, autofs can't resolve this. # # On sycamore, you can only exports files LOCAL to sycamore # sycamore: vi /etc/exports /uxx *.bbn.com(ro) sycamore: /etc/rc.d/init.d/nfs stop ; /etc/rc.d/init.d/nfs start now you should be able to mount it: # # try it manually....if you cannot mount it manually, autofs will fail too # linux# mount sycamore.bbn.com:/uxx /mnt linux# ls -l /mnt should give you want you want # # If the above manual mount and ls gives you want you want.. # autofs will work too # have fun alvin > What I need to do is extend the key in auto.master somehow: > > /etc/auto.master: > /net/sycamore/* sycamore.bbn.com:/nfs/sycamore/& -fstype=nfs,port=2049.... > > So the mount point becomes /net/sycamore for any partition on > sycamore.bbn.com:/nfs/sycamore/. Then I'd have to do this: > > mkdir /net/sycamore > mkdir /nfs; cd /nfs; ln -s sycamore /net/sycamore > > I think the Sun `multiple mounts' feature is designed to let you do > this. I'm not sure I can emulate it with autofs (3.1.1). > > Thanks again, > > -Fred > From owner-autofs@linux.kernel.org Fri Jan 15 20:23:42 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id UAA20232 for ; Fri, 15 Jan 1999 20:23:41 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id TAA12294 for autofs-list; Fri, 15 Jan 1999 19:38:24 -0800 Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id TAA12290 for ; Fri, 15 Jan 1999 19:38:21 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id WAA18220; Fri, 15 Jan 1999 22:44:55 -0500 (EST) Message-Id: <199901160344.WAA18220@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com, aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, autofs@linux.kernel.org, Brent.Callaghan@eng.Sun.COM Subject: Re: Sun style YP/NIS maps and autofs In-reply-to: Your message of "Fri, 15 Jan 1999 19:19:35 PST." <199901160319.TAA13784@leopard.imedia.com> Reply-to: frichard@bbn.com Date: Fri, 15 Jan 1999 22:36:56 -0500 From: Fred Richardson Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi Alvin- Thanks for the quick response! Actually, this isn't quite it. I don't want to mount (or export) the root from each machine. Instead, by some ancient ad-hoc convention, each file system is exported as: /nfs//u?? So the mount entery looks like: /nfs//u?? .:/nfs//u?? I think the Sun/Solaris/SGI boxes have been accessing these using the `multiple mounts' feature of the Sun and SGI automounters (I just figured this out by reading the Solaris automounter man page more carefully). I'm pretty sure autofs-3.1.1 doesn't have this feature yet. The TODO file indicates that they are slated for autofs-4.x. -Fred From frichard@cyan.bbn.com Fri Jan 15 21:28:34 1999 Return-Path: Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id VAA20272 for ; Fri, 15 Jan 1999 21:28:34 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id AAA18903; Sat, 16 Jan 1999 00:35:53 -0500 (EST) Message-Id: <199901160535.AAA18903@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com, alvin@linux-consulting.com, aoga@leopard.imedia.com (Alvin Oga), Brent.Callaghan@eng.Sun.COM Subject: Re: Sun style YP/NIS maps and autofs (fwd) In-reply-to: Your message of "Fri, 15 Jan 1999 20:14:17 PST." <199901160414.UAA14097@leopard.imedia.com> Reply-to: frichard@bbn.com Date: Sat, 16 Jan 1999 00:27:54 -0500 From: Fred Richardson Status: RO X-Status: Hi Alvin- Thanks for all your help, but it's still no good... I tried what you suggested: mkdir /.autofs mkdir /nfs cd /nfs; ln -s /.autofs/sycamore sycamore /etc/auto.master: /.autofs /etc/auto.misc --timeout 60 /etc/auto.misc: sycamore sycamore.bbn.com:/nfs/sycamore -fstype=nfs,port=2049,rw,hard,intr,noquota,grpid,timeo=40,retrans=0 The problem is that sycamore doesn't export /nfs/sycamore. It exports /nfs/sycamore/u??. So, autofs can't resolve this. What I need to do is extend the key in auto.master somehow: /etc/auto.master: /net/sycamore/* sycamore.bbn.com:/nfs/sycamore/& -fstype=nfs,port=2049.... So the mount point becomes /net/sycamore for any partition on sycamore.bbn.com:/nfs/sycamore/. Then I'd have to do this: mkdir /net/sycamore mkdir /nfs; cd /nfs; ln -s sycamore /net/sycamore I think the Sun `multiple mounts' feature is designed to let you do this. I'm not sure I can emulate it with autofs (3.1.1). Thanks again, -Fred From frichard@cyan.bbn.com Sat Jan 16 06:33:14 1999 Return-Path: Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id GAA20730 for ; Sat, 16 Jan 1999 06:33:13 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id JAA21523; Sat, 16 Jan 1999 09:40:32 -0500 (EST) Message-Id: <199901161440.JAA21523@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com, alvin@linux-consulting.com, aoga@leopard.imedia.com (Alvin Oga) Subject: Re: Sun style YP/NIS maps and autofs (fwd) In-reply-to: Your message of "Fri, 15 Jan 1999 22:27:50 PST." <199901160627.WAA16485@leopard.imedia.com> Reply-to: frichard@bbn.com Date: Sat, 16 Jan 1999 09:32:33 -0500 From: Fred Richardson Status: RO X-Status: Hi Alvin- Again, thanks for the help! Actually, in our setup, sycamore has these filesystems mounted from local disks: /dev/... /nfs/sycamore/u?? and sycamore is configured to export /nfs/sycamore/u??. All the other directories on sycamore under /nfs/* are links to automount points and are not exported. But, sycamore doesn't export /nfs/sycamore. So when I try the setup we were talking about (making the change you pointed out by reversing the options and nfs mount points), I get this complaint: automount[6257]: attempting to mount entry /.autofs/sycamore automount[6263]: >> mount: sycamore.bbn.com:/nfs/sycamore failed, reason given by server: Permission denied automount[6263]: mount(nfs): nfs: mount failure sycamore.bbn.com:/nfs/sycamore on /.autofs/sycamore So I'm telling automount to do the wrong thing. What I really want to do is extend the key so that /.autofs/sycamore will be a directory used as a mount point for /.autofs/sycamore/u??. I think this is why we use /net/sycamore: they're directories that we set up and the automounter can use. -Fred From owner-autofs@linux.kernel.org Sat Jan 16 20:11:14 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id UAA21215 for ; Sat, 16 Jan 1999 20:11:14 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id TAA32657 for autofs-list; Sat, 16 Jan 1999 19:37:55 -0800 Received: from snowy.nsw.cmis.CSIRO.AU (snowy.nsw.cmis.CSIRO.AU [130.155.16.108]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id TAA32654 for ; Sat, 16 Jan 1999 19:37:51 -0800 Received: by snowy.nsw.cmis.CSIRO.AU (SMI-8.6/SMI-SVR4) id OAA07077; Sun, 17 Jan 1999 14:37:48 +1100 Received: from edb.home.nsw.cmis.CSIRO.AU(130.155.16.90), claiming to be "vodka.gnarus.com" via SMTP by snowy.nsw.cmis.CSIRO.AU, id smtpdAAAa001iZ; Sun Jan 17 14:37:45 1999 Received: from Altavista.net (really [192.168.110.12]) by edb.home.dms.csiro.au via smail with esmtp (ident edb using rfc1413) id (Debian Smail3.2.0.101) for ; Sun, 17 Jan 1999 14:37:41 +1100 (EST) Message-ID: <36A15AE0.44AD4AA6@Altavista.net> Date: Sun, 17 Jan 1999 03:37:04 +0000 From: Ed Breen Organization: eic_free X-Mailer: Mozilla 4.5 [en] (X11; I; Linux 2.0.34 i686) MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: autofs woes: fail to mount home directories Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi I have tried using autofs to automate the mounting of home directories, but it is not working as expected. Clearly I am missing something, but I don't know what? On the NIS server (kernel 2.0.34), wax, I have autofs running. # cat /proc/filesystems ... nodev autofs # /sbin/lsmod Module Pages Used by autofs 2 1 Wax's auto.master file contains other than comments: # Sample auto.master file /home /etc/auto.home The auto.home file is, missing comments: exdb -rw,hard,intr grun:/home2/exdb In wax's nsswitch file there is: automount: files nis Wax's export file has: /home grun(rw) And I have included auto.home and auto.master in /var/yp/Makefile. I can log in on wax as exdb and the /home/exdb exists and every thing is fine. However, while I can log into grun (kernel 2.0.36) as exdb, exdb home directory fails to mount -- and this is crux of my problem. On grun: # cat /proc/filesystems ... nodev autofs # /sbin/lsmod Module Pages Used by autofs 2 1 # cat /etc/nsswitch.conf ... automount: nis # cat /etc/auto.master +auto.master /home auto.home --timeout 60 # cat /etc/auto.home +auto.home # cat /etc/exports ... /home2 wax(rw) >From the log file /var/log/syslog on grun: Jan 17 11:44:16 grun automount[7194]: attempting to mount entry /home/exdb Jan 17 11:44:16 grun automount[7323]: lookup(file): looking up exdb Jan 17 11:44:16 grun automount[7323]: lookup(file): lookup for exdb failed So it seems to fail to find exdb, but why?? Remember I can log in as exdb on grun. exdb's passwd/shadow is handled using NIS on wax. When I expect wax's /var/log/syslog I find: Jan 17 19:44:54 wax automount[5095]: lookup(file): looking up edb Jan 17 19:44:54 wax automount[5095]: lookup(file): edb -> -rw,hard,intr^I grun:/home2/exdb Jan 17 19:44:54 wax automount[5095]: expanded entry: -rw,hard,intr^I grun:/home2/exdb Jan 17 19:44:54 wax automount[5095]: parse(sun): gathered options: rw,hard,intr Jan 17 19:44:54 wax automount[5095]: parse(sun): core of entry: grun:/home2/exdb Jan 17 19:44:54 wax automount[5095]: mount(nfs): calling mkdir /home/exdb Jan 17 19:44:54 wax automount[5095]: mount(nfs): calling mount -t nfs -o rw,hard,intr grun:/home2/exdb /home/exdb Jan 17 19:44:54 wax automount[5095]: mount(nfs): mounted grun:/home2/exdb on /home/exdb Any help would be greatly appreciated -- this problem is really slowing things up. Also amd is installed but not used on the server wax, but it is not installed on grun, should I install amd on grun also? Ed. From owner-autofs@linux.kernel.org Sun Jan 17 02:15:20 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA21466 for ; Sun, 17 Jan 1999 02:15:20 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA14384 for autofs-list; Sun, 17 Jan 1999 01:39:24 -0800 Received: from imag.imag.fr (imag.imag.fr [129.88.30.1]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA14381 for ; Sun, 17 Jan 1999 01:39:21 -0800 Received: from simian.localdomain (ppp-habraken.imag.fr [129.88.13.43]) by imag.imag.fr (8.8.5/8.8.5) with ESMTP id KAA09428; Sun, 17 Jan 1999 10:39:17 +0100 (MET) Received: from imag.fr (localhost [127.0.0.1]) by simian.localdomain (8.8.7/8.8.7) with ESMTP id KAA00839; Sun, 17 Jan 1999 10:37:42 +0100 Message-ID: <36A1AF65.C2D250DA@imag.fr> Date: Sun, 17 Jan 1999 10:37:41 +0100 From: Pierre Habraken Organization: Université Joseph Fourier X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.0.36 i586) MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Mouting local file system References: <199901160246.VAA17721@jubilee.bbn.com> Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello, I would like to use autofs for mounting local vfat file systems residing on removable media such as floppies or zip disks. Once mounted the file system should be seen as owned by the user who attempted to access this file system, not by the super user. Without autofs, the above result is obtained by putting the option 'suid' in /etc/fstab, like in the following entry: /dev/fd0 /mnt/floppy vfat user,noauto,suid 0 0 Expecting to obtain the same result with autofs, I put the following entry in /etc/auto.misc: floppy -fstype=vfat,suid :/dev/fd0 However, when an ordinary user executes 'ls -l /misc/floppy' all files and directories on the floppy are listed as owned by root, not by this user. Apparently autofs executes mount with the real id. of root, where I would expect that it executes mount with the real id. of the end user. Am I right ? Does anybody know how to obtain the right behavior (file system mounted as owned by the user not by root) ? Thanks in advance for any advice. Pierre -- ________________________________________________________________________ Pierre HABRAKEN - mailto:Pierre.Habraken@imag.fr Tél: 04 76 82 72 48 - Fax: 04 76 82 72 87 IMAG-LSR BP72 38402 SAINT MARTIN D'HERES Cedex ________________________________________________________________________ From owner-autofs@linux.kernel.org Sun Jan 17 10:11:43 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA21953 for ; Sun, 17 Jan 1999 10:11:43 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA30935 for autofs-list; Sun, 17 Jan 1999 09:27:26 -0800 Received: from ifi.informatik.uni-stuttgart.de (ifi.informatik.uni-stuttgart.de [129.69.211.1]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA30932 for ; Sun, 17 Jan 1999 09:27:23 -0800 Received: by trick.informatik.uni-stuttgart.de with SMTP; Sun, 17 Jan 1999 18:27:46 +0100 (MET) Date: Sun, 17 Jan 1999 18:28:02 +0100 (MET) From: Siegfried Langauf X-Sender: langausd@dagobert.informatik.uni-stuttgart.de To: Pierre Habraken cc: autofs@linux.kernel.org Subject: Re: Mouting local file system In-Reply-To: <36A1AF65.C2D250DA@imag.fr> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: On Sun, 17 Jan 1999, Pierre Habraken wrote: > I would like to use autofs for mounting local vfat file systems residing > on removable media such as floppies or zip disks. generally, supermount is more suitable for this purpose. > Once mounted the file system should be seen as owned by the user who > attempted to access this file system, not by the super user. > > Without autofs, the above result is obtained by putting the option > 'suid' in /etc/fstab, like in the following entry: > > /dev/fd0 /mnt/floppy vfat user,noauto,suid 0 0 actually, it is the "user" option that allows any user to mount the file system with their own permissions. "suid" allows SUID executables to be executed from the mounted filesystem which is a huge security hole, as anybody could become root by simply mounting a floppy with SUID root shell un it... > Expecting to obtain the same result with autofs, I put the following > entry in /etc/auto.misc: > > floppy -fstype=vfat,suid :/dev/fd0 > > However, when an ordinary user executes 'ls -l /misc/floppy' all files > and directories on the floppy are listed as owned by root, not by this > user. > > Apparently autofs executes mount with the real id. of root, where I > would expect that it executes mount with the real id. of the end user. > > Am I right ? Does anybody know how to obtain the right behavior (file > system mounted as owned by the user not by root) ? AFAIR, there is no simple way to achieve this via autofs, as the automountd is running as root and it has no chance to ever get to know who accessed the mount point. ------------------------------------------------------------------------ Siegfried Langauf Student der Softwaretechnik, Uni Stuttgart "If one is master of one thing and understands one thing well, one has at the same time, insight into and understanding of many things." -- Van Gogh ------------------------------------------------------------------------ From owner-autofs@linux.kernel.org Sun Jan 17 15:01:25 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA22132 for ; Sun, 17 Jan 1999 15:01:25 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA10593 for autofs-list; Sun, 17 Jan 1999 14:24:12 -0800 Received: from fate.eng.buffalo.edu (fate.eng.buffalo.edu [128.205.25.5]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA10590 for ; Sun, 17 Jan 1999 14:24:10 -0800 Received: from volta.eng.buffalo.edu (volta.eng.buffalo.edu [128.205.28.60]) by fate.eng.buffalo.edu (8.9.0/8.9.0) with ESMTP id RAA23997; Sun, 17 Jan 1999 17:24:08 -0500 (EST) Received: from localhost by volta.eng.buffalo.edu (8.9.0/8.9.0) with ESMTP id RAA01455; Sun, 17 Jan 1999 17:24:05 -0500 (EST) X-Authentication-Warning: volta.eng.buffalo.edu: senthil owned process doing -bs Date: Sun, 17 Jan 1999 17:24:05 -0500 (EST) From: Senthil Kumar Narayanasamy To: Pierre Habraken cc: autofs@linux.kernel.org Subject: Re: Mouting local file system In-Reply-To: <36A1AF65.C2D250DA@imag.fr> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by linux.kernel.org id OAA10591 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: On Sun, 17 Jan 1999, Pierre Habraken wrote: hi , This regards to my previous post dont use suid or exec option to mount a removable media.probably you know why regards, senthil #Hello, # #I would like to use autofs for mounting local vfat file systems residing #on removable media such as floppies or zip disks. # #Once mounted the file system should be seen as owned by the user who #attempted to access this file system, not by the super user. # #Without autofs, the above result is obtained by putting the option #'suid' in /etc/fstab, like in the following entry: # #/dev/fd0 /mnt/floppy vfat user,noauto,suid 0 0 # #Expecting to obtain the same result with autofs, I put the following #entry in /etc/auto.misc: # #floppy -fstype=vfat,suid :/dev/fd0 # #However, when an ordinary user executes 'ls -l /misc/floppy' all files #and directories on the floppy are listed as owned by root, not by this #user. # #Apparently autofs executes mount with the real id. of root, where I #would expect that it executes mount with the real id. of the end user. # #Am I right ? Does anybody know how to obtain the right behavior (file #system mounted as owned by the user not by root) ? # #Thanks in advance for any advice. # #Pierre #-- #________________________________________________________________________ #Pierre HABRAKEN - mailto:Pierre.Habraken@imag.fr #Tél: 04 76 82 72 48 - Fax: 04 76 82 72 87 #IMAG-LSR BP72 38402 SAINT MARTIN D'HERES Cedex #________________________________________________________________________ # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Senthil Kumar Narayanasamy http://www.eng.buffalo.edu/~senthil ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Home: Office: 1525,Amherst Manor Drive, #804, 809, Furnas Hall Williamsville, University at Buffalo NY, 14221. Buffalo, NY 14260 Ph: (716)632-3135 Ph:(716)645 2593 x2260 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From owner-autofs@linux.kernel.org Sun Jan 17 15:04:38 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA22136 for ; Sun, 17 Jan 1999 15:04:37 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA10402 for autofs-list; Sun, 17 Jan 1999 14:21:37 -0800 Received: from fate.eng.buffalo.edu (fate.eng.buffalo.edu [128.205.25.5]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA10398 for ; Sun, 17 Jan 1999 14:21:34 -0800 Received: from volta.eng.buffalo.edu (volta.eng.buffalo.edu [128.205.28.60]) by fate.eng.buffalo.edu (8.9.0/8.9.0) with ESMTP id RAA23869; Sun, 17 Jan 1999 17:21:31 -0500 (EST) Received: from localhost by volta.eng.buffalo.edu (8.9.0/8.9.0) with ESMTP id RAA01450; Sun, 17 Jan 1999 17:21:28 -0500 (EST) X-Authentication-Warning: volta.eng.buffalo.edu: senthil owned process doing -bs Date: Sun, 17 Jan 1999 17:21:28 -0500 (EST) From: Senthil Kumar Narayanasamy To: Pierre Habraken cc: autofs@linux.kernel.org Subject: Re: Mouting local file system In-Reply-To: <36A1AF65.C2D250DA@imag.fr> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by linux.kernel.org id OAA10399 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: On Sun, 17 Jan 1999, Pierre Habraken wrote: #Hello, # #I would like to use autofs for mounting local vfat file systems residing #on removable media such as floppies or zip disks. # #Once mounted the file system should be seen as owned by the user who #attempted to access this file system, not by the super user. # #Without autofs, the above result is obtained by putting the option #'suid' in /etc/fstab, like in the following entry: # #/dev/fd0 /mnt/floppy vfat user,noauto,suid 0 0 # #Expecting to obtain the same result with autofs, I put the following #entry in /etc/auto.misc: # #floppy -fstype=vfat,suid :/dev/fd0 # #However, when an ordinary user executes 'ls -l /misc/floppy' all files #and directories on the floppy are listed as owned by root, not by this #user. Hi , I dont know why you want the files in a removable media to be owned by a user. If you concern is to write into removable media , set proper umask value your /etc/auto.misc should contain floppy -fstype=vfat,umask=000 :/dev/fd0 note that this umask read.write,execute permission for user,group.and others. IF yours is a single user system you can set uid= and have rw permissions. If yours is a multi user system it is better to play with umask settings . you can also specify "sync" option to immediate write All these options are documented in manual pages. hope that helps regards, senthil # #Apparently autofs executes mount with the real id. of root, where I #would expect that it executes mount with the real id. of the end user. # #Am I right ? Does anybody know how to obtain the right behavior (file #system mounted as owned by the user not by root) ? # #Thanks in advance for any advice. # #Pierre #-- #________________________________________________________________________ #Pierre HABRAKEN - mailto:Pierre.Habraken@imag.fr #Tél: 04 76 82 72 48 - Fax: 04 76 82 72 87 #IMAG-LSR BP72 38402 SAINT MARTIN D'HERES Cedex #________________________________________________________________________ # ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Senthil Kumar Narayanasamy http://www.eng.buffalo.edu/~senthil ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Home: Office: 1525,Amherst Manor Drive, #804, 809, Furnas Hall Williamsville, University at Buffalo NY, 14221. Buffalo, NY 14260 Ph: (716)632-3135 Ph:(716)645 2593 x2260 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From alvin Mon Jan 18 01:44:57 1999 Return-Path: Received: (from alvin@localhost) by mail.Linux-Consulting.com (8.8.7/8.8.7) id BAA22542; Mon, 18 Jan 1999 01:44:47 -0800 From: Alvin Oga Message-Id: <199901180944.BAA22542@mail.Linux-Consulting.com> Subject: Re: Sun style YP/NIS maps and autofs (fwd) To: frichard@bbn.com Date: Mon, 18 Jan 1999 01:44:47 -0800 (PST) Cc: alvin@mail.Linux-Consulting.com (Alvin Oga) In-Reply-To: <199901161440.JAA21523@jubilee.bbn.com> from "Fred Richardson" at Jan 16, 99 09:32:33 am Content-Type: text Status: RO X-Status: hi ya fred the error message from autofs automounter below is correct if that directory is not exported ( sycamore:/etc/export ) from syncamore.bbn.com when sycamore is properly exporting it... you should be manually able to mount it... if manual mounting fails...autofs will fail # # this must work # linux: mount sycamore:/u?? /mnt except directory names and file names cannot be ?? in unix have fun alvin > Again, thanks for the help! Actually, in our setup, sycamore has > these filesystems mounted from local disks: > > /dev/... /nfs/sycamore/u?? > > and sycamore is configured to export /nfs/sycamore/u??. All the other > directories on sycamore under /nfs/* are links to automount points and > are not exported. > > But, sycamore doesn't export /nfs/sycamore. So when I try the setup > we were talking about (making the change you pointed out by reversing > the options and nfs mount points), I get this complaint: > > automount[6257]: attempting to mount entry /.autofs/sycamore > automount[6263]: >> mount: sycamore.bbn.com:/nfs/sycamore failed, reason given by server: Permission denied > automount[6263]: mount(nfs): nfs: mount failure sycamore.bbn.com:/nfs/sycamore on /.autofs/sycamore > > So I'm telling automount to do the wrong thing. What I really want to > do is extend the key so that /.autofs/sycamore will be a directory > used as a mount point for /.autofs/sycamore/u??. I think this is why > we use /net/sycamore: they're directories that we set up and the > automounter can use. > > -Fred > From owner-autofs@linux.kernel.org Mon Jan 18 02:36:23 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA22592 for ; Mon, 18 Jan 1999 02:36:23 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA08423 for autofs-list; Mon, 18 Jan 1999 01:44:59 -0800 Received: from lcnis.sorosis.ro (IDENT:tvlad@lcnis.sorosis.ro [193.226.103.240]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA07982 for ; Mon, 18 Jan 1999 01:35:26 -0800 Received: from localhost (tvlad@localhost) by lcnis.sorosis.ro (8.8.7/8.8.7) with ESMTP id LAA00914 for ; Mon, 18 Jan 1999 11:28:38 +0200 Date: Mon, 18 Jan 1999 11:28:38 +0200 (EET) From: Vlad Tautu To: autofs@linux.kernel.org Subject: autofs patch for cdrom jukeboxes (cdchangers) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I wrote a quick module for cdrom changers. For example, to use my 2nd cdrom in the jukebox, I type something like `cd /cdrom/2`, and the CD's in drive will be automatically switched. I included the auto.master and auto.cdrom files in the tgz file. Download it from: http://home.dntis.ro/~tvlad/proj/autofs/autofs_cdchanger_patch.tgz and get only the files new in package, or from http://home.dntis.ro/~tvlad/proj/autofs/autofs-3.1.1.tar.gz to get the whole package and the new module. Put in /etc/auto.master something like: # $Id: auto.master,v 1.2 1997/10/06 21:52:03 hpa Exp $ # Sample auto.master file # Format of this file: # mountpoint map options # For details of the format look at autofs(8). /cdrom /etc/auto.cdrom and the auto.cdrom should be: # $Id: auto.cdrom # This is an automounter map and it has the following format # key [ -mount-options-separated-by-comma ] location # Details may be found in the autofs(5) manpage 1 -fstype=cdch,slot=1,uid=99,mode=755 :/dev/cdrom 2 -fstype=cdch,uid=99,slot=2 :/dev/cdrom 3 -fstype=cdch,slot=3 :/dev/cdrom 4 -fstype=cdch,slot=4 :/dev/cdrom type 'cd /cdrom/[1234]' to access the first four drives... Good luck and tell me if there is something is wrong. Vlad. From owner-autofs@linux.kernel.org Mon Jan 18 03:24:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA22629 for ; Mon, 18 Jan 1999 03:24:31 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA11438 for autofs-list; Mon, 18 Jan 1999 02:43:56 -0800 Received: from gwa.fr.bosch.de (firewall-user@gwa.fr.bosch.de [194.120.36.67]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA11434 for ; Mon, 18 Jan 1999 02:43:53 -0800 Received: (from uucp@localhost) by gwa.fr.bosch.de (8.9.1/8.9.1) id LAA18687 for ; Mon, 18 Jan 1999 11:43:50 +0100 (MET) Received: from mailgate.fr.internet.bosch.de( 194.120.36.134) by gwa.fr.bosch.de via smap (V2.1) id xma017757; Mon, 18 Jan 99 11:42:45 +0100 Received: from himh1.hi.bosch.de (HIMH1.hi.bosch.de [133.2.13.123]) by mailgate.fr.internet.bosch.de (8.9.1/8.9.1) with ESMTP id LAA12418 for ; Mon, 18 Jan 1999 11:42:44 +0100 (MET) Received: (from Rudolf.Dederer@fr.bosch.de) by himh1.hi.bosch.de (8.7.1/8.7.1) with ESMTP id LAA28601 for ; Mon, 18 Jan 1999 11:42:37 +0100 Message-ID: <36A3103E.D7A19CC3@fr.bosch.de> Date: Mon, 18 Jan 1999 11:43:10 +0100 From: Rudolf Dederer Organization: Robert Bosch GmbH X-Mailer: Mozilla 4.5 [en] (X11; I; Linux 2.0.36 i586) X-Accept-Language: ru, de, en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Problems with Solaris options Content-Type: text/plain; charset=iso-8859-1 X-MIME-Autoconverted: from 8bit to quoted-printable by mailgate.fr.internet.bosch.de id LAA12418 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by linux.kernel.org id CAA11436 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello, I run into Problems trying to use automounter version 3.1.3 to mount directories exported by a Sun via nisplus: mount cannot deal with the Solaris specific options, e.g. ´noquota´: Entry in nisplus ´scratch.org_dir´ (Solaris) dataSOL -noquota,soft SOL:/export/home/dataSOL Jan 18 10:32:44 LIN automount[3520]: attempting to mount entry /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: lookup(nisplus): looking up dataSOL Jan 18 10:32:44 LIN automount[3579]: lookup(nisplus): dataSOL -> -noquota,soft SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: expanded entry: -noquota,soft SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: parse(sun): gathered options: noquota,soft Jan 18 10:32:44 LIN automount[3579]: parse(sun): core of entry: SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: mount(nfs): calling mkdir /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: mount(nfs): calling mount -t nfs -o noquota,soft SOL:/usr/local/data /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: >> unknown nfs mount option: noquota Jan 18 10:32:44 LIN automount[3579]: mount(nfs): nfs: mount failure SOL:/usr/local/data on /scratch/dataSOL Is there a possibility for automounter to filter out the unknown options without stopping at this point? -- ____________________________________________________________________ Rudolf Dederer mailto:Rudolf.Dederer@fr.bosch.de ____________________________________________________________________ From owner-autofs@linux.kernel.org Mon Jan 18 14:50:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA23152 for ; Mon, 18 Jan 1999 14:50:17 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA10550 for autofs-list; Mon, 18 Jan 1999 14:12:56 -0800 Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA10546 for ; Mon, 18 Jan 1999 14:12:54 -0800 Received: from bbn.com (root@SHERRY.BBN.COM [128.33.161.12]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id RAA00760; Mon, 18 Jan 1999 17:20:45 -0500 (EST) Received: (from root@localhost) by bbn.com (8.8.7/8.8.7) id RAA03585; Mon, 18 Jan 1999 17:12:40 -0500 Date: Mon, 18 Jan 1999 17:12:40 -0500 Message-Id: <199901182212.RAA03585@bbn.com> From: root To: hpa@zytor.com, autofs@linux.kernel.org Cc: frichard@bbn.com Subject: autofs: multiple mount support Reply-to: frichard@bbn.com Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi- I was wondering if there's a version of autofs available which offers support for multiple mounts. I saw this entery in autofs-3.1.x/TODO: autofs v4 TODO list ------------------- . . * Support multimount entries. (Quite messy: can't unmount the tree as a unit, probably needs locking while constructing the tree, and needs scaffolding if no entry for /. Appears *very* difficult to do without threading!) . . We were forced to use amd because autofs offers no convenient way to handle our auto.master map which make use of the multimount feature like this: auto.master: /net/ auto. - auto.: * .:/nfs//& Right now, I'm forced to run an automount deamon for each entery in auto.master. doesn't export /nfs/ itself, but just the disk partitions under /nfs/. It would be difficult to change this. A multi-mount capable autofs would really save the day. Otherwise we'll have to keep using amd. Thanks, -Fred From owner-autofs@linux.kernel.org Mon Jan 18 14:35:04 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA23146 for ; Mon, 18 Jan 1999 14:35:04 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA09241 for autofs-list; Mon, 18 Jan 1999 13:45:59 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA09237 for ; Mon, 18 Jan 1999 13:45:56 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id NAA01274; Mon, 18 Jan 1999 13:45:08 -0800 From: Alvin Oga Message-Id: <199901182145.NAA01274@leopard.imedia.com> Subject: Re: autofs woes: fail to mount home directories To: Ed.Breen@Altavista.net (Ed Breen) Date: Mon, 18 Jan 1999 13:45:07 -0800 (PST) Cc: autofs@linux.kernel.org, aoga@leopard.imedia.com (Alvin Oga) In-Reply-To: <36A15AE0.44AD4AA6@Altavista.net> from "Ed Breen" at Jan 17, 99 03:37:04 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi why is grun exporting home2 to wax and wax is exporting home to grun ? > Wax's auto.master file contains other than comments: > > # Sample auto.master file > > /home /etc/auto.home Make sure that /home is not the system /home or you'll lose all the defaults stuff that came with teh distribution /home/{httpd,ftp,gopher} you might consider /.home /etc/auto.home for your auto.master file > The auto.home file is, missing comments: > > exdb -rw,hard,intr grun:/home2/exdb > > In wax's nsswitch file there is: > > automount: files nis > > Wax's export file has: > /home grun(rw) so far so good... > And I have included auto.home and auto.master in /var/yp/Makefile. > > I can log in on wax as exdb and the /home/exdb exists and every > thing is fine. However, while I can log into grun (kernel 2.0.36) as exdb, exdb > home directory fails to mount -- and this is crux of my problem. > > On grun: > > # cat /proc/filesystems > ... > nodev autofs > > # /sbin/lsmod > Module Pages Used by > autofs 2 1 > > # cat /etc/nsswitch.conf > ... > automount: nis not sure about this entry > # cat /etc/auto.master > > +auto.master > /home auto.home --timeout 60 shouldn't that be /etc/auto.home in order to find the file ??? not sure that +auto.master is supported feature > # cat /etc/auto.home > +auto.home not sure that +auto.master is supported feature try using: home -fstype=nfs wax:/home > # cat /etc/exports > ... > /home2 wax(rw) okay...exporting home2 to wax.... > >From the log file /var/log/syslog on grun: > Jan 17 11:44:16 grun automount[7194]: attempting to mount entry /home/exdb > Jan 17 11:44:16 grun automount[7323]: lookup(file): looking up exdb > Jan 17 11:44:16 grun automount[7323]: lookup(file): lookup for exdb failed probably needed /etc/auto.home entry in the auto.master file > So it seems to fail to find exdb, but why?? Remember I can log > in as exdb on grun. exdb's passwd/shadow is handled using NIS on wax. > > When I expect wax's /var/log/syslog I find: > > Jan 17 19:44:54 wax automount[5095]: lookup(file): looking up edb > Jan 17 19:44:54 wax automount[5095]: lookup(file): edb -> -rw,hard,intr^I > grun:/home2/exdb > Jan 17 19:44:54 wax automount[5095]: expanded entry: -rw,hard,intr^I > grun:/home2/exdb > Jan 17 19:44:54 wax automount[5095]: parse(sun): gathered options: rw,hard,intr > Jan 17 19:44:54 wax automount[5095]: parse(sun): core of entry: grun:/home2/exdb > Jan 17 19:44:54 wax automount[5095]: mount(nfs): calling mkdir /home/exdb > Jan 17 19:44:54 wax automount[5095]: mount(nfs): calling mount -t nfs -o > rw,hard,intr grun:/home2/exdb /home/exdb > Jan 17 19:44:54 wax automount[5095]: mount(nfs): mounted grun:/home2/exdb on > /home/exdb good.... > Any help would be greatly appreciated -- this problem is really slowing things > up. Also amd is installed but not used on the server wax, but it is not > installed on > grun, should I install amd on grun also? are you running amd or autofs ??? ... the log messages is for autofs as is the mailing list and files... have fun alvin From frichard@cyan.bbn.com Mon Jan 18 09:44:21 1999 Return-Path: Received: from jubilee.bbn.com (mail.d4m.bbn.com [128.33.160.97]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA22966 for ; Mon, 18 Jan 1999 09:44:19 -0800 Received: from cyan.bbn.com (frichard@cyan.bbn.com [128.33.160.15]) by jubilee.bbn.com (8.8.5/8.6.9) with ESMTP id MAA06932; Mon, 18 Jan 1999 12:51:37 -0500 (EST) Message-Id: <199901181751.MAA06932@jubilee.bbn.com> To: Alvin Oga cc: frichard@bbn.com Subject: Re: Sun style YP/NIS maps and autofs (fwd) In-reply-to: Your message of "Mon, 18 Jan 1999 01:44:47 PST." <199901180944.BAA22542@mail.Linux-Consulting.com> Reply-to: frichard@bbn.com Date: Mon, 18 Jan 1999 12:43:34 -0500 From: Fred Richardson Status: RO X-Status: Alvin- Again, thanks for all your work on trying to help with this. We currently have many servers which export something like this: On orchid (which is a Solaris server), exportfs includes: /nfs/orchid/u13 but orchid doesn't export "/nfs/orchid". For Solaris and Irix we set things up so that the auto.master map uses the `Multiple Mount' feature of NFS. This feature is on the TODO list for autofs-4, but autofs-3.x doesn't (and probably won't) support it. For now I'm going to try running with a hundred or so automount deamons and see if things hold up. Unfortunately, I can't change each NFS servers to export /nfs// ... Again, thanks for your efforts, -Fred From owner-autofs@linux.kernel.org Mon Jan 18 07:33:49 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA22879 for ; Mon, 18 Jan 1999 07:33:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA20611 for autofs-list; Mon, 18 Jan 1999 06:32:04 -0800 Received: from eclectic.kluge.net (root@eclectic.kluge.net [206.114.160.248]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA20608 for ; Mon, 18 Jan 1999 06:32:02 -0800 Received: from eclectic.kluge.net (felicity@localhost [127.0.0.1]) by eclectic.kluge.net (8.9.1a/8.9.1a) with ESMTP id JAA14175; Mon, 18 Jan 1999 09:31:45 -0500 Message-Id: <199901181431.JAA14175@eclectic.kluge.net> X-Mailer: exmh version 2.0.2 2/24/98 To: Rudolf Dederer cc: autofs@linux.kernel.org Subject: Re: Problems with Solaris options In-Reply-To: Message from Rudolf Dederer of "Mon, 18 Jan 1999 11:43:10 +0100." <36A3103E.D7A19CC3@fr.bosch.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Date: Mon, 18 Jan 1999 09:31:45 -0500 From: Theo Van Dinter Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by linux.kernel.org id GAA20609 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: | Is there a possibility for automounter to filter out the unknown options without | stopping at this point? it isn't automounter's responsibility to filter options (how does it know what is unknown?) there's a version of mount which ignores unknown options (2.8? I forget exactly). you can also filter out specific options (like noquota) with the rc.autofs file (it already parses the options a little bit). -- Randomly Generated Tagline: "Everyone lives by selling something." - R.L. Stevenson From owner-autofs@linux.kernel.org Mon Jan 18 03:24:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA22629 for ; Mon, 18 Jan 1999 03:24:31 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA11438 for autofs-list; Mon, 18 Jan 1999 02:43:56 -0800 Received: from gwa.fr.bosch.de (firewall-user@gwa.fr.bosch.de [194.120.36.67]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA11434 for ; Mon, 18 Jan 1999 02:43:53 -0800 Received: (from uucp@localhost) by gwa.fr.bosch.de (8.9.1/8.9.1) id LAA18687 for ; Mon, 18 Jan 1999 11:43:50 +0100 (MET) Received: from mailgate.fr.internet.bosch.de( 194.120.36.134) by gwa.fr.bosch.de via smap (V2.1) id xma017757; Mon, 18 Jan 99 11:42:45 +0100 Received: from himh1.hi.bosch.de (HIMH1.hi.bosch.de [133.2.13.123]) by mailgate.fr.internet.bosch.de (8.9.1/8.9.1) with ESMTP id LAA12418 for ; Mon, 18 Jan 1999 11:42:44 +0100 (MET) Received: (from Rudolf.Dederer@fr.bosch.de) by himh1.hi.bosch.de (8.7.1/8.7.1) with ESMTP id LAA28601 for ; Mon, 18 Jan 1999 11:42:37 +0100 Message-ID: <36A3103E.D7A19CC3@fr.bosch.de> Date: Mon, 18 Jan 1999 11:43:10 +0100 From: Rudolf Dederer Organization: Robert Bosch GmbH X-Mailer: Mozilla 4.5 [en] (X11; I; Linux 2.0.36 i586) X-Accept-Language: ru, de, en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Problems with Solaris options Content-Type: text/plain; charset=iso-8859-1 X-MIME-Autoconverted: from 8bit to quoted-printable by mailgate.fr.internet.bosch.de id LAA12418 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by linux.kernel.org id CAA11436 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello, I run into Problems trying to use automounter version 3.1.3 to mount directories exported by a Sun via nisplus: mount cannot deal with the Solaris specific options, e.g. ´noquota´: Entry in nisplus ´scratch.org_dir´ (Solaris) dataSOL -noquota,soft SOL:/export/home/dataSOL Jan 18 10:32:44 LIN automount[3520]: attempting to mount entry /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: lookup(nisplus): looking up dataSOL Jan 18 10:32:44 LIN automount[3579]: lookup(nisplus): dataSOL -> -noquota,soft SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: expanded entry: -noquota,soft SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: parse(sun): gathered options: noquota,soft Jan 18 10:32:44 LIN automount[3579]: parse(sun): core of entry: SOL:/usr/local/data Jan 18 10:32:44 LIN automount[3579]: mount(nfs): calling mkdir /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: mount(nfs): calling mount -t nfs -o noquota,soft SOL:/usr/local/data /scratch/dataSOL Jan 18 10:32:44 LIN automount[3579]: >> unknown nfs mount option: noquota Jan 18 10:32:44 LIN automount[3579]: mount(nfs): nfs: mount failure SOL:/usr/local/data on /scratch/dataSOL Is there a possibility for automounter to filter out the unknown options without stopping at this point? -- ____________________________________________________________________ Rudolf Dederer mailto:Rudolf.Dederer@fr.bosch.de ____________________________________________________________________ From owner-autofs@linux.kernel.org Mon Jan 18 16:18:24 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA23231 for ; Mon, 18 Jan 1999 16:18:24 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA15792 for autofs-list; Mon, 18 Jan 1999 15:46:27 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA15789 for ; Mon, 18 Jan 1999 15:46:24 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA25974; Mon, 18 Jan 1999 15:46:19 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA21433; Mon, 18 Jan 1999 15:46:18 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA10721; Mon, 18 Jan 1999 15:46:18 -0800 Message-Id: <199901182346.PAA10721@cesium.transmeta.com> Subject: Re: autofs: multiple mount support In-Reply-To: <199901182212.RAA03585@bbn.com> from root at "Jan 18, 99 05:12:40 pm" To: frichard@bbn.com Date: Mon, 18 Jan 1999 15:46:18 -0800 (PST) Cc: hpa@zytor.com, autofs@linux.kernel.org, frichard@bbn.com X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Don't hold your breath. This will go into 2.3. -hpa > Hi- > > I was wondering if there's a version of autofs available which offers > support for multiple mounts. I saw this entery in autofs-3.1.x/TODO: > > autofs v4 TODO list > ------------------- > . > . > * Support multimount entries. (Quite messy: can't unmount the > tree as a unit, probably needs locking while constructing the tree, > and needs scaffolding if no entry for /. Appears *very* difficult > to do without threading!) > . > . > > We were forced to use amd because autofs offers no convenient way to > handle our auto.master map which make use of the multimount feature > like this: > > auto.master: > /net/ auto. - > > auto.: > * .:/nfs//& > > Right now, I'm forced to run an automount deamon for each entery in > auto.master. doesn't export /nfs/ itself, but just > the disk partitions under /nfs/. It would be difficult to > change this. > > A multi-mount capable autofs would really save the day. Otherwise > we'll have to keep using amd. > > Thanks, > > -Fred > From owner-autofs@linux.kernel.org Mon Jan 25 08:09:51 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA30001 for ; Mon, 25 Jan 1999 08:09:51 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA08086 for autofs-list; Mon, 25 Jan 1999 07:34:28 -0800 Received: from penguin.wise.edt.ericsson.se (penguin-ext.wise.edt.ericsson.se [194.237.142.5]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA08078 for ; Mon, 25 Jan 1999 07:34:25 -0800 Received: from ratatosk.aom.ericsson.se (ratatosk.aom.ericsson.se [130.100.222.46]) by penguin.wise.edt.ericsson.se (8.9.0/8.9.0/WIREfire-1.2) with SMTP id QAA20920 for ; Mon, 25 Jan 1999 16:34:18 +0100 (MET) Received: from arioch15.aom.ericsson.se by ratatosk.aom.ericsson.se (SMI-8.6/LME-DOM-2.2.6) id QAA23980; Mon, 25 Jan 1999 16:34:20 +0100 Date: Mon, 25 Jan 1999 16:34:20 +0100 (MET) From: Fredrik Sorensson EDT/L/G 16333 xxxxx To: autofs@linux.kernel.org Subject: Multiple star entries Message-ID: Return-View-To: edtsorn@aom.ericsson.se MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by linux.kernel.org id HAA08083 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello. I would like to point out a difference between autofs and the Sun automounter. I've discussed this with Peter Anvin, but after a brief discussion he disagreed with me that the functionality proposed by me wasn't a good solution. I didn't agree, but I respect his opinions. Therefore, I don't think that I have much call for my idea, but since there is a difference, I still want to point it out. If you think I'm obnoxious and should go away, please ignore my ramblings... Ok, the scenario is as follows: We have a large Unix installation (ca 1000 Suns, ca 10 HP and 2 Linux), and we distribute software via NFS. I.e., only the OS is installed locally, and all software is installed on large NFS servers, and the software is "mounted" on the client when needed. All the software is then reachable via /usr/local/ . So, if I want a perl interpreter, I find it under /usr/local/perl5/bin/perl, where /usr/local/perl5 is mounted from my designated NFS server when I need it. Now, we don't specify every software in the autmounter configuration files, we use a couple of * entries, like this ('taz' is my NFSserver, an UltraSparc running Solaris 2.6): /etc/auto_local: * -rw,intr taz:/export/linux.core/& So if I ask for /usr/local/perl5 it mounts it from taz:/export/linux.core/perl5. And this works in autofs to. Like a charm. Now, the problem is, that we seperate different types of software into different directory. .core is software that must exist, like perl, global cshrc-files etc. All the boring bought software like framemaker goes into .local, and all the fun freeware goes into .godis (Swedish for candy). So, I added two more entries: /etc/auto.local * -rw,intr taz:/export/linux.core/& * -rw,intr taz:/export/linux.local/& * -rw,intr taz:/export/linux.godis/& Now it stops working. Or to be exact, autofs only looks att the first entry, linux.core, and doesn't care about the two other. It only looks at the first entry. (If I remove the linux.core-entry, autofs finds the linux.local-entry and works with that and skips the linux.godis-entry) I saw something about MultiMounts in autofs v4.0 in the docs, but I don't see that that is the solution to my problems. If I'm wrong, do tell me so. Ok, I hope I made myself understood now. If you say that "hey, that's the way it supposed to work", I can only say "fine...". But you should be aware that this is not the way it works in either Solaris or HPUX. I find a solution with a cronjob fixing links somewhat patchy. Fredrik Sörensson - fredrik.sorensson@aom.ericsson.se System/Network Administrator Ericsson Data Göteborg From owner-autofs@linux.kernel.org Mon Jan 25 07:44:53 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA29990 for ; Mon, 25 Jan 1999 07:44:52 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA01074 for autofs-list; Mon, 25 Jan 1999 05:49:24 -0800 Received: from nenuphar.saclay.cea.fr (nenuphar.saclay.cea.fr [132.166.192.7]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA01070 for ; Mon, 25 Jan 1999 05:49:22 -0800 Received: from muguet.saclay.cea.fr (muguet.saclay.cea.fr [132.166.192.6]) by nenuphar.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id OAA05751 for ; Mon, 25 Jan 1999 14:50:10 +0100 (MET) Received: from dphdse.saclay.cea.fr (dphdse.saclay.cea.fr [132.166.30.5]) by muguet.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id OAA29091 for ; Mon, 25 Jan 1999 14:46:16 +0100 (MET) Received: from seipca185.saclay.cea.fr (seipca185 [132.166.29.185]) by dphdse.saclay.cea.fr (8.9.0/8.9.0) with SMTP id PAA09842 for ; Mon, 25 Jan 1999 15:05:00 +0100 (MET) Date: Mon, 25 Jan 1999 15:05:00 +0100 (MET) Message-Id: <199901251405.PAA09842@dphdse.saclay.cea.fr> From: Jean-Paul Le Fevre To: autofs@linux.kernel.org Subject: automount exits immediatly Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I have just installed autofs 3.1.3 on my linux box but I am unable to make it working properly. My kernel is 2.1.105, the loader is ld-2.0.6.so, the lib libc-2.0.6.so. I looked at my daemon log file and noticed that parse_sun.so complained that the symbol 'un' was undefined, so I tried to declared it static but next time 'processor' was found undefined. I made it static, 'chunklen' was undefined etc. Eventually every symbols were defined, I have no error message in the log file. But nothing happens : the program exits, no daemon running. In the log file the last messages are : Jan 25 13:36:20 seipca185 automount[996]: starting automounter version 3.1.3, pa th = /mnt, maptype = file, mapname = /etc/auto.mnt Jan 25 13:36:20 seipca185 automount[996]: Map argc = 1 Jan 25 13:36:20 seipca185 automount[996]: Map argv[0] = /etc/auto.mnt Help would be appreciated, thank you, -- ____________________________________________________________________________ Jean-Paul Le Fevre http://www.multimania.com/lefevre E-mail : Jean-Paul.LeFevre@cea.fr Tel : (33) [0]1 69 08 74 50 From owner-autofs@linux.kernel.org Wed Jan 27 15:26:49 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA32298 for ; Wed, 27 Jan 1999 15:26:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA30594 for autofs-list; Wed, 27 Jan 1999 13:49:29 -0800 Received: from nua.lampf.lanl.gov (nua.lampf.lanl.gov [128.165.53.157]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA30590 for ; Wed, 27 Jan 1999 13:49:25 -0800 Received: from nua.lampf.lanl.gov (localhost [127.0.0.1]) by nua.lampf.lanl.gov (980427.SGI.8.8.8/970903.SGI.AUTOCF) via ESMTP id NAA73199; Wed, 27 Jan 1999 13:49:22 -0800 (PST) Message-ID: <36AF89E1.40C318E0@nua.lampf.lanl.gov> Date: Wed, 27 Jan 1999 14:49:21 -0700 From: Ben Sapp Reply-To: bsapp@lanl.gov Organization: Los Alamos National Laboratory X-Mailer: Mozilla 4.05C-SGI [en] (X11; I; IRIX 6.5 IP22) MIME-Version: 1.0 To: hpa@transmeta.com CC: autofs@linux.kernel.org Subject: useless if() Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I have the source for autofs-3.1.3 package. In the spawnv routine there is a set of if statements like this: if(f < 0){ sigprocmask(SIG_SETMASK, &oldsig, NULL); return -1; else if(f == 0){ . . . } else { . . . if( f < 0){ close(pipefd[0]); return -1; } . . . } When the program arrives at the second "if(f < 0)" f must be greater than zero. This of course only takes a few clock cycles to execute and so it does not slow the program down much, but you might as well get rid of it in future releases. -- Ben Sapp Los Alamos National Laboratory email: Phone: (505)667-3277 Fax: (505)665-7920 URL: http://www.neutrino.lanl.gov/ -- From owner-autofs@linux.kernel.org Tue Jan 26 09:04:11 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA31022 for ; Tue, 26 Jan 1999 09:04:11 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA23572 for autofs-list; Tue, 26 Jan 1999 07:39:49 -0800 Received: from nenuphar.saclay.cea.fr (nenuphar.saclay.cea.fr [132.166.192.7]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA23564 for ; Tue, 26 Jan 1999 07:39:45 -0800 Received: from muguet.saclay.cea.fr (muguet.saclay.cea.fr [132.166.192.6]) by nenuphar.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id QAA16522 for ; Tue, 26 Jan 1999 16:40:33 +0100 (MET) Received: from dphdse.saclay.cea.fr (dphdse.saclay.cea.fr [132.166.30.5]) by muguet.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id QAA21092 for ; Tue, 26 Jan 1999 16:36:38 +0100 (MET) Received: from seipca185.saclay.cea.fr (seipca185 [132.166.29.185]) by dphdse.saclay.cea.fr (8.9.0/8.9.0) with SMTP id QAA30397 for ; Tue, 26 Jan 1999 16:55:26 +0100 (MET) Date: Tue, 26 Jan 1999 16:55:26 +0100 (MET) Message-Id: <199901261555.QAA30397@dphdse.saclay.cea.fr> From: Jean-Paul Le Fevre To: autofs@linux.kernel.org Subject: automount exits immediatly Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I recently posted a question to the list because autofs does not work for me : it exits almost immediatly without doing anything. (kernel is 2.1.105, loader is ld-2.0.6.so, libc-2.0.6.so, autofs 3.1.3) I've just found where the problem occurs, it allows me to use a (dirty) workaround but I would like to understand what is wrong with my configuration. The program vanishes mysteriously in mount_init() of mount_nfs.c. It turns out that getprotobyname() is a black hole (as well as getservbyname()) : you never returns from these functions. I've commented out the calls to these functions using hardcoded default values instead and the daemon started to work happily. I've checked using a small standalone program that getprotobyname("udp") returns 17 as expected. BTW : It is really painful to debug automount, since it forks immediatly I do not know how to use gdb : I have inserted hundreds of syslog(LOG_DEBUG) to find out what was going on. Cheers, -- ____________________________________________________________________________ Jean-Paul Le Fevre http://www.multimania.com/lefevre E-mail : Jean-Paul.LeFevre@cea.fr Tel : (33) [0]1 69 08 74 50 From owner-autofs@linux.kernel.org Tue Feb 2 13:27:42 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA13580 for ; Tue, 2 Feb 1999 13:27:42 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA31050 for autofs-list; Tue, 2 Feb 1999 12:40:37 -0800 Received: from gate.aval.com (gate.aval.com [199.108.11.14]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA31043 for ; Tue, 2 Feb 1999 12:40:34 -0800 Received: from aval.com (neda.aval.com [199.108.11.205]) by gate.aval.com (8.8.7/8.7.3) with SMTP id MAA03781 for ; Tue, 2 Feb 1999 12:40:33 -0800 (PST) Received: from SMTP by aval.com (SMI-8.6/SMI-SVR4) id MAA28445; Tue, 2 Feb 1999 12:42:21 -0800 Received: from aval.com ([199.108.11.248]) by 199.108.11.240 (Norton AntiVirus for Internet Email Gateways 1.0) ; Tue, 02 Feb 1999 20:45:04 0000 (GMT) Message-ID: <36B762EA.DC9257EB@aval.com> Date: Tue, 02 Feb 1999 12:41:15 -0800 From: "Derek R. Pizzagoni" X-Mailer: Mozilla 4.5 [en] (X11; U; Linux 2.0.36 i686) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Automount Question Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hello, I just compiled automounter support in my Linux 2.0.36 kernel. Then, I downloaded, compiled, and installed the files from ftp.kernel.org:/pub/linux/daemons/autofs. After restarting everything, I tried to do the following: 1. Create mount point /test 2. Edit /etc/auto.master to include the following: /test /etc/auto.test 3. Edit /etc/auto.test to include the following: host:/test /test nfs rw 0 0 4. Confirm that /test is exported via NFS on host. I assume this should allow a local mountpoint /test to automount the /test subdirectory on 'host'. I started the autofs daemon, and I got the following: [root@cheese]# /etc/rc.d/init.d/autofs reload Checking for changes to /etc/auto.master .... Start /usr/sbin/automount /test file /etc/auto.test Now, when I go into /test, I get nothing. Shouldn't /test automount machine:/test? -- Derek R. Pizzagoni Information Technology Manager Aval Communications, Inc. URL: http://www.aval.com From owner-autofs@linux.kernel.org Tue Feb 2 13:32:22 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA13586 for ; Tue, 2 Feb 1999 13:32:22 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA32387 for autofs-list; Tue, 2 Feb 1999 12:56:36 -0800 Received: from gate.aval.com (gate.aval.com [199.108.11.14]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA32382 for ; Tue, 2 Feb 1999 12:56:33 -0800 Received: from aval.com (neda.aval.com [199.108.11.205]) by gate.aval.com (8.8.7/8.7.3) with SMTP id MAA03822 for ; Tue, 2 Feb 1999 12:56:31 -0800 (PST) Received: from SMTP by aval.com (SMI-8.6/SMI-SVR4) id MAA28809; Tue, 2 Feb 1999 12:58:19 -0800 Received: from aval.com ([199.108.11.248]) by 199.108.11.240 (Norton AntiVirus for Internet Email Gateways 1.0) ; Tue, 02 Feb 1999 21:01:01 0000 (GMT) Message-ID: <36B766A9.5F88AC8B@aval.com> Date: Tue, 02 Feb 1999 12:57:13 -0800 From: "Derek R. Pizzagoni" X-Mailer: Mozilla 4.5 [en] (X11; U; Linux 2.0.36 i686) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Got it Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Ok, I looked at the samples, and have had some mild success. Now onto mounting Microsoft shares via SMB... Sorry for the unnecessary traffic. -- Derek R. Pizzagoni Information Technology Manager Aval Communications, Inc. URL: http://www.aval.com From owner-autofs@linux.kernel.org Tue Feb 2 13:54:55 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA13592 for ; Tue, 2 Feb 1999 13:54:55 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA00512 for autofs-list; Tue, 2 Feb 1999 13:04:51 -0800 Received: from eclectic.kluge.net (root@eclectic.kluge.net [206.114.160.248]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA00509 for ; Tue, 2 Feb 1999 13:04:50 -0800 Received: from eclectic.kluge.net (felicity@localhost [127.0.0.1]) by eclectic.kluge.net (8.9.2/8.9.2) with ESMTP id QAA32239; Tue, 2 Feb 1999 16:01:32 -0500 (EST) Message-Id: <199902022101.QAA32239@eclectic.kluge.net> X-Mailer: exmh version 2.0.2 2/24/98 To: "Derek R. Pizzagoni" cc: autofs@linux.kernel.org Subject: Re: Automount Question In-Reply-To: Message from "Derek R. Pizzagoni" of "Tue, 02 Feb 1999 12:41:15 PST." <36B762EA.DC9257EB@aval.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 02 Feb 1999 16:01:32 -0500 From: Theo Van Dinter Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: | 2. Edit /etc/auto.master to include the following: | /test /etc/auto.test ok. | 3. Edit /etc/auto.test to include the following: | host:/test /test nfs rw 0 0 completely wrong. this is an automount map, not the fstab. | Now, when I go into /test, I get nothing. Shouldn't /test automount | machine:/test? no. entries under auto.test will mount as /test/ ... ie (auto.test): foobar host:/test will mount host:/test on /test/foobar on the automount client. -- Randomly Generated Tagline: "In space, no one can hear you scream!" From owner-autofs@linux.kernel.org Tue Feb 2 14:01:31 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA13600 for ; Tue, 2 Feb 1999 14:01:30 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA01783 for autofs-list; Tue, 2 Feb 1999 13:17:43 -0800 Received: from gate.aval.com (gate.aval.com [199.108.11.14]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA01772 for ; Tue, 2 Feb 1999 13:17:40 -0800 Received: from aval.com (neda.aval.com [199.108.11.205]) by gate.aval.com (8.8.7/8.7.3) with SMTP id NAA03878 for ; Tue, 2 Feb 1999 13:17:38 -0800 (PST) Received: from SMTP by aval.com (SMI-8.6/SMI-SVR4) id NAA29245; Tue, 2 Feb 1999 13:19:26 -0800 Received: from aval.com ([199.108.11.248]) by 199.108.11.240 (Norton AntiVirus for Internet Email Gateways 1.0) ; Tue, 02 Feb 1999 21:22:08 0000 (GMT) Message-ID: <36B76B9B.61CADE7E@aval.com> Date: Tue, 02 Feb 1999 13:18:19 -0800 From: "Derek R. Pizzagoni" X-Mailer: Mozilla 4.5 [en] (X11; U; Linux 2.0.36 i686) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: Re: Automount Question References: <199902022059.MAA22681@cesium.transmeta.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thank you for your responses (I have received 2 so far). What I did was copy the auto.master and auto.misc files in the samples directory. Then, I added my line item. Now, I just go under /misc/xxx, and I get it. Would anyone know how to enter the line in the auto.misc file to allow access to a SMB share? I tried the following: mshost -rw,soft,intr //mshost/shared This apparently does not work. I noticed in the man page that it says you don't need the double //, but even with 1, it doesn't work. I also tried: mshost -fstype=smb ://mshost/shared This actually lets me change into the /misc/mshost/shared directory, but no files exist. Thank you in advance, Derek "H. Peter Anvin" wrote: > > Hello, > > > > I just compiled automounter support in my Linux 2.0.36 kernel. Then, I > > downloaded, compiled, and installed the files from > > ftp.kernel.org:/pub/linux/daemons/autofs. > > > > After restarting everything, I tried to do the following: > > > > 1. Create mount point /test > > 2. Edit /etc/auto.master to include the following: > > /test /etc/auto.test > > 3. Edit /etc/auto.test to include the following: > > host:/test /test nfs rw 0 0 > > 4. Confirm that /test is exported via NFS on host. > > > > I assume this should allow a local mountpoint /test to automount the > > /test subdirectory on 'host'. I started the autofs daemon, and I got > > the following: > > > > [root@cheese]# /etc/rc.d/init.d/autofs reload > > Checking for changes to /etc/auto.master .... > > Start /usr/sbin/automount /test file /etc/auto.test > > > > Now, when I go into /test, I get nothing. Shouldn't /test automount > > machine:/test? > > > > Please read the documentation first; the automount maps are *NOT* the > same format as /etc/fstab. > > /etc/auto.test should look like > > test host:/test > > ... except that will create a mount point /test/test, not /test. What > it sounds like you're trying to do is called a direct mount, which > isn't supported. > > -hpa From owner-autofs@linux.kernel.org Tue Feb 2 14:09:10 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA13606 for ; Tue, 2 Feb 1999 14:09:10 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA03710 for autofs-list; Tue, 2 Feb 1999 13:38:44 -0800 Received: from gate.aval.com (gate.aval.com [199.108.11.14]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA03701 for ; Tue, 2 Feb 1999 13:38:41 -0800 Received: from aval.com (neda.aval.com [199.108.11.205]) by gate.aval.com (8.8.7/8.7.3) with SMTP id NAA03919 for ; Tue, 2 Feb 1999 13:38:40 -0800 (PST) Received: from SMTP by aval.com (SMI-8.6/SMI-SVR4) id NAA29623; Tue, 2 Feb 1999 13:40:28 -0800 Received: from aval.com ([199.108.11.248]) by 199.108.11.240 (Norton AntiVirus for Internet Email Gateways 1.0) ; Tue, 02 Feb 1999 21:43:10 0000 (GMT) Message-ID: <36B77089.BCACA17A@aval.com> Date: Tue, 02 Feb 1999 13:39:21 -0800 From: "Derek R. Pizzagoni" X-Mailer: Mozilla 4.5 [en] (X11; U; Linux 2.0.36 i686) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: [Fwd: Automount Questiony] Content-Type: multipart/mixed; boundary="------------2CE486DDB8C20F1843B6A371" Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: This is a multi-part message in MIME format. --------------2CE486DDB8C20F1843B6A371 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I apologize for asking very basic questions regarding autofs. I have been referred to the docs, but am unable to find them. I downloaded autofs-3.1.3.tar.gz, and get the following: COPYING: English text COPYRIGHT: English text Makefile: English text Makefile.conf: English text Makefile.conf.in: English text Makefile.rules: English text NEWS: International language text README: English text README.options: English text TODO: English text aclocal.m4: English text config.log: English text config.status: Bourne shell script text configure: Bourne shell script text configure.in: English text daemon: directory include: directory man: directory modules: directory multiserver_mount.patch: C program text The README files give some basic information, but nothing too detailed. I did not see any detailed documentation. If someone can tell me where the docs can be found, I will definitely be gainfully employed for a while, and will hopefully be more self-sufficient. With the help of H. Peter Anvin, and Theo Van Dinter, I have automounting working with NFS and local filesystems, and am currently working on SMB. I believe I have the syntax ok, but I need to find out if the automounter will forward the request for a password when mounting a password protected share. Again, thank you for your help so far. - Derek --------------2CE486DDB8C20F1843B6A371 Content-Type: message/rfc822 Content-Transfer-Encoding: 7bit Content-Disposition: inline Return-Path: Received: from gate.aval.com by aval.com (SMI-8.6/SMI-SVR4) id NAA29357; Tue, 2 Feb 1999 13:27:26 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by gate.aval.com (8.8.7/8.7.3) with ESMTP id NAA03896 for ; Tue, 2 Feb 1999 13:25:31 -0800 (PST) Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id NAA25316 for ; Tue, 2 Feb 1999 13:24:00 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id NAA29890 for ; Tue, 2 Feb 1999 13:24:35 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id NAA22978 for drp@aval.com; Tue, 2 Feb 1999 13:24:35 -0800 Message-Id: <199902022124.NAA22978@cesium.transmeta.com> Subject: Re: Automount Questiony In-Reply-To: <36B76B9B.61CADE7E@aval.com> from "Derek R. Pizzagoni" at "Feb 2, 99 01:18:19 pm" To: drp@aval.com (Derek R. Pizzagoni) Date: Tue, 2 Feb 1999 13:24:35 -0800 (PST) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > > I also tried: > > mshost -fstype=smb ://mshost/shared > > This actually lets me change into the /misc/mshost/shared directory, but no > files exist. > The filesystem is called "smbfs". This is also in the docs. -hpa --------------2CE486DDB8C20F1843B6A371-- From owner-autofs@linux.kernel.org Tue Feb 2 17:01:35 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA13713 for ; Tue, 2 Feb 1999 17:01:35 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA19050 for autofs-list; Tue, 2 Feb 1999 16:26:42 -0800 Received: from fuzzy.foxtaur.com (tnt-dal-stat-47.dallas.net [209.44.42.47]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA19040 for ; Tue, 2 Feb 1999 16:26:38 -0800 Received: (from fox@localhost) by fuzzy.foxtaur.com (8.7.5/8.6.11) id SAA15480 for autofs@linux.kernel.org; Tue, 2 Feb 1999 18:23:02 -0600 Date: Tue, 2 Feb 1999 18:23:02 -0600 From: Fuzzy Fox To: autofs@linux.kernel.org Subject: Re: Automount Question Message-ID: <19990202182302.B15366@fuzzy.foxtaur.com> Mail-Followup-To: autofs@linux.kernel.org References: <199902022059.MAA22681@cesium.transmeta.com> <36B76B9B.61CADE7E@aval.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.1i In-Reply-To: <36B76B9B.61CADE7E@aval.com>; from Derek R. Pizzagoni on Tue, Feb 02, 1999 at 01:18:19PM -0800 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Derek R. Pizzagoni wrote: > > Would anyone know how to enter the line in the auto.misc file to allow > access to a SMB share? I think you've got it. > I also tried: > > mshost -fstype=smb ://mshost/shared That's about right. Here's mine: dir -fstype=smbfs,uid=501,gid=100,filemod=3664,dirmod=3775 ://server/share Your problem is probably that you are not using the correct "smbmount" script that knows how to tell the newer smbmount the proper options in the old format. -- fox@dallas.net (Fuzzy Fox) || "Nothing takes the taste out of peanut sometimes known as David DeSimone || butter quite like unrequited love." http://www.dallas.net/~fox/ || -- Charlie Brown From owner-autofs@linux.kernel.org Tue Feb 2 18:30:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id SAA13777 for ; Tue, 2 Feb 1999 18:30:16 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id SAA26846 for autofs-list; Tue, 2 Feb 1999 18:01:35 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id SAA26841 for ; Tue, 2 Feb 1999 18:01:32 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id SAA05374; Tue, 2 Feb 1999 18:00:59 -0800 From: Alvin Oga Message-Id: <199902030200.SAA05374@leopard.imedia.com> Subject: Re: Automount Question To: fox@dallas.net (Fuzzy Fox) Date: Tue, 2 Feb 1999 18:00:59 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <19990202182302.B15366@fuzzy.foxtaur.com> from "Fuzzy Fox" at Feb 2, 99 06:23:02 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi... i'd try one experiment too: linux# smbclient -L server -U Administrator - you should get a list of shares... linux# smbmount //WinNT/C /mnt/WinNT -U Administrator -C -P passwd linux# ls -l /mnt/WinNT - you should see everything... linux# umount /mnt/WinNT than I think autofs can do it too... have fun alvin > Derek R. Pizzagoni wrote: > > > > Would anyone know how to enter the line in the auto.misc file to allow > > access to a SMB share? > > I think you've got it. > > > I also tried: > > > > mshost -fstype=smb ://mshost/shared > > That's about right. Here's mine: > > dir -fstype=smbfs,uid=501,gid=100,filemod=3664,dirmod=3775 ://server/share > > Your problem is probably that you are not using the correct "smbmount" > script that knows how to tell the newer smbmount the proper options in > the old format. > > -- > fox@dallas.net (Fuzzy Fox) || "Nothing takes the taste out of peanut > sometimes known as David DeSimone || butter quite like unrequited love." > http://www.dallas.net/~fox/ || -- Charlie Brown > From owner-autofs@linux.kernel.org Wed Feb 3 05:29:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id FAA14365 for ; Wed, 3 Feb 1999 05:29:27 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id EAA04470 for autofs-list; Wed, 3 Feb 1999 04:11:21 -0800 Received: from mailgate.fore.com (mailgate.fore.com [169.144.68.6]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id EAA04467 for ; Wed, 3 Feb 1999 04:11:18 -0800 Received: from mailman.fore.com (mailman.fore.com [169.144.2.12]) by mailgate.fore.com (8.9.1/8.9.1) with ESMTP id HAA10856; Wed, 3 Feb 1999 07:11:12 -0500 (EST) Received: from sol.eng.fore.com (sol [169.144.155.73]) by mailman.fore.com (8.8.8/8.8.8) with ESMTP id HAA21040; Wed, 3 Feb 1999 07:11:38 -0500 (EST) Received: from spud.eng.fore.com (spud [169.144.87.22]) by sol.eng.fore.com (8.8.8/8.8.8) with ESMTP id HAA00814; Wed, 3 Feb 1999 07:11:16 -0500 (EST) Received: from localhost (bgarcia@localhost) by spud.eng.fore.com (8.8.7/8.8.7) with ESMTP id HAA05262; Wed, 3 Feb 1999 07:13:37 -0500 X-Authentication-Warning: spud.eng.fore.com: bgarcia owned process doing -bs Date: Wed, 3 Feb 1999 07:13:37 -0500 (EST) From: "Brad M. Garcia" X-Sender: bgarcia@spud.eng.fore.com To: grep@oriole.sbay.org cc: Alvin Oga , Roger Gordon , autofs@linux.kernel.org Subject: Re: autofs and the new linux 2.2.1 kernel In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Wed, 3 Feb 1999, George Bonser wrote: > I can confirm that Red Hat 5.x is not 2.2 ready. You need to update the > stuff listed in the Documentation/Changes file in the source tree. I can confirm that Red Hat 5.2 _is_ 2.2 ready. (although I've been keeping all my rpm's updated with the errata list - perhaps a rh52 system that hasn't been upgraded won't work?). I'm currently running 2.2pre7ac4 on two machines, and 2.1.131 on another with no problems. I had to tweak my rc.* files to do things like load some modules, run ipchains in place of ipfwadm, etc, but otherwise it was a simple drop-in replacement. autofs runs much better under 2.2 than under 2.0.*. I kept leaking resources (mount points) under 2.0.* (until I could no longer mount partitions), but have had no problems with 2.2. Brad Garcia ___/ __ / __ / ___/ "Being the Linux of digital media __/ / / / _/ __/ would be a very good life." _/ ____/ _/ _| ____/ - Jean-Louis Gassee, CEO of Be, Inc. From owner-autofs@linux.kernel.org Wed Feb 3 02:31:13 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA14155 for ; Wed, 3 Feb 1999 02:31:12 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA25934 for autofs-list; Wed, 3 Feb 1999 01:27:35 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA25923 for ; Wed, 3 Feb 1999 01:27:26 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id KAA01140; Wed, 3 Feb 1999 10:25:33 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id KAA12335; Wed, 3 Feb 1999 10:25:32 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Wed, 03 Feb 1999 10:25:32 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: grep@oriole.sbay.org Subject: Re: autofs and the new linux 2.2.1 kernel Cc: Roger Gordon , autofs@linux.kernel.org, Alvin Oga Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke George Bonser on 03-Feb-99 : > On Tue, 2 Feb 1999, Alvin Oga wrote: > >> >> hi roger >> >> not sure what the incompatibilities are....but simply >> replacing the kernel might break some of your other >> binaries... >> - I've "broken" my old test rh-5.0 box with >> linux-2.2.1 kernel >> > > I can confirm that Red Hat 5.x is not 2.2 ready. You need to update the > stuff listed in the Documentation/Changes file in the source tree. There > are instructions in there showing which versions you need and how to tell > which version you have. Users of Debian slink and potato should be fine, > 2.2 will drop in. It will even autodetect and use ipchains rather than > ipfwadm to set up the anti-spoofing at network start. > I just want to add Slackware 3.5 to the 2.2.0(1) compartibility list. I compiled the kernel last week and it runs smoothly with the autofs-3.1.3. I guess, the newer Slackware 3.6 will work just fine, too :) Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 3 01:34:15 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA14116 for ; Wed, 3 Feb 1999 01:34:14 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA22999 for autofs-list; Wed, 3 Feb 1999 00:47:46 -0800 Received: from shorelink.com (root@calvin.shorelink.com [209.31.227.112]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA22995 for ; Wed, 3 Feb 1999 00:47:43 -0800 Received: from localhost ([127.0.0.1]) by shorelink.com with smtp (ident grep using rfc1413) id m107xyE-002vkpC (Debian Smail-3.2.0.101 1997-Dec-17 #2); Wed, 3 Feb 1999 00:47:38 -0800 (PST) Date: Wed, 3 Feb 1999 00:47:32 -0800 (PST) From: George Bonser Reply-To: grep@oriole.sbay.org To: Alvin Oga cc: Roger Gordon , Alvin Oga , autofs@linux.kernel.org Subject: Re: autofs and the new linux 2.2.1 kernel In-Reply-To: <199902030708.XAA12411@leopard.imedia.com> Message-ID: X-No-Archive: yes MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 2 Feb 1999, Alvin Oga wrote: > > hi roger > > not sure what the incompatibilities are....but simply > replacing the kernel might break some of your other > binaries... > - I've "broken" my old test rh-5.0 box with > linux-2.2.1 kernel > I can confirm that Red Hat 5.x is not 2.2 ready. You need to update the stuff listed in the Documentation/Changes file in the source tree. There are instructions in there showing which versions you need and how to tell which version you have. Users of Debian slink and potato should be fine, 2.2 will drop in. It will even autodetect and use ipchains rather than ipfwadm to set up the anti-spoofing at network start. I just went through this with a Red Hat box at work. (1 Red Hat, 8 Debian) George Bonser From owner-autofs@linux.kernel.org Tue Feb 2 23:56:22 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id XAA14031 for ; Tue, 2 Feb 1999 23:56:22 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id XAA16890 for autofs-list; Tue, 2 Feb 1999 23:09:10 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id XAA16883 for ; Tue, 2 Feb 1999 23:09:07 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id XAA12411; Tue, 2 Feb 1999 23:08:46 -0800 From: Alvin Oga Message-Id: <199902030708.XAA12411@leopard.imedia.com> Subject: Re: autofs and the new linux 2.2.1 kernel To: rmg745@gwis.com (Roger Gordon) Date: Tue, 2 Feb 1999 23:08:45 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <36B7C93C.B53D0DCD@gwis.com> from "Roger Gordon" at Feb 2, 99 10:57:48 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi roger not sure what the incompatibilities are....but simply replacing the kernel might break some of your other binaries... - I've "broken" my old test rh-5.0 box with linux-2.2.1 kernel you need to upgrade your binaries and libs too.... to have a true linux-2.2 machine http://www.linux-consulting.com/Linux-2.2/upgrade-2.2.pl - to see what needs to change along with references to other linux-2.2 docs - - I downloaded most of the sources you will be needing to recompile under linux-2-2 For changing amd to autofs... please check: http://www.linux-consulting.com/Amd_AutoFS/autofs.html have fun alvin > I'm relatively new to Linux about a year of using! I multiboot three > different versions off a pentium. Linux Suse 5.2, Caldera 1.2 and Redhat > 5.0. I found that Linux is great operating system ! > > I have download and compiled the new 2.2.1 kernel. I compiled several > zImage new kernels with various features built into the new kernel. And > I plan to run and test each one. > > But before installing any of these new creations I was wondering how to > implement /autofs on distributions such as Caldera 1.2 and replace the > /etc/sysconfig/daemons/amd. If it is possible at all. I'm experimenting > and not afraid to test. And take risks with new configurations. > > Placing the new kernel and updating the Suse 5.2 distribution should be > straight foward since /autofs is in place. > > Thanks in advance for any tips or pointers! > > rmg > From owner-autofs@linux.kernel.org Tue Feb 2 21:35:56 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id VAA13953 for ; Tue, 2 Feb 1999 21:35:56 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id UAA07178 for autofs-list; Tue, 2 Feb 1999 20:57:15 -0800 Received: from gwis.com (root@darcy.gwis.com [209.57.72.3]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id UAA07173 for ; Tue, 2 Feb 1999 20:57:12 -0800 Received: from gwis.com (pap19.gwis.com [209.57.71.49]) by gwis.com (8.9.0/8.9.0) with ESMTP id XAA26274 for ; Tue, 2 Feb 1999 23:57:09 -0500 (EST) Message-ID: <36B7C93C.B53D0DCD@gwis.com> Date: Tue, 02 Feb 1999 22:57:48 -0500 From: Roger Gordon X-Mailer: Mozilla 4.04C-Caldera [en] (X11; I; Linux 2.0.33 i586) MIME-Version: 1.0 To: "autofs@linux.kernel.org" Subject: autofs and the new linux 2.2.1 kernel Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: I'm relatively new to Linux about a year of using! I multiboot three different versions off a pentium. Linux Suse 5.2, Caldera 1.2 and Redhat 5.0. I found that Linux is great operating system ! I have download and compiled the new 2.2.1 kernel. I compiled several zImage new kernels with various features built into the new kernel. And I plan to run and test each one. But before installing any of these new creations I was wondering how to implement /autofs on distributions such as Caldera 1.2 and replace the /etc/sysconfig/daemons/amd. If it is possible at all. I'm experimenting and not afraid to test. And take risks with new configurations. Placing the new kernel and updating the Suse 5.2 distribution should be straight foward since /autofs is in place. Thanks in advance for any tips or pointers! rmg From owner-autofs@linux.kernel.org Sat Feb 6 14:39:13 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA28710 for ; Sat, 6 Feb 1999 14:39:13 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA05566 for autofs-list; Sat, 6 Feb 1999 13:53:52 -0800 Received: from lunix.mit.edu (IDENT:sarantos@LUNIX.MIT.EDU [18.172.0.105]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA05563 for ; Sat, 6 Feb 1999 13:53:48 -0800 Received: (from sarantos@localhost) by lunix.mit.edu (8.8.7/8.8.7) id QAA02560 for autofs@linux.kernel.org; Sat, 6 Feb 1999 16:53:43 -0500 Date: Sat, 6 Feb 1999 16:53:43 -0500 From: Sarantos Kapidakis Message-Id: <199902062153.QAA02560@lunix.mit.edu> To: autofs@linux.kernel.org Subject: timeout Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: It would be nice if autofs could use a different timeout per filesystem (on one mount point). I would use a lower timeout for the floppy, and higher for other filesystems. sarantos From aoga@leopard.imedia.com Thu Feb 11 15:55:57 1999 Return-Path: Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA13038 for ; Thu, 11 Feb 1999 15:55:57 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA17813; Thu, 11 Feb 1999 15:51:03 -0800 From: Alvin Oga Message-Id: <199902112351.PAA17813@leopard.imedia.com> Subject: Re: autofs and the new linux 2.2.1 kernel To: schnitzer@rz.rwth-aachen.de (H.-J. Schnitzer) Date: Thu, 11 Feb 1999 15:51:03 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com In-Reply-To: <36C3105E.C7D0260F@rz.RWTH-Aachen.DE> from "H.-J. Schnitzer" at Feb 11, 99 06:16:14 pm Content-Type: text Status: RO X-Status: hi h.j.s. I just started playing with rh-5.2 and applied all rh-5.2 patched and applied all new versions of application binaries that the linux-2.2.1 kernel required....and am using autofs-3.1.3 - all works fine -- at least my test machine has no problems... http://www.linux-consulting.com/Linux-2.2 upgrade-2.2.Virgin-RH5.2.txt for binaries that need to be upgraded have fun alvin http://www.linux-consulting.com/Amd_AutoFS/autofs-HOWTO.html > I am running several machine with Red Hat 5.2 and autofs 3.1.3, > and have rather big problems with autofs/NFS. Especially for maps > referencing submounts ('-fstype=autofs') the automounter sometimes > stops working after some time. > > We are running the Linux machines as NFS clients with an AIX-NFS server, > and, independently of autofs, this combination does practically > not work (*very* slow connection, the file contents are sometimes corrupted > when writing > to the server!) since we have upgraded to the 2.2.1 kernel. Hence, it is not > quite clear > if the hanging automounter are due to an autofs or due to an NFS problem. > > I would be interested to hear if other people are able to run Linux 2.2.x as > > clients in combination with an AIX file server. > > H.J. Schnitzer > > From owner-autofs@linux.kernel.org Thu Feb 11 10:14:14 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA12816 for ; Thu, 11 Feb 1999 10:14:14 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA21854 for autofs-list; Thu, 11 Feb 1999 09:16:55 -0800 Received: from mail.rwth-aachen.de (mail.RWTH-Aachen.DE [137.226.144.9]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA21850 for ; Thu, 11 Feb 1999 09:16:49 -0800 Received: from ibmc8.rz.RWTH-Aachen.DE by mail.rwth-aachen.de (PMDF V5.1-12 #30440) with ESMTP id <01J7MKPVU17000003N@mail.rwth-aachen.de> for autofs@linux.kernel.org; Thu, 11 Feb 1999 18:15:06 +0100 Received: from rz.RWTH-Aachen.DE (loopback [127.0.0.1]) by ibmc8.rz.RWTH-Aachen.DE (AIX4.2/UCB 8.7/8.7) with ESMTP id SAA57382 for ; Thu, 11 Feb 1999 18:16:15 +0100 (MEZ) Date: Thu, 11 Feb 1999 18:16:14 +0100 From: "H.-J. Schnitzer" Subject: Re: autofs and the new linux 2.2.1 kernel To: autofs@linux.kernel.org Message-id: <36C3105E.C7D0260F@rz.RWTH-Aachen.DE> Organization: Rechenzentrum der RWTH-Aachen MIME-version: 1.0 X-Mailer: Mozilla 4.5 [en] (X11; I; AIX 4.2) Content-type: text/plain; charset=us-ascii Content-transfer-encoding: 7bit X-Accept-Language: en References: Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi. I am running several machine with Red Hat 5.2 and autofs 3.1.3, and have rather big problems with autofs/NFS. Especially for maps referencing submounts ('-fstype=autofs') the automounter sometimes stops working after some time. We are running the Linux machines as NFS clients with an AIX-NFS server, and, independently of autofs, this combination does practically not work (*very* slow connection, the file contents are sometimes corrupted when writing to the server!) since we have upgraded to the 2.2.1 kernel. Hence, it is not quite clear if the hanging automounter are due to an autofs or due to an NFS problem. I would be interested to hear if other people are able to run Linux 2.2.x as clients in combination with an AIX file server. H.J. Schnitzer From owner-autofs@linux.kernel.org Mon Feb 15 18:17:18 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id SAA18238 for ; Mon, 15 Feb 1999 18:17:18 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id RAA27601 for autofs-list; Mon, 15 Feb 1999 17:33:38 -0800 Received: from artful.grumblesmurf.net (dsl-209-162-215-45.easystreet.com [209.162.215.45]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id RAA27597 for ; Mon, 15 Feb 1999 17:33:35 -0800 Received: from localhost (darrell@localhost) by artful.grumblesmurf.net (8.8.8/8.8.8) with SMTP id RAA10913 for ; Mon, 15 Feb 1999 17:33:33 -0800 (PST) Date: Mon, 15 Feb 1999 17:33:33 -0800 (PST) From: Darrell Fuhriman X-Sender: darrell@artful To: autofs@linux.kernel.org Subject: patch to add a couple of variables. Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: So, I was setting up an automounter environ using NIS and redhat 5.x. I realized that the $HOST on redhat by default is fully qualified, which makes using map entries like this: foobar localhost:/disk/$HOST/0/foobar kind of annoying. My options were to end up creating a directory like: /disk/foo.mydomain.com/0/ or to change the hostname to be just 'foo', or hack the autofs code. I chose to hack the code. Thus, I added to variables to the available ones, namely $NODE, which is the part of the hostname up to the first period, and $NISDOMAIN, which is well, the NIS domain. It's included below. While I'm here, does anyone know a way to fix the braindead-ism that redhat seems to exhibit? Namely, choosing the loopback interface as the default instead of eth0... Darrell Fuhriman diff -r -u autofs-3.1.1.old/man/autofs.5 autofs-3.1.1/man/autofs.5 --- autofs-3.1.1.old/man/autofs.5 Thu Apr 9 16:34:04 1998 +++ autofs-3.1.1/man/autofs.5 Mon Feb 15 17:11:55 1999 @@ -105,10 +105,12 @@ .nf ARCH Architecture (uname -m) CPU Processor Type -HOST Hostname (uname -n) +HOST Full Hostname (uname -n) OSNAME Operating System (uname -s) OSREL Release of OS (uname -r) OSVERS Version of OS (uname -v) +NISDOMAIN NIS Domainname (nisdomainname) +NODE Nodename (uname -n | sed 's/\\..*$//') .fi .RE .sp diff -r -u autofs-3.1.1.old/modules/parse_sun.c autofs-3.1.1/modules/parse_sun.c --- autofs-3.1.1.old/modules/parse_sun.c Thu Apr 9 16:34:03 1998 +++ autofs-3.1.1/modules/parse_sun.c Mon Feb 15 17:07:03 1999 @@ -54,6 +54,8 @@ struct utsname un; char processor[65]; /* Not defined on Linux, so we make our own */ +char nisdomain[65]; /* This is a nice var to have */ +char nodename[65]; /* This is a nice var to have as well */ static struct substvar /* Predefined variables: tail of link chain */ sv_arch = { "ARCH", un.machine, NULL }, @@ -61,7 +63,10 @@ sv_host = { "HOST", un.nodename, &sv_cpu }, sv_osname = { "OSNAME", un.sysname, &sv_host }, sv_osrel = { "OSREL", un.release, &sv_osname }, - sv_osvers = { "OSVERS", un.version, &sv_osrel }; + sv_nisdomain = { "NISDOMAIN", nisdomain, &sv_osrel }, + sv_node = { "NODE", nodename, &sv_nisdomain }, + sv_osvers = { "OSVERS", un.version, &sv_node }; + /* Default context pattern */ @@ -259,6 +264,7 @@ const char *xopt; int optlen, len; int i, bval; + int j; /* Get processor information for predefined escapes */ @@ -270,7 +276,23 @@ if ( processor[0] == 'i' && processor[1] >= '3' && !strcmp(processor+2, "86") ) processor[1] = '3'; + + /* Get NIS Domainname */ + if ((getdomainname(nisdomain, 1024)) != 0) + nisdomain[0] = '\0'; + /* Get our nodename, this is the hostname up to the first '.' */ + uname(&un); + strcpy(nodename, un.nodename); + nodename[65] = '\0'; + for(j=0; j <= 64; j++) { + if (nodename[j] == '.') { + nodename[j] = '\0'; + break; + } + } } + + /* Set up context and escape chain */ From owner-autofs@linux.kernel.org Wed Feb 17 08:05:12 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA19949 for ; Wed, 17 Feb 1999 08:05:12 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA23594 for autofs-list; Wed, 17 Feb 1999 06:00:13 -0800 Received: from hns.com (nbeckerpc.hns.com [139.85.24.35]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA23586 for ; Wed, 17 Feb 1999 06:00:06 -0800 Received: from nbecker by hns.com with local (Exim 2.10 #1) id 10D7WD-0001sS-00 for autofs@linux.kernel.org; Wed, 17 Feb 1999 09:00:01 -0500 To: autofs@linux.kernel.org Subject: newbie questions Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII From: nbecker@fred.net Date: 17 Feb 1999 09:00:01 -0500 Message-ID: Lines: 7 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: 1) Do people use autofs to automount home directories? Any suggested setup? 2) How does autofs react to local (loopback) mounts? Will it create a symlink (ala amd) or and nfs loopback? 3) Is autofs being actively developed? From owner-autofs@linux.kernel.org Wed Feb 17 08:38:02 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA19980 for ; Wed, 17 Feb 1999 08:38:01 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA28007 for autofs-list; Wed, 17 Feb 1999 07:47:41 -0800 Received: from smtp.adc.com (smtp.adc.com [155.226.10.207]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA28003 for ; Wed, 17 Feb 1999 07:47:39 -0800 Received: from bbg.adc.com (wst564.bbg.adc.com [155.226.177.77]) by smtp.adc.com (8.9.0/8.9.0) with ESMTP id JAA09696 for ; Wed, 17 Feb 1999 09:47:38 -0600 (CST) Received: from adc.com (wst161 [155.226.177.184]) by bbg.adc.com (8.9.1/8.9.1) with ESMTP id JAA18013 for ; Wed, 17 Feb 1999 09:47:37 -0600 (CST) Message-ID: <36CAE498.134B026D@adc.com> Date: Wed, 17 Feb 1999 09:47:36 -0600 From: Peter Loveman X-Mailer: Mozilla 4.5 [en] (X11; U; SunOS 5.6 sun4m) X-Accept-Language: en MIME-Version: 1.0 To: autofs@linux.kernel.org Subject: shared objects Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I am trying to build autofs on a powerpc, and it builds fine, but when I run automount I get the following error: Feb 17 09:03:58 lyn108 automount[7790]: cannot open lookup module file (/usr/lib/autofs//lookup_file.so: undefined symbol: close_parse) Also, While I'm building I get the warning: gcc: unrecognized option `-rdynamic' Can anyone help me with this please? -peter -- _____ Peter Loveman | ADC Telecommunications |/// | peter_loveman@adc.com | Transport Systems Group | ADC| (972)680-4852 | 2240 Campbell Creek Blvd |_____| (972)680-7672 fax | Richardson, TX 75082 From owner-autofs@linux.kernel.org Wed Feb 17 08:45:04 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA19986 for ; Wed, 17 Feb 1999 08:45:04 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA27358 for autofs-list; Wed, 17 Feb 1999 07:30:37 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA27342 for ; Wed, 17 Feb 1999 07:30:17 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id QAA07403; Wed, 17 Feb 1999 16:28:36 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id QAA28063; Wed, 17 Feb 1999 16:28:36 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Wed, 17 Feb 1999 16:28:36 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: nbecker@fred.net Subject: RE: newbie questions Cc: autofs@linux.kernel.org Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thus spoke nbecker@fred.net on 17-Feb-99 : > 1) Do people use autofs to automount home directories? yes > Any suggested setup? auto.home can look like user1 lec:/lec/user/& user2 -rw,rsize=8192,wsize=8192,soft,intr lec:/lec/user/& user2 -rw,rsize=8192,wsize=8192,soft,intr star:/star/user/& > > 2) How does autofs react to local (loopback) mounts? Will it create a > symlink (ala amd) or and nfs loopback? symlinks ala amd > > 3) Is autofs being actively developed? yes FA "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 17 10:06:01 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA20031 for ; Wed, 17 Feb 1999 10:06:01 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA32483 for autofs-list; Wed, 17 Feb 1999 09:09:31 -0800 Received: from nenuphar.saclay.cea.fr (nenuphar.saclay.cea.fr [132.166.192.7]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA32479 for ; Wed, 17 Feb 1999 09:09:29 -0800 Received: from muguet.saclay.cea.fr (muguet.saclay.cea.fr [132.166.192.6]) by nenuphar.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id SAA14439 for ; Wed, 17 Feb 1999 18:10:28 +0100 (MET) Received: from dphdse.saclay.cea.fr (dphdse.saclay.cea.fr [132.166.30.5]) by muguet.saclay.cea.fr (8.9.1a/8.9.1/CEAnet-relay-5.0.0.1) with ESMTP id SAA01914; Wed, 17 Feb 1999 18:06:03 +0100 (MET) Received: from seipca185.saclay.cea.fr (seipca185 [132.166.29.185]) by dphdse.saclay.cea.fr (8.9.0/8.9.0) with SMTP id SAA31828; Wed, 17 Feb 1999 18:26:15 +0100 (MET) Date: Wed, 17 Feb 1999 18:26:15 +0100 (MET) Message-Id: <199902171726.SAA31828@dphdse.saclay.cea.fr> From: Jean-Paul Le Fevre To: peter_loveman@adc.com CC: autofs@linux.kernel.org In-reply-to: <36CAE498.134B026D@adc.com> (message from Peter Loveman on Wed, 17 Feb 1999 09:47:36 -0600) Subject: Re: shared objects References: <36CAE498.134B026D@adc.com> Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: The problem giving the warning is the cause of the failure of automount. Check the version of your gcc which should accept the option -rdynamic. -- ____________________________________________________________________________ Jean-Paul Le Fevre http://www.multimania.com/lefevre E-mail : Jean-Paul.LeFevre@cea.fr Tel : (33) [0]1 69 08 74 50 From owner-autofs@linux.kernel.org Wed Feb 17 10:35:11 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA20062 for ; Wed, 17 Feb 1999 10:35:10 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA01089 for autofs-list; Wed, 17 Feb 1999 09:39:07 -0800 Received: from tera.jprc.com (TERA.JPRC.COM [207.86.147.221]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id JAA01086 for ; Wed, 17 Feb 1999 09:39:05 -0800 Received: from THELIA.JPRC.COM (THELIA.JPRC.COM [207.86.147.157]) by tera.jprc.com (NTMail 3.03.0017/1.agyw) with ESMTP id pa328733 for ; Wed, 17 Feb 1999 12:45:43 -0500 Date: Wed, 17 Feb 1999 12:38:12 -0500 (EST) From: "James O'Kane" To: autofs@linux.kernel.org Subject: point of confusion In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I'm not sure I understand this part of the manual: ---------------- A * in the key field matches all keys. An example for the usefulness is the following entry: * &:/home/& This will enable you to access all the home directory of local hosts using the path /mountpoint/hostname/local- path. ---------------- I'm trying to convert my amd mounts to autofs. Currently, when I access /homes/jo2y it creates a symlink to /home/jo2y if I'm on my home machine or /net/thelia/home/jo2y if I'm on a remote machine. Is the above mentioned syntax what I want to look into? thanks -james From owner-autofs@linux.kernel.org Wed Feb 17 12:24:52 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA20136 for ; Wed, 17 Feb 1999 12:24:51 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id LAA07991 for autofs-list; Wed, 17 Feb 1999 11:48:11 -0800 Received: from smtp.adc.com (smtp.adc.com [155.226.10.207]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id LAA07986 for ; Wed, 17 Feb 1999 11:48:09 -0800 Received: from bbg.adc.com (wst564.bbg.adc.com [155.226.177.77]) by smtp.adc.com (8.9.0/8.9.0) with ESMTP id NAA01338 for ; Wed, 17 Feb 1999 13:48:07 -0600 (CST) Received: from adc.com (wst161 [155.226.177.184]) by bbg.adc.com (8.9.1/8.9.1) with ESMTP id NAA24661 for ; Wed, 17 Feb 1999 13:48:06 -0600 (CST) Message-ID: <36CB1CF5.524AAC4D@adc.com> Date: Wed, 17 Feb 1999 13:48:05 -0600 From: Peter Loveman X-Mailer: Mozilla 4.5 [en] (X11; U; SunOS 5.6 sun4m) X-Accept-Language: en MIME-Version: 1.0 To: autofs Subject: kernel support Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I finally got automount to compile, and when I run automount I get the following error: Feb 17 13:32:10 lyn108 automount[13174]: starting automounter version 3.1.3, path = /amount1, maptype = file, mapname = /root/build/autofs-3.1.3/daemon/mapfile Feb 17 13:32:11 lyn108 automount[13174]: >> mount: fs type autofs not supported by kernel Feb 17 13:32:11 lyn108 automount[13174]: /amount1: mount failed! Is there something missing in my kernel, or is there something I can change in configuration files? -peter -- _____ Peter Loveman | ADC Telecommunications |/// | peter_loveman@adc.com | Transport Systems Group | ADC| (972)680-4852 | 2240 Campbell Creek Blvd |_____| (972)680-7672 fax | Richardson, TX 75082 From owner-autofs@linux.kernel.org Wed Feb 17 13:10:29 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA20159 for ; Wed, 17 Feb 1999 13:10:29 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA10116 for autofs-list; Wed, 17 Feb 1999 12:22:28 -0800 Received: from snowy.nsw.cmis.CSIRO.AU (snowy.nsw.cmis.CSIRO.AU [130.155.16.108]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id MAA10109 for ; Wed, 17 Feb 1999 12:22:24 -0800 Received: by snowy.nsw.cmis.CSIRO.AU (SMI-8.6/SMI-SVR4) id HAA11166; Thu, 18 Feb 1999 07:22:08 +1100 Received: from edb.home.nsw.cmis.CSIRO.AU(130.155.16.90), claiming to be "vodka.gnarus.com" via SMTP by snowy.nsw.cmis.CSIRO.AU, id smtpdAAAa002iJ; Thu Feb 18 07:21:58 1999 Received: from vodka (really [192.168.110.10]) by edb.home.dms.csiro.au via smail with smtp (ident edb using rfc1413) id (Debian Smail3.2.0.101) for ; Thu, 18 Feb 1999 07:22:57 +1100 (EST) Message-ID: <36CB2521.5316282A@Altavista.net> Date: Wed, 17 Feb 1999 20:22:57 +0000 From: Ed Breen Organization: eic_free X-Mailer: Mozilla 3.04 (X11; I; Linux 2.0.34 i586) MIME-Version: 1.0 To: "James O'Kane" CC: autofs@linux.kernel.org Subject: Re: point of confusion References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: James O'Kane wrote: > > I'm not sure I understand this part of the manual: > > ---------------- > A * in the key field matches all keys. An example for the > usefulness is the following entry: > > * &:/home/& > > This will enable you to access all the home directory of > local hosts using the path /mountpoint/hostname/local- > path. I'm confused also, cause the only way it makes sense to me is via: * machine:/home/& Another problem is that auto.home seems to only allow one wild card entry. For example I would have thought the following would work, but I can't seem to get it to work: * machine1:/dir/& * machine2:/dir/& What appears to happen is the first entry gets accepted and 2nd gets ignored. Ed. -- Ed. Breen Email: Ed.Breen@Altavista.net (http://www.pobox.com/~eic) ----------------------------------------- From owner-autofs@linux.kernel.org Wed Feb 17 14:09:35 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA20198 for ; Wed, 17 Feb 1999 14:09:34 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA13928 for autofs-list; Wed, 17 Feb 1999 13:31:38 -0800 Received: from www2.netops.com (www2.netops.com [207.17.13.2]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA13924 for ; Wed, 17 Feb 1999 13:31:36 -0800 Received: from localhost (mandrake@localhost) by www2.netops.com (8.9.1/8.9.1) with ESMTP id QAA21496; Wed, 17 Feb 1999 16:26:53 -0500 Date: Wed, 17 Feb 1999 16:26:52 -0500 (EST) From: Isaac Ojeda To: Ed Breen cc: "James O'Kane" , autofs@linux.kernel.org Subject: Re: point of confusion In-Reply-To: <36CB2521.5316282A@Altavista.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: But would user machine1:/dir/& user machine2:/dir/& work if machine1 was down ? Thanks Isaac On Wed, 17 Feb 1999, Ed Breen wrote: > James O'Kane wrote: > > > > I'm not sure I understand this part of the manual: > > > > ---------------- > > A * in the key field matches all keys. An example for the > > usefulness is the following entry: > > > > * &:/home/& > > > > This will enable you to access all the home directory of > > local hosts using the path /mountpoint/hostname/local- > > path. > > I'm confused also, cause the only way it makes sense to > me is via: > > * machine:/home/& > > Another problem is that auto.home seems to only allow > one wild card entry. For example I would have thought > the following would work, but I can't seem to get it to work: > > * machine1:/dir/& > * machine2:/dir/& > > > What appears to happen is the first entry gets accepted > and 2nd gets ignored. > > Ed. > > -- > Ed. Breen > Email: Ed.Breen@Altavista.net > (http://www.pobox.com/~eic) > ----------------------------------------- > From owner-autofs@linux.kernel.org Wed Feb 17 15:16:24 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA20268 for ; Wed, 17 Feb 1999 15:16:24 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA17199 for autofs-list; Wed, 17 Feb 1999 14:40:40 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA17196 for ; Wed, 17 Feb 1999 14:40:39 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id OAA09198; Wed, 17 Feb 1999 14:40:02 -0800 From: Alvin Oga Message-Id: <199902172240.OAA09198@leopard.imedia.com> Subject: Re: kernel support To: peter_loveman@adc.com (Peter Loveman) Date: Wed, 17 Feb 1999 14:40:02 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), alvin@linux-consulting.com, autofs@linux.kernel.org In-Reply-To: <36CB1CF5.524AAC4D@adc.com> from "Peter Loveman" at Feb 17, 99 01:48:05 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi Peter... Make sure your kernel supports autofs a. cat /proc/filesystems - see if autofs is listed b. if autofs is NOT listed: - recompile your kernel w/ autofs support --OR-- -or- insmod autofs.o c. make sure you have autofs and the config files installed... - ls -l /usr/lib/autofs the autofs libraries - /etc/auto.{misc,master} or equivalents - make sure your autofs mount directory exists ( /amount1 in your case ?? ) - you already have the automount binary somewhere have fun alvin http://www.linux-consulting.com/Amd_AutoFS/autofs-HOWTO.html > I finally got automount to compile, and when I run automount I get the > following error: > > Feb 17 13:32:10 lyn108 automount[13174]: starting automounter version > 3.1.3, path = /amount1, maptype = file, mapname = > /root/build/autofs-3.1.3/daemon/mapfile > Feb 17 13:32:11 lyn108 automount[13174]: >> mount: fs type autofs not > supported by kernel > Feb 17 13:32:11 lyn108 automount[13174]: /amount1: mount failed! > > Is there something missing in my kernel, or is there something I can > change in configuration files? > -peter > -- > _____ Peter Loveman | ADC Telecommunications > |/// | peter_loveman@adc.com | Transport Systems Group > | ADC| (972)680-4852 | 2240 Campbell Creek Blvd > |_____| (972)680-7672 fax | Richardson, TX 75082 > From owner-autofs@linux.kernel.org Wed Feb 17 15:23:28 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA20273 for ; Wed, 17 Feb 1999 15:23:28 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA17813 for autofs-list; Wed, 17 Feb 1999 14:53:12 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA17810 for ; Wed, 17 Feb 1999 14:53:11 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id OAA11613; Wed, 17 Feb 1999 14:53:08 -0800 Received: from swtemp.transmeta.com (hpa@swtemp.transmeta.com [10.1.27.22]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id OAA24857; Wed, 17 Feb 1999 14:53:08 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by swtemp.transmeta.com (8.8.4/8.7.3) id OAA17671; Wed, 17 Feb 1999 14:53:08 -0800 Message-Id: <199902172253.OAA17671@swtemp.transmeta.com> Subject: Re: point of confusion In-Reply-To: from Lori at "Feb 17, 99 02:48:33 pm" To: lori@Frogtown.com (Lori) Date: Wed, 17 Feb 1999 14:53:08 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: > > > I think the syntax is something like: > > * machine1,machine2:/dir/& > > this way who ever answers first gets mounted. if one fails then you fail > to the other, if the dir can be unmounted, ie not busy. I believe it will > fail over if you use -ro > > can't mount both on the same mount point, anyway, and I belive the file is > read from top to bottom. > No, it is a (key,value) type lookup; therefore having two entries with the same key is just plain WRONG. The syntax above "should" work, but mount(8) and the Linux NFS subsystems don't support using more than one server. -hpa From owner-autofs@linux.kernel.org Wed Feb 17 15:30:24 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA20279 for ; Wed, 17 Feb 1999 15:30:24 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA17694 for autofs-list; Wed, 17 Feb 1999 14:50:54 -0800 Received: from scooter.frogtown.com (IDENT:root@robin.frogtown.com [209.220.40.226]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA17690 for ; Wed, 17 Feb 1999 14:50:53 -0800 Received: from robin.frogtown.com (lori@robin.frogtown.com [209.220.40.226]) by scooter.frogtown.com (8.8.7/8.8.7) with ESMTP id OAA05519 for ; Wed, 17 Feb 1999 14:48:35 -0800 Date: Wed, 17 Feb 1999 14:48:33 -0800 (PST) From: Lori X-Sender: lori@scooter.corp.frogtown.com cc: autofs@linux.kernel.org Subject: Re: point of confusion In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: I think the syntax is something like: * machine1,machine2:/dir/& this way who ever answers first gets mounted. if one fails then you fail to the other, if the dir can be unmounted, ie not busy. I believe it will fail over if you use -ro can't mount both on the same mount point, anyway, and I belive the file is read from top to bottom. cheers, On Wed, 17 Feb 1999, Isaac Ojeda wrote: > But would > > user machine1:/dir/& > user machine2:/dir/& > > work if machine1 was down ? > > Thanks > Isaac > > > On Wed, 17 Feb 1999, Ed Breen wrote: > > > James O'Kane wrote: > > > > > > I'm not sure I understand this part of the manual: > > > > > > ---------------- > > > A * in the key field matches all keys. An example for the > > > usefulness is the following entry: > > > > > > * &:/home/& > > > > > > This will enable you to access all the home directory of > > > local hosts using the path /mountpoint/hostname/local- > > > path. > > > > I'm confused also, cause the only way it makes sense to > > me is via: > > > > * machine:/home/& > > > > Another problem is that auto.home seems to only allow > > one wild card entry. For example I would have thought > > the following would work, but I can't seem to get it to work: > > > > * machine1:/dir/& > > * machine2:/dir/& > > > > > > What appears to happen is the first entry gets accepted > > and 2nd gets ignored. > > > > Ed. > > > > -- > > Ed. Breen > > Email: Ed.Breen@Altavista.net > > (http://www.pobox.com/~eic) > > ----------------------------------------- > > > From owner-autofs@linux.kernel.org Wed Feb 17 23:07:22 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id XAA20843 for ; Wed, 17 Feb 1999 23:07:22 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id WAA05173 for autofs-list; Wed, 17 Feb 1999 22:31:29 -0800 Received: from scooter.frogtown.com (IDENT:root@robin.frogtown.com [209.220.40.226]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id WAA05169 for ; Wed, 17 Feb 1999 22:31:26 -0800 Received: from animal.corp.frogtown.com (mblake@animal.corp.frogtown.com [10.0.0.9] (may be forged)) by scooter.frogtown.com (8.8.7/8.8.7) with ESMTP id WAA09828 for ; Wed, 17 Feb 1999 22:29:08 -0800 Date: Wed, 17 Feb 1999 22:31:11 -0800 (PST) From: Michael Newlyn Blake X-Sender: mblake@animal.corp.frogtown.com To: The autofs list Subject: Re: point of confusion In-Reply-To: <199902172253.OAA17671@swtemp.transmeta.com> Message-ID: X-Url: http://www.frogtown.com/~mblake MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: On Wed, 17 Feb 1999, H. Peter Anvin wrote: > > I think the syntax is something like: > > * machine1,machine2:/dir/& > No, it is a (key,value) type lookup; therefore having two entries with > the same key is just plain WRONG. Actually, I think she's refering to the way Solaris does autofs. The comma seperated list of hostnames is evaluated by autofs itself, not passed along to mount. The hostnames are passed along to seperate mount requests, providing a sort of failover. (For example, two nfs servers providing access to identical filesystems. If one stops responding, Solaris will try to force a remount from the secondary server.) A cool feature, but probably not supported in Linux autofs. (Or is it?) ----- Flirting is the gentle art of making a man feel pleased with himself. -- Helen Rowland From OceanBabe2@aol.com Thu Feb 18 08:32:09 1999 Return-Path: Received: from imo12.mx.aol.com (imo12.mx.aol.com [198.81.17.2]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA21425 for ; Thu, 18 Feb 1999 08:32:03 -0800 From: OceanBabe2@aol.com Received: from OceanBabe2@aol.com by imo12.mx.aol.com (IMOv18.1) id PIOBa04795; Thu, 18 Feb 1999 11:04:49 -0500 (EST) Message-ID: Date: Thu, 18 Feb 1999 11:04:49 EST To: alvarozuniga@earthlink.net, alvaslocomotion@hotmail.com, alvaslocomotion@supernews.com, alvat@delphi.com, alvaught@delphi.com, alvaz@hotmail.com, alvco@blueneptune.com, alvco@netcom.com, alvcpa@delphi.com, alvd@earthlink.net, alvd@ix.netcom.com, alvdetol@ccr.net, alvdog@pacbell.net, alve@alvenet.com, alve@mlstand.com, alveajo@ix.netcom.com, alven@edumedia.com, alven@pudknockers.com, alvena@eskimo.com, alvenee@topaz.hknet.com, alvenor@ix.netcom.com, alveoli@mail.ptd.net, alveoli@prolog.net, alveoli@ptdprolog.net, alver@ix.netcom.com, alvercom@webtv.net, alvergona@delphi.com, alvernaz@delphi.com, alverne@ix.netcom.com, alverno@iquest.net, alveroj@logica.com, alvers@hotmail.com, alvers7@imicom.net, alverson@delphi.com, alversond@delphi.com, alves@chez.com, alves@interlynx.net, alves@ix.netcom.com, alves@mlsonline.com, alves@mlstand.com, alves@stratus.com, alves@studio.disney.com, alves007@delphi.com, alvesp@cadvision.com, alvesp@vivanet.com, alvest@megalinx.my.net, alvest@megalinx.net, alvey@ausfish.com, alvey@genie.com, alveybs@delphi.com, alveys@intrepid.net, alvh@doitnow.com, alvi@ix.netcom.com, alvi1@ix.netcom.com, alviani@ix.netcom.com, alviani@mcs.com, alvic@bellsouth.net, alvic@snet.net, alvicari@delphi.com, alvickery@delphi.com, alvidrez@mtview.enet.dec.com, alvidrez@mtview.enetdec.com, alvidrez@tpflow.enet.dec.com, alvidrez@tpflow.enetdec.com, alvie@ix.netcom.com, alvier@hotmail.com, alvig@delphi.com, alvige@urgentmail.com, alvigil@airmail.net, alvigil@netcom.com, alvigo@mindspring.com, alvilla@earthlink.net, alvilla@worldnet.att.net, alvin@accent.net, alvin@accessv.com, alvin@aimnet.com, alvin@ari.net, alvin@azstarnet.com, alvin@c-square.com, alvin@chevalier.net, alvin@compass.com, alvin@cryogen.com, alvin@cyberramp.net, alvin@delphi.com, alvin@electrotex.com, alvin@eskimo.com, alvin@expressway.com, alvin@eyepoint.com, alvin@fremark.com, alvin@gaia.torolab.ibm.com, alvin@heart.net, alvin@hkg.com, alvin@hkstar.com, alvin@iglou.com, alvin@iplink.net, alvin@ix.netcom.com, alvin@linux-consulting.com, alvin@merlion.singnet.com, alvin@milkyway.infoimaging.com, alvin@neander.com, alvin@netcom.com, alvin@netib.com, alvin@netv.com, alvin@nmia.com, alvin@numeritech.com, alvin@picknowl.com, alvin@planet.fef.com, alvin@plato.nmia.com, alvin@politician.com, alvin@primenet.com, alvin@pwrnet.com, alvin@pyrotechnics.com, alvin@quick-checks.com, alvin@segel.com, alvin@singnet.com, alvin@spice.com, alvin@sprynet.com, alvin@stanton.cts.com, alvin@swcp.com, alvin@tandem.com, alvin@usxpress.com, alvin@webpro.com, alvin@winternet.com, alvin@worldchat.com, alvin@zany.mti.sgi.com, alvin.carter@filebank.cts.com, alvin.clark@mailport.delta-air.com, alvin.n@usa.net, alvin.r.mills@mci.com, alvin.wilson@giffer.gifl.com, alvin01@mbox3.singnet.com, alvin04@sprynet.com, alvin1@delphi.com, alvin1@idt.net, alvin1@mbox2.singnet.com, alvin1.3@worldnet.att.net, alvin11@ix.netcom.com, alvin123@netvigator.com, alvin133@delphi.com, alvin185@delphi.com, alvin2@erols.com, alvin21@ms18.hinet.net, alvin3340@delphi.com, alvin44@hotmail.com, alvin5@ix.netcom.com, alvin5609@delphi.com, alvin5901@delphi.com, alvin7@ix.netcom.com, alvin8@ix.netcom.com, alvin88@ms2.hinet.net, alvin9@ix.netcom.com, alvin_chen@idecc.com, alvin_crofts@stratus.com, alvin_hernandez@nugraphix.com, alvin_koh@pts.mot.com, alvin_m_hall@msn.com, alvin_mootoo@msn.com, alvin_plummer@nt.com, alvin_tso@hotmail.com, alvina@ari.net, alvina@delphi.com, alvina@genie.com, alvina@gordon.com, alvina@hamlin.com, alvina@leesburg.com, alvina@nallen.com, alvina@oakland.com, alvinam@worldnet.att.net, alvinau@websuite.com, alvinb@com.com, alvinb@delphi.com, alvinb@mindspring.com, alvinb@mindsprlng.com, alvinb@msn.com, alvinb2@ix.netcom.com, alvinb2@netcom.com, alvinbattrick@geocities.com, alvinbb@prodigy.com, alvinc@baynetworks.com, alvinc@delphi.com, alvinc@icsi.net, alvinc@jumppoint.com, alvinc@tanisys.com, alvinc@total.net, alvinc@ut.airiger.com, alvince@linkline.com, alvince@stbbs.com, alvinchang@delphi.com, alvinchua@post1.com, alvinchua@usa.net, alvincmc@concentric.net, alvincys@news.hk.super.net, alvindavid@delphi.com, alvindavis@delphi.com, alvindelgado@mindspring.com, alvine@hprnd.rose.hp.com, alvinez@delphi.com, alvinfan@netvigator.com, alving@infocomm.net, alvingan@singnet.com, alvinh@delphi.com, alvinh@pwrnet.com, alvinh@teleport.com, alvinho@snp.com, alvinj@iglou.com, alvinj@ix.netcom.com, alvink@aloha.net, alvink@delphi.com, alvink@haleakala.aloha.net, alvink@ipoline.com, alvinkearse@delphi.com, alvinkei@netvigator.com, alvinklein@delphi.com, alvinkoh@singnet.com, alvinkoh@writeme.com, alvinksu@hotmail.com, alvinkuo@ix.netcom.com, alvinl@delphi.com, alvinl@genie.com, alvinl@stlo.smhs.com, alvinlaw@net.com, alvinlaw@netcom.com, alvinlaw@netcom23.netcom.com, alvinlaw@netcom4.netcom.com, alvinlee@asiaonline.net, alvinlee@cyberec.com, alvinlee@hkstar.com, alvinlhanks@remote.occ.com, alvinli@delphi.com, alvinli@tcts1.globalnet.com, alvinlim@asia-ads.com, alvinlo@hotmail.com, alvinlo@is-dept.com, alvinlow@mailhost.net, alvinlu@geocities.com, alvinlwh@cyberway.com, alvinm@tyrell.net, alvinmak@asiaonline.net, alvinmak@netvigator.com, alvinmak@wlink.net, alvinnkc@singnet.com, alvino@pwrnet.com, alvinp@hotmail.com, alvinpan@hkstar.com, alvins@bga.com, alvins@realtime.com, alvinsee@mbox2.singnet.com, alvinsee@singnet.com, alvinsh@hotmail.com, alvinsiu@hkstar.com, alvint@earthlink.net, alvinteo@asia-ads.com, alvintrotman@delphi.com, alvinx@ut.airtiger.com, alviny@delphi.com, alvirez@dallas.net, alvis@toast.net, alvis@worldnet.att.net, alvis1@ix.netcom.com, alvision@gworld.net, alvision@hkstar.com, alvislives@webtv.net, alvism@wingsbbs.com, alvisoca@juno.com, alvisw@erols.com, alvitar@typhoon.liveonthenet.com, alvitar@cyclone.traveller.com, alvitar@traveller.com, alvitar@xavax.com, alviti@ids.net, alviti@lanline.com, alvk@post1.com, alvlim@ix.netcom.com, alvmor@delphi.com, alvo@spdf.dnet.hac.com, alvo@spdf.dnethac.com, alvon@delphi.com, alvon@flash.net, alvon@ibm.net, alvonderhaar@genie.com, alvorada@citynet.net, alvord@cia-g.com, alvord@interaccess.com, alvord@lightspeed.net, alvord@tek.com, alvorne@maui.net, alvp@sprynet.com, alvp@usa.pipeline.com, alvsr@worldnet.net, alvust@hotmail.com, alvw@ix.netcom.com, alvy@ix.netcom.com, alvydas_a@equinoxlt.com, alw@al-williams.com, alw@delphi.com, alw@enter.net, alw@execpc.com, alw@friend.ly.net, alw@impulse.net, alw@livnet.com, alw@lsrlaw.com, alw@ncrbeth.bethlehempa.ncr.com, alw@netcom.com, alw@philabs.philips.com, alw@sequent.com, alw1@delphi.com, alw1@netcom.com, alw1900@delphi.com, alw9000@delphi.com, alw985@ix.netcom.com, alwaechter@hq.ensco.com, alwagner@delphi.com, alwagner@mcs.com, alwahl@delphi.com, alwaite@erols.com, alwakeel@delphi.com, alwalker@junosystems.com, alwaller@shore.intercom.net, alwalsh@ibm.net, alwalton@delphi.com, alwanderski@delphi.com, alwang@doubt.com, alwankl@hkstar.com, alward@agate.net, alward@london.sbi.com, alward@simware.com, alwarner@idt.net, alwarr@msn.com, alwarren@delphi.com, alwas@gte.net, alwassat@post1.com, alwatson@sedona.net, alway@accesscom.com, alway@delphi.com, alway@goodnet.com, alway@gte.net, alway@ix.netcom.com, alway@netscape.com, alwayda@morganfinnegan.com, alwayne@delphi.com, alwayne@suba.com, always@alwaysontop.com, always@beaves.com, always@best.com, always@faithful.com, always@harddrive.net, always@hknet.com, always@home.com, always@ix.netcom.com, always@lookin.com, always@lucky.asiainvestments.com, always@mailmasher.com, always@serv.net, always@spacestar.com, always@the.com, always@thefridge.com, always@witty.com, always@wolfenet.com, always@work.com, always-live@spankovision.com, always1@ix.netcom.com, always1@ms10.hinet.net, always_free@earthling.net, always_ready@hotmail.com, alwaysan@earthling.net, alwayscc@ix.netcom.com, alwayscc@mindspring.com, alwayscc@prodigy.com, alwaysflysolo@hotmail.com, alwaysintouch@hotmail.com, alwayski@juno.com, alwayslive@assfuck.com, alwayslive@xxfuck.com, alwayslookin@msn.com, alwaysme@delphi.com, alwaysnice@delphi.com, alwayspolite@hotmail.com, alwaysup@delphi.com, alwayswearprotectivegear@earthling.net, alwayswin@msn.com, alwayz@primenet.com, alwayz@writeme.com, alwazstrmy@computershoppe.com, alweaver@delphi.com, alwebb@delphi.com, alwebb@genie.com, alweekes@audiophile.com, alwegener@specwire.com, alwei@hotmail.com, alwei@ms13.hinet.net, alweiner@alweiner.clark.net, alweiner@clark.net, alweiner@presstar.com, alweiner@printing.presstar.com, alweiner@ricochet.net.com, alweiss@interlog.com, alweiss@jetlink.net, alwen@bangornews.infi.net, alwerder@delphi.com, alwerder@worldnet.att.net, alwes@swbell.net, alwest@delphi.com, alwest@netcom.com, alweston@ogr.com, alweston@pixi.com, alwet@pacbell.net, alwexpre@capecod.net, alwhalen@delphi.com, alwhipp@gil.com, alwhipp@mail.ipswich.gil.com, alwhitaker@delphi.com, alwhite@emh1.pa.net, alwhite@sparc.isl.net, alwhite@mail.cvn.net, alwhitehurst@delphi.com, alwhitt@mich.com, alwi@ananke.com, alwi@digntn1.csgi.com, alwicher@pi.net, alwickman@delphi.com, alwiensc@mail.netshop.net, alwiensc@netshop.net, alwilcox@delphi.com, alwild@gate.net, alwilhel@mail.delcoelect.com, alwilk@ix.netcom.com, alwilkno@shore.intercom.net, alwill@bconnex.net, alwill@netcom.com, alwilli@iafrica.com, alwillis@brokersys.com, alwills@ix.netcom.com, alwills@rconnect.com, alwilson@cpqm.saic.com, alwilson@ix.netcom.com, alwimberly@delphi.com, alwin@cld9.com, alwin@delphi.com, alwin@ix.netcom.com, alwin@nynexst.com, alwin@onlink.net, alwin@vaxxine.com, alwin.dieperink@elca.com, alwin90@delphi.com, alwinch@com.com, alwinch@delphi.com, alwine@xensei.com, alwinli@delphi.com, alwinv@edmark.com, alwinv@pstbbs.com, alwis@csgi.com, alwis@digntn1.csgi.com, alwis_adrianhohlwegaahohlw@kakadu3.com, alwiss@dave-world.net, alwith@delphi.com, alwithoff@delphi.com, alwito@icanect.net, alwlswv@delphi.com, alwnew@psbnewton.com, alwo@bigfoot.com, alwo@eagle.dinet.com, alwofford@delphi.com, alwohl@winternet.com, alwolf@delphi.com, alwolfe@iactivebbs.com, alwong@aznet.net, alwong@interlog.com, alwong@netvigator.com, alwong@uniserve.com, alwong@value.net, alwoo@alacritas.com, alwood@hiwaay.net, alwood@netcom.com, alwoods@delphi.com, alwoodward@delphi.com, alwready@hotmail.com, alwreed@buba9.ip.portal.com, alwri@idt.net, alwright@concentric.net, alwrite@delphi.com, alwshthy@sonic.net, alwsome@msn.com, alwt71a@prodigy.com, alwu@epc.lmms.lmco.com, alwu@esd.sgi.com, alwu@sunfse.ese.lmsc.lockheed.com, alwyn@ari.net, alwyn@inforamp.net, alwyn@netvigator.com, alwynallan@prodigy.net, alwynh@caribsurf.com, alwynne@ozemail.com, alwys420@ix.netcom.com, alwyse@com.com, alwyse@delphi.com, alwzbusy@ix.netcom.com, alx@access.mountain.net, alx@alx.com, alx@earth.execpc.com, alx@elvis.sovusa.com, alx@erols.com, alx@execpc.com, alx@gte.net, alx@hotwired.com, alx@i2020.net, alx@ikp.atm.com, alx@ix.netcom.com, alx@linex.com, alx@soft.hotwired.com, alx@speednet.com, alx@webtv.net, alx.pavl@ix.netcom.com, alx1@kaman.com, alx1275@hotmail.com, alx2@ix.netcom.com, alx3@ix.netcom.com, alx94@delphi.com, alx_97@hotmail.com, alxander@axionet.com, alxander@nyc.pipeline.com, alxander@panix.com, alxander@pipeline.com, alxander@spamfree.pipeline.com, alxander@usa.pipeline.com, alxandr@ibm.net, alxandra@pacbell.net, alxar@infinet.com, alxberg@delphi.com, alxcheung@hknet.com, alxdgreat@delphi.com, alxdulage@death.com, alxevans@concentric.net, alxgebs@ftn.net, alxgomez@tricom.net, alxh@delphi.com, alxh41a@prodigy.com, alxinfo@delphi.com, alxis@netpass.com, alxj60a@prodigy.com, alxko@hotmail.com, alxkuper@centurion.flash.net, alxl77a@prodigy.com, alxlan@ace.com, alxlan@world.std.com, alxlewis@bellsouth.net, alxlozano@delphi.com, alxluzv@wwonline.com, alxmar@com.com, alxmar@delphi.com, alxmdc4226@delphi.com, alxmelis@gate.net, alxndr@com.com, alxndr@delphi.com, alxndr@mindspring.com, alxndr@pe.net, alxnet@deltanet.com, alxny@shani.net, alxor@world.com, alxshing@netvigator.com, alxsl@msn.com, alxtech@delphi.com, alxthorn@ix.netcom.com, alxx@myself.com, aly@bdt.com, aly@fugawi.com, aly@hotmail.com, aly@ix.netcom.com, aly@merlin.dev.cdx.mot.com, aly@niven.imsweb.net, aly@somtel.com, aly@unforgettable.com, aly@usa.net, aly1@ix.netcom.com, aly1@netcom.com, aly33@why.net, aly8888@msn.com, alya@ix.netcom.com, alya@paragraph.com, alya1@ix.netcom.com, alyafei@hotmail.com, alyajss@ibm.net, alyal@delphi.com, alyal@hotmail.com, alyana@bga.com, alyau@ix.netcom.com, alyb@ix.netcom.com, alybooks@epix.net, alybotz@worldnet.att.net, alycas@singnet.com, alycat@slip.net, alyce@kool.studio.sgi.com, alyce_hatch@bendnet.com, alyceg@vegas.quik.com, alycek@sundial.net, alycem@msn.com, alycew@delphi.com, alychan@asiaonline.net, alycia@ari.net, alycia@delphi.com, alyclepal@mindspring.com, alyda@iquest.net, alydaar@mailbox.neosoft.com, alydar@cnct.com, alye6048@mlb.com, alyen@newjersey.sptc.hdshq.com, alyeska@execpc.com, alyeska@ix.netcom.com, alyeska@m1.interserv.com, alyfong@netcom.com, alyg@delphi.com, alyg28a@prodigy.com, alyh95a@prodigy.com, alyi@chevron.com, alykat@alykat.com, alykat@geocities.com, alykat@midcoat.com, alykat@pobox.com, alykat@thepentagon.com, alyke@mbay.net, alykhan_virani@mindlink.net, alykken@delphi.com, alyles@mindspring.com, alyman@delphi.com, alyman1@delphi.com, alymburn@iitworld.com, alymike@msn.com, alyn@kirov.com, alyn@netcom.com, alyn@sirius.com, alyn@wolfenet.com, alyn.ward@amd.com, alyn3@nwlink.com, alyn55a@prodigy.com, alyn71a@prodigy.com, alyn_n_jante@sbphrd.com, alyn_n_jante%notes@sb.com, alyn_n_jantenotes@sb.com, alyna@delphi.com, alynch@getonthe.net, alynch@ix.netcom.com, alynch@netcom.com, alynchawol@worldnet.att.net, alynda@webtv.com, alyne@harris.com, alynfw@ix.netcom.com, alynn@pgh.nauticom.net, alynn@bmcd.com, alynn@concentric.net, alynn@east.delfin.com, alynn@maxim.com, alynn@mich.com, alynn@nauticom.net, alynn@pgh.nauticom.net, alynn@rzdn.com, alynn@shl.com, alynn@sktp.com, alynn@svq-net.com, alynn_tanya@msn.com Mime-Version: 1.0 Subject: Would you like to pay off your bills? Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7bit X-Mailer: AOL 4.0 for Windows 95 sub 205 Status: RO X-Status: TAKE 10 MINUTES TO READ THIS...IT COULD CHANGE YOUR LIFE!!! EARN $100,000 PER YEAR SENDING E-MAIL!!! Dear Friend, You can earn $50,000 or more in the next 90 days sending e-mail!! Seem impossible? Read on for details (no, there is no 'catch')... "AS SEEN ON NATIONAL T.V." Thank you for your time and interest. PLEASE CONTINUE READING. This is the letter you've been reading about in the news lately. Due to the popularity of this letter on the Internet, a major nightly news program recently devoted an entire show to the investigation of the program described below, to see if it really can make people money. The show also investigated whether or not the program was legal. Their findings proved once and for all that there are, absolutely no laws prohibiting the participation in the program. This has helped to show people that this is a simple, harmless and fun way to make some extra money at home. The results of this show have been truly remarkable. So many people are participating that those involved are doing much better than before. Since everyone makes more as more people try it out, its been very exciting to be a part of lately. You will understand once you experience it. "HERE IT IS BELOW" **Print This Now For Future Reference** The following income opportunity is one you may be interested in taking a look at. It can be started with VERY LITTLE investment and the income return is TREMENDOUS!!! If you would like to make at least $50,000 in less than 90 days, please read the enclosed program...THEN READ IT AGAIN!!! THIS IS A LEGITIMATE, LEGAL, MONEY MAKING OPPORTUNITY. It does not require you to come into contact with people, do any hard work, and best of all, you never have to leave the house except to get the mail. If you believe that someday you'll get that big break that you've been waiting for, THIS IS IT! Simply follow the instructions, and your dreams will come true. This multi-level e-mail order marketing program works perfectly...100% EVERY TIME. E-mail is the sales tool of the future. Take advantage of this non-commercialized method of advertising NOW!! The longer you wait, the more people will be doing business using e-mail. Get your piece of this action!!! MULTI-LEVEL MARKETING (MLM) has finally gained respectability. It is being taught in the Harvard Business School, and both Stanford Research and the Wall Street Journal have stated that between 50% and 65% of all goods and services will be sold through multi-level methods by the mid to late 1990's. This is a Initially, I sent out 10,000 e-mails. It cost me about $15 for my time on-line. The great thing about e-mail is that I don't need any money for printing to send out the program, and because all of my orders are fulfilled via e-mail, the only expense is my time. I am telling you like it is, I hope it doesn't turn you off, but I promised myself that I would not "rip-off" anyone, no matter how much money it cost me. In less than one week, I was starting to receive orders for REPORT #1. By January 13, I had received 26 orders for REPORT #1. Your goal is to "RECEIVE at least 20 ORDERS FOR REPORT #1 WITHIN 2 WEEKS. IF YOU DON'T, SEND OUT MORE PROGRAMS UNTIL YOU DO!" My first step in making $50,000 in 90 days was done. By January 30, I had received 196 orders for REPORT #2. Your goal is to "RECEIVE AT LEAST 100+ ORDERS FOR REPORT #2 WITHIN 2 WEEKS. IF NOT, SEND OUT MORE PROGRAMS UNTIL YOU DO. ONCE YOU HAVE 100 ORDERS, THE REST IS EASY, RELAX, YOU WILL MAKE YOUR $50,000 GOAL." Well, I had 196 orders for REPORT #2, 96 more than I needed. So I sat back and relaxed. By March 1, of my e-mailing of 10,000, I received $58,000 with more coming in every day. I paid off ALL of my debts and bought a much needed new car. Please take time to read the attached program, IT WILL CHANGE YOUR LIFE FOREVER!!! Remember, it won't work if you don't try it. This program does work, but you must follow it EXACTLY! Especially the rules of not trying to place your name in a different place. It won't work, you'll lose out on a lot of money! In order for this program to work, you must meet you goal of 20+ orders for REPORT #1, and 100+ orders for REPORT #2 and you will make $50,000 or more in 90 days. I AM LIVING PROOF THAT IT WORKS!!! If you choose not to participate in this program, I am sorry. It really is a great opportunity with little cost or risk to you. If you choose to participate, follow the program and you will be on your way to financial security. If you are a fellow business owner and are in financial trouble like I was, or you want to start your own business, consider this a sign. I DID! Sincerely, Johnathon Rourke P.S. Do you have any idea what 11,700 $5 bills (58,000) look like piled up on a kitchen table? IT'S AWESOME! A PERSONAL NOTE FROM THE ORIGINATOR OF THIS PROGRAM: By the time you have read the enclosed program and reports, you should have concluded that such a program, and one that is legal, could not have been created by an amateur. Let me tell you a little about myself. I had a profitable business for 10 years. Then in 1979 my business began falling off. I was doing the same things that were previously successful for me, but it wasn't working. Finally, I figured it out. It wasn't me, it was the economy. Inflation and recession had replaced the stable economy that had been with us since 1945. I don't have to tell you what happened to the unemployment rate... because many of you know from first hand experience. There were more failures and bankruptcies than ever before. The middle class was vanishing. Those who knew that they were doing invested wisely and moved up. Those who did not, including those who never had anything to save or invest, were moving down into the ranks of the poor. As the saying goes, "THE RICH GET RICHER AND THE POOR GET POORER". The traditional methods of making money will never allow you to "move up" or "get rich", inflation will see to that. You have just received information that can give you financial freedom for the rest of your life, with "NO RISK" and "JUST A LITTLE BIT OF EFFORT." You can make more money in the next few months than you have ever imagined. I should also point out that I will not see a penny of this money, nor anyone else who has provided a testimonial for this program. I have already made over 4 MILLION DOLLARS! I have retired from the program after sending out over 16,000 programs. Now I have several offices that make this and several other programs here and over seas. Follow the program EXACTLY AS INSTRUCTED. Do not change it in any way. It works exceedingly well as it is now. Remember to e-mail a copy of this exciting report to everyone you can think of. One of the people you send this to may send out 50,000...and your name will be on everyone of them! Remember though, the more you send out the more potential customers you will reach. So my friend, I have given you the ideas, information, materials and opportunity to become financially independent, IT IS UP TO YOU NOW! "THINK ABOUT IT" Before you delete this program from your mailbox, as I almost did, take a little time to read it and REALLY THINK ABOUT IT. Get a pencil and figure out what could happen when YOU participate. Figure out the worst possible response and no matter how you calculate it, you will still make a lot of money! You will definitely get back what you invested. Any doubts you have will vanish when your first orders come in. IT WORKS! Jody Jacobs, Richmond, VA HERE'S HOW THIS AMAZING PROGRAM WILL MAKE YOU THOUSANDS OF DOLLAR$ INSTRUCTIONS: This method of raising capital REALLY WORKS 100% EVERY TIME. I am sure that you could use up to $50,000 or more in the next 90 days. Before you say "BULL... ", please read this program carefully. This is not a chain letter, but a perfectly legal money making opportunity. Basically, this is what you do: As with all multi-level businesses, we build our business by recruiting new partners and selling our products. Every state in the USA allows you to recruit new multi-level business partners, and we offer a product for EVERY dollar sent. YOUR ORDERS COME BY MAIL AND ARE FILLED BY E-MAIL, so you are not involved in personal selling. You do it privately in your own home, store or office. This is the GREATEST Multi-Level Mail Order Marketing anywhere: This is what you MUST do: 1. Order all 4 reports shown on the list below (you can't sell them if you don't order them). * For each report, send $5.00 CASH, the NAME & NUMBER OF THE REPORT YOU ARE ORDERING, YOUR E-MAIL ADDRESS, and YOUR NAME & RETURN ADDRESS (in case of a problem) to the person whose name appears on the list next to the report. MAKE SURE YOUR RETURN ADDRESS IS ON YOUR ENVELOPE IN CASE OF ANY MAIL PROBLEMS! * When you place your order, make sure you order each of the four reports. You will need all four reports so that you can save them on your computer and resell them. * Within a few days you will receive, via e-mail, each of the four reports. Save them on your computer so they will be accessible for you to send to the 1,000's of people who will order them from you. 2. IMPORTANT-- DO NOT alter the names of the people who are listed next to each report, or their sequence on the list, in any way other than is instructed below in steps "a" through "f" or you will lose out on the majority of your profits. Once you understand the way this works, you'll also see how it doesn't work if you change it. Remember, this method has been tested, and if you alter it, it will not work. a. Look below for the listing of available reports. b. After you've ordered the four reports, take this advertisement and remove the name and address under REPORT #4. This person has made it through the cycle and is no doubt counting their $50,000! c. Move the name and address under REPORT #3 down to REPORT #4. d. Move the name and address under REPORT #2 down to REPORT #3. e. Move the name and address under REPORT #1 down to REPORT #2. f. Insert your name/address in the REPORT #1 position. Please make sure you copy every name and address ACCURATELY! 3. Take this entire letter, including the modified list of names, and save it to your computer. Make NO changes to the instruction portion of this letter. Your cost to participate in this is practically nothing (surely you can afford $20). You obviously already have an Internet connection and e-mail is FREE! There are two primary methods of building your downline: METHOD #1: SENDING BULK E-MAIL Let's say that you decide to start small, just to see how it goes, and we'll assume you and all those involved send out only 2,000 programs each. Let's also assume that the mailing receives a 0.5% response. Using a good list the response could be much better. Also, many people will send out hundreds of thousands of programs instead of 2,000. But continuing with this example, you send out only 2,000 programs. With a 0.5% response, that is only 10 orders for REPORT #1. Those 10 people respond by sending out 2,000 programs each for a total of 20,000. Out of those 0.5%, 100 people respond and order REPORT #2. Those 100 mail out 2,000 programs each for a total of 200,000. The 0.5% response to that is 1,000 orders for REPORT #3. Those 1,000 send out 2,000 programs each for a 2,000,000 total. The 0.5% response to that is 10,000 orders for REPORT #4. That's 10,000 $5 bills for you. CASH!!! Your total income in this example is $50 + $500 + $5,000 + $50,000 for a total of $55,550!!! REMEMBER FRIEND, THIS IS ASSUMING 1,990 OUT OF THE 2,000 PEOPLE YOU MAIL TO WILL DO ABSOLUTELY NOTHING AND TRASH THIS PROGRAM! DARE TO THINK FOR A MOMENT WHAT WOULD HAPPEN IF EVERYONE, OR HALF SENT OUT 100,000 PROGRAMS INSTEAD OF 2,000. Believe me, many people will do just that, and more! By the way, your cost to participate in this is practically nothing. You obviously already have an internet connection and e-mail is FREE!!! REPORT #2 will show you the best methods for bulk e-mailing, tell you where to obtain free bulk e-mail software and where to obtain e-mail lists. METHOD #2 - PLACING FREE ADS ON THE INTERNET 1. Advertising on the 'Net is very, very inexpensive, and there are HUNDREDS of FREE places to advertise. Let's say you decide to start small just to see how well it works. Assume your goal is to get ONLY 10 people to participate on your first level. (Placing a lot of FREE ads on the internet will EASILY get a larger response.) Also assume that everyone else in YOUR ORGANIZATION gets ONLY 10 downline members. Follow this example to achieve the STAGGERING results below. 1st level--your 10 members with $5.....................................$50 2nd level--10 members from those 10 ($5 x 100)...............$500 3rd level--10 members from those 100 ($5 x 1,000)........$5,000 4th level--10 members from those 1,000 ($5 x 10,000)..$50,000 THIS TOTALS -----------------------$55,550 Remember friends, this assumes that the people who participate only recruit 10 people each. Think for a moment what would happen if they got 20 people to participate! Most people get 100's of participants! THINK ABOUT IT! For every $5.00 you receive, all you must do is e-mail them the report they ordered. THAT'S IT! ALWAYS PROVIDE SAME-DAY SERVICE ON ALL ORDERS! This will guarantee that the e-mail THEY send out, with YOUR name and address on it, will be prompt because they can't advertise until they receive the report! ------------------------------------------ AVAILABLE REPORTS ------------------------------------------ *** Order Each REPORT by NUMBER and NAME *** Notes: - ALWAYS SEND $5 CASH (U.S. CURRENCY) FOR EACH REPORT CHEQUES NOT ACCEPTED - ALWAYS SEND YOUR ORDER VIA FIRST CLASS MAIL - Make sure the cash is concealed by wrapping it in at least two sheets of paper - On one of those sheets of paper, include: (a) the number & name of the report you are ordering, (b) your e-mail address, and (c) your name & postal address. PLACE YOUR ORDER FOR THESE REPORTS NOW: ______________________________________________________ REPORT #1 "The Insider's Guide to Usenet Newsgroups and Classified Sites." ORDER REPORT #1 FROM: Lone Eagle 2 Kiel Avenue #286 Kinnelon, NJ 07405 ______________________________________________ REPORT #2 "How to Strike Gold in network Marketing." ORDER REPORT #2 FROM: Tom Pederson 9701 Forrest City RD. Altamonte Springs, FL. 32714 ______________________________________________ REPORT #3 "The Authoritative Internet Manual." ORDER REPORT #3 FROM: Bowling Syndicate P.O. Box181114 Casselberry, FL. 32718-1114 ________________________________________________ REPORT #4 "Own Your Own Million Dollar Corporation in as Little as 4 Weeks." ORDER REPORT #4 FROM: Edith Hiroto 148 Sausalito Blvd. #130 Casselberry, FL.32707 _______________________________________________ About 50,000 new people get online every month! ******* TIPS FOR SUCCESS ******* * TREAT THIS AS YOUR BUSINESS! Be prompt, professional, and follow the directions accurately. * Send for the four reports IMMEDIATELY so you will have them when the orders start coming in because: When you receive a $5 order, you MUST send out the requested product/report. * ALWAYS PROVIDE SAME-DAY SERVICE ON THE ORDERS YOU RECEIVE. * Be patient and persistent with this program. If you follow the instructions exactly, your results WILL BE SUCCESSFUL! * ABOVE ALL, HAVE FAITH IN YOURSELF AND KNOW YOU WILL SUCCEED! ******* YOUR SUCCESS GUIDELINES ******* Follow these guidelines to guarantee your success: If you don't receive 20 orders for REPORT #1 within two weeks, continue advertising or sending e-mails until you do. Then, a couple of weeks later you should receive at least 100 orders for REPORT#2. If you don't, continue advertising or sending e-mails until you do. Once you have received 100 or more orders for REPORT #2, YOU CAN RELAX, because the system is already working for you, and the cash will continue to roll in! THIS IS IMPORTANT TO REMEMBER: Every time your name is moved down on the list, you are placed in front of a DIFFERENT report. You can KEEP TRACK of your PROGRESS by watching which report people are ordering from you. If you want to generate more income, send another batch of e-mails or continue placing ads and start the whole process again! There is no limit to the income you will generate from this business! Before you make your decision as to whether or not you participate in this program. Please answer one question. DO YOU WANT TO CHANGE YOUR LIFE? If the answer is yes, please look at the following facts about this program: 1. YOU ARE SELLING A PRODUCT WHICH DOES NOT COST ANYTHING TO PRODUCE! 2. YOU ARE SELLING A PRODUCT WHICH DOES NOT COST ANYTHING TO SHIP! 3. YOU ARE SELLING A PRODUCT WHICH DOES NOT COST YOU ANYTHING TO ADVERTISE! 4. YOU ARE UTILIZING THE POWER OF THE INTERNET AND THE POWER OF MULTI-LEVEL MARKETING TO DISTRIBUTE YOUR PRODUCT ALL OVER THE WORLD! 5.. YOUR ONLY EXPENSES OTHER THAN YOUR INITIAL $20 INVESTMENT IS YOUR TIME! 6. VIRTUALLY ALL OF THE INCOME YOU GENERATE FROM THIS PROGRAM IS PURE PROFIT! 7. THIS PROGRAM WILL CHANGE YOUR LIFE FOREVER. ******* T E S T I M O N I A L S ******* This program does work, but you must follow it EXACTLY! Especially the rule of not trying to place your name in a different position, it won't work and you'll lose a lot of potential income. I'm living proof that it works. It really is a great opportunity to make relatively easy money, with little cost to you. If you do choose to participate, follow the program exactly, and you'll be on your way to financial security. Steven Bardfield, Portland, OR My name is Mitchell. My wife, Jody, and I live in Chicago, IL. I am a cost accountant with a major U.S. Corporation and I make pretty good money. When I received the program I grumbled to Jody about receiving "junk mail." I made fun of the whole thing, spouting my knowledge of the population and percentages involved. I "knew" it wouldn't work. Jody totally ignored my supposed intelligence and jumped in with both feet. I made merciless fun of her, and was ready to lay the old "I told you so" on her when the thing didn't work... well, the laugh was on me! Within two weeks she had received over 50 responses. Within 45 days she had received over $147,200 in $5 bills! I was shocked! I was sure that I had it all figured and that it wouldn't work. I AM a believer now. I have joined Jody in her "hobby." I did have seven more years until retirement, but I think of the "rat race" and it's not for me. We owe it all to MLM. Mitchell Wolf MD., Chicago, IL The main reason for this letter is to convince you that this system is honest, lawful, extremely profitable, and is a way to get a large amount of money in a short time. I was approached several times before I checked this out. I joined just to see what one could expect in return for the minimal effort and money required. To my astonishment, I received $36,470.00 in the first 14 weeks, with money still coming in. Sincerely yours, Charles Morris, Esq. Not being the gambling type, it took me several weeks to make up my mind to participate in this plan. But conservative that I am, I decided that the initial investment was so little that there was just no way that I wouldn't get enough orders to at least get my money back. Boy, was I surprised when I found my medium-size post office box crammed with orders! For awhile, it got so overloaded that I had to start picking up my mail at the window. I'll make more money this year than any 10 years of my life before. The nice thing about this deal is that it doesn't matter where people live. There simply isn't a better investment with a faster return. Paige Willis, Des Moines, IA I had received this program before. I deleted it, but later I wondered if I shouldn't have given it a try. Of course, I had no idea who to contact to get another copy, so I had to wait until I was e-mailed another program, ...11 months passed then it came...I didn't delete this one!...I made more than $41,000 on the first try!! Violet Wilson, Johnstown, PA This is my third time to participate in this plan. We have quit our jobs, and will soon buy a home on the beach and live off the interest on our money. The only way on earth that this plan will work for you is if you do it. For your sake, and for your family's sake don't pass up this golden opportunity. Good luck and happy spending! Kerry Ford, Centerport, NY ORDER YOUR REPORTS TODAY AND GET STARTED ON YOUR ROAD TO FINANCIAL FREEDOM! NOW IS THE TIME FOR YOUR TURN DECISIVE ACTION YIELDS POWERFUL RESULTS PLEASE NOTE: If you need help with starting a business, registering a business name, learning how income tax is handled, etc., contact your local office of the Small Business Administration (a Federal agency) 1-(800)827-5722 for free help and answers to questions. Also, the Internal Revenue Service offers free help via telephone and free seminars about business tax requirements. Your earnings and results are highly dependant on your activities and advertising. This letter constitutes no guarantees stated nor implied. In the event that it is determined that this letter constitutes a guarantee of any kind, that guarantee is now void. Any testimonials or amounts of earnings listed in this letter may be factual or fictitious. If you have any question of the legality of this letter contact the Office of Associate Director for Marketing Practices Federal Trade Commission Bureau of Consumer Protection in Washington DC. From owner-autofs@linux.kernel.org Thu Feb 18 03:11:09 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA21142 for ; Thu, 18 Feb 1999 03:11:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA14065 for autofs-list; Thu, 18 Feb 1999 01:44:36 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA14027 for ; Thu, 18 Feb 1999 01:44:03 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id KAA13619 for ; Thu, 18 Feb 1999 10:42:27 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id KAA00357 for autofs@linux.kernel.org; Thu, 18 Feb 1999 10:42:26 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Thu, 18 Feb 1999 10:42:26 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: autofs@linux.kernel.org Subject: Re: point of confusion Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thus spoke Michael Newlyn Blake on 18-Feb-99 : > On Wed, 17 Feb 1999, H. Peter Anvin wrote: >> > I think the syntax is something like: >> > * machine1,machine2:/dir/& >> No, it is a (key,value) type lookup; therefore having two entries with >> the same key is just plain WRONG. > > Actually, I think she's refering to the way Solaris does autofs. The > comma seperated list of hostnames is evaluated by autofs itself, not > passed along to mount. The hostnames are passed along to seperate mount > requests, providing a sort of failover. (For example, two nfs servers > providing access to identical filesystems. If one stops responding, > Solaris will try to force a remount from the secondary server.) > > A cool feature, but probably not supported in Linux autofs. (Or is it?) That would be very nice, if that is supported in autofs. Good old amd, which we run on our linux boxes in the pre autofs days (1994 - 1998) would do that for you. FA "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Thu Feb 18 13:04:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA21576 for ; Thu, 18 Feb 1999 13:04:17 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id LAA08086 for autofs-list; Thu, 18 Feb 1999 11:59:08 -0800 Received: from phobos.talarian.com (mailhost.talarian.com [207.5.32.17]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id LAA08080 for ; Thu, 18 Feb 1999 11:59:05 -0800 Received: from iota (iota.talarian.com [207.5.32.205]) by phobos.talarian.com (8.9.0/8.9.0) with SMTP id LAA08388; Thu, 18 Feb 1999 11:59:03 -0800 (PST) Message-ID: <199902181158580185.0A105C1B@mailhost.talarian.com> In-Reply-To: References: X-Mailer: Calypso Version 2.40.41.05 Date: Thu, 18 Feb 1999 11:58:58 -0800 From: "Peter Walker" To: frithjof@fkp.tu-darmstadt.de, autofs@linux.kernel.org Subject: Re: point of confusion Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: This feature is not supported in autofs. However, in the source distribution of autofs there is a patch that can be applied to mount so that mount will parse out these multiple machine names and just try to use the first one. This is not an ideal solution but at least it allows a linux box to participate in a heterogenous network where there are other machines (Sun, HP etc) that do support and use this feature of autofs, without having to change all your autofs maps to support linux. Peter *********** REPLY SEPARATOR *********** On 2/18/99, at 3:18 AM, Frithjof Anders wrote: >Thus spoke Michael Newlyn Blake on 18-Feb-99 : >> On Wed, 17 Feb 1999, H. Peter Anvin wrote: >>> > I think the syntax is something like: >>> > * machine1,machine2:/dir/& >>> No, it is a (key,value) type lookup; therefore having two entries with >>> the same key is just plain WRONG. >> >> Actually, I think she's refering to the way Solaris does autofs. The >> comma seperated list of hostnames is evaluated by autofs itself, not >> passed along to mount. The hostnames are passed along to seperate mount >> requests, providing a sort of failover. (For example, two nfs servers >> providing access to identical filesystems. If one stops responding, >> Solaris will try to force a remount from the secondary server.) >> >> A cool feature, but probably not supported in Linux autofs. (Or is it?) > >That would be very nice, if that is supported in autofs. Good >old amd, which we run on our linux boxes in the pre autofs days (1994 - 1998) >would do that for you. > >FA > > > > > "If you see someone without a smile, give him one of yours" > >Frithjof Anders >Institut fuer Festkoerperphysik >Technische Universitaet Darmstadt >Hochschulstr. 6 >64289 Darmstadt, GERMANY > > >Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de >FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Fri Feb 19 15:48:19 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA22812 for ; Fri, 19 Feb 1999 15:48:18 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA15197 for autofs-list; Fri, 19 Feb 1999 14:57:40 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA15193 for ; Fri, 19 Feb 1999 14:57:37 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id OAA25818; Fri, 19 Feb 1999 14:56:30 -0800 From: Alvin Oga Message-Id: <199902192256.OAA25818@leopard.imedia.com> Subject: Re: Automounter does not start except at bootup To: sjl96v@ecs.soton.ac.uk (Simon Liddington) Date: Fri, 19 Feb 1999 14:56:30 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Simon Liddington" at Feb 19, 99 05:12:17 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi ... what's the exact nature of the problems ? what is your error messages ? is autofs properly installed and configured ? have fun alvin http://www.linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > I am using the autofs-3.1.1-4.i386.rpm package on redhat 5.2. > > If I restart the automounter with these commands it stops but doesn't > restart again. > > /etc/rc.d/init.d/autofs stop > /etc/rc.d/init.d/autofs start > > If I try to start it by hand and type > > /usr/sbin/automount --timeout 60 /net file /etc/auto.net > > then it returns immediately and does nothing. > > It seems to only start if started at boot up. > > Help, > Simon > > ----------------------------------------------------------------------- > | Simon Liddington | | > | E-Mail : sjl96v@ecs.soton.ac.uk | Tel (work) : +44 (0)1703 592422 | > ----------------------------------------------------------------------- > From owner-autofs@linux.kernel.org Fri Feb 19 09:58:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA22583 for ; Fri, 19 Feb 1999 09:58:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA31029 for autofs-list; Fri, 19 Feb 1999 09:13:20 -0800 Received: from rayleigh.ecs.soton.ac.uk (IDENT:root@rayleigh.ecs.soton.ac.uk [152.78.66.94]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA30994 for ; Fri, 19 Feb 1999 09:12:43 -0800 Received: from hartley.ecs.soton.ac.uk (sjl@hartley.ecs.soton.ac.uk [152.78.66.91]) by rayleigh.ecs.soton.ac.uk (8.8.7/8.8.7) with ESMTP id RAA10266 for ; Fri, 19 Feb 1999 17:12:18 GMT Date: Fri, 19 Feb 1999 17:12:17 +0000 (GMT) From: Simon Liddington To: autofs@linux.kernel.org Subject: Automounter does not start except at bootup Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi, I am using the autofs-3.1.1-4.i386.rpm package on redhat 5.2. If I restart the automounter with these commands it stops but doesn't restart again. /etc/rc.d/init.d/autofs stop /etc/rc.d/init.d/autofs start If I try to start it by hand and type /usr/sbin/automount --timeout 60 /net file /etc/auto.net then it returns immediately and does nothing. It seems to only start if started at boot up. Help, Simon ----------------------------------------------------------------------- | Simon Liddington | | | E-Mail : sjl96v@ecs.soton.ac.uk | Tel (work) : +44 (0)1703 592422 | ----------------------------------------------------------------------- From owner-autofs@linux.kernel.org Sun Feb 21 02:43:07 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA24253 for ; Sun, 21 Feb 1999 02:43:07 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA05383 for autofs-list; Sun, 21 Feb 1999 01:41:00 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA05380 for ; Sun, 21 Feb 1999 01:40:58 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id KAA04873; Sun, 21 Feb 1999 10:41:01 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id KAA10191; Sun, 21 Feb 1999 10:41:01 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Sun, 21 Feb 1999 10:41:01 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Simon Liddington Subject: RE: Automounter does not start except at bootup Cc: autofs@linux.kernel.org Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thus spoke Simon Liddington on 19-Feb-99 : > Hi, > > I am using the autofs-3.1.1-4.i386.rpm package on redhat 5.2. > > If I restart the automounter with these commands it stops but doesn't > restart again. > > /etc/rc.d/init.d/autofs stop > /etc/rc.d/init.d/autofs start > > If I try to start it by hand and type > > /usr/sbin/automount --timeout 60 /net file /etc/auto.net > > then it returns immediately and does nothing. > > It seems to only start if started at boot up. type /etc/rc.d/init.d/autofs status it gives you, which point moints a configured and with are active e.g something like: Configured Mount Points: ------------------------ automount /home yp auto.home automount /imports yp auto.imports Active Mount Points: -------------------- automount /home yp auto.home automount /imports yp auto.imports If you stop the automounter with the script and the mount points are still in use, the automounter process will be killed and if you do give a started afterwards it will not be started again on the mount points which are still in use :) You have to kill all processes, which use the automounter mount point, kill eventuall all still running automount processes. Then you can start new. This is a weekness of the autofs implementation. I am not an expert, just a user of autofs, but I guess it has to do with the fact, that part of the autofs runs in the kernel. The amd automounter was nicer with that respect: you could kill and restart it and it worked fine. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Mon Feb 22 03:55:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA25687 for ; Mon, 22 Feb 1999 03:55:31 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA02021 for autofs-list; Mon, 22 Feb 1999 02:46:59 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA02017 for ; Mon, 22 Feb 1999 02:46:57 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id CAA04046; Mon, 22 Feb 1999 02:36:55 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id CAA05208; Mon, 22 Feb 1999 02:36:55 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id CAA27025; Mon, 22 Feb 1999 02:36:55 -0800 Message-Id: <199902221036.CAA27025@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from Simon Liddington at "Feb 22, 99 10:24:15 am" To: sjl96v@ecs.soton.ac.uk (Simon Liddington) Date: Mon, 22 Feb 1999 02:36:54 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > Unfortunately there are other user processes using the mounts so I > can't really just kill them all off to restart automount. It would be > nice if automount could keep some table like mtab of mounts that were > setup by it and then when it is restarted it takes control of them > again. > > I imagine the official line on this is that you don't need to restart > an automount process because you can just change the map file and it > should take effect immediately. The "autofs reload|restart" only kills > or starts mounts which have changed in the auto.master file. > > I guess I'll either have to kill those processes or permanently mount > the directories until the next reboot! > Basically, in autofs v3 there is no way of taking control over an abandoned automount, and I haven't been planning on that kind of support for autofs v4 either; it is really hard to do and I see very little benefit to it. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 03:59:24 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA25691 for ; Mon, 22 Feb 1999 03:59:24 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA01347 for autofs-list; Mon, 22 Feb 1999 02:27:56 -0800 Received: from hartley.ecs.soton.ac.uk (hartley.ecs.soton.ac.uk [152.78.66.91]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA01293 for ; Mon, 22 Feb 1999 02:25:29 -0800 Received: (from sjl@localhost) by hartley.ecs.soton.ac.uk (8.8.7/8.8.7) id KAA09152; Mon, 22 Feb 1999 10:24:16 GMT To: autofs@linux.kernel.org Subject: Re: Automounter does not start except at bootup References: From: Simon Liddington Date: 22 Feb 1999 10:24:15 +0000 In-Reply-To: Frithjof Anders's message of "Sun, 21 Feb 1999 10:41:01 +0100 (CET)" Message-ID: Lines: 92 X-Mailer: Gnus v5.5/Emacs 20.3 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Frithjof Anders writes: > Thus spoke Simon Liddington on 19-Feb-99 : > > Hi, > > > > I am using the autofs-3.1.1-4.i386.rpm package on redhat 5.2. > > > > If I restart the automounter with these commands it stops but doesn't > > restart again. > > > > /etc/rc.d/init.d/autofs stop > > /etc/rc.d/init.d/autofs start > > > > If I try to start it by hand and type > > > > /usr/sbin/automount --timeout 60 /net file /etc/auto.net > > > > then it returns immediately and does nothing. > > > > It seems to only start if started at boot up. > > type > /etc/rc.d/init.d/autofs status > > it gives you, which point moints a configured and with are active e.g > something like: > > Configured Mount Points: > ------------------------ > automount /home yp auto.home > automount /imports yp auto.imports > > Active Mount Points: > -------------------- > automount /home yp auto.home > automount /imports yp auto.imports > > > If you stop the automounter with the script and the mount points are still in > use, the automounter process will be killed and if you do give a started > afterwards it will not be started again on the mount points which are still in > use :) You have to kill all > processes, which use the automounter mount point, kill eventuall all still > running automount processes. Then you can start new. > > This is a weekness of the autofs implementation. I am not an expert, just a > user of autofs, but I guess it has to do with the fact, that part of the autofs > runs in the kernel. The amd automounter was nicer with that respect: you > could kill and restart it and it worked fine. > Thanks, this makes sense. Here's some more info: I found this in the logs (Dunno why I didn't look there in the first place) The stop command doesn't unmount from /net because some of the mounts were still busy and so the restart failed. Here's the logs: Feb 18 11:58:42 rayleigh automount[333]: shutting down, path = /net Feb 18 11:58:42 rayleigh automount[333]: could not unmount /net/hartley Feb 18 11:58:42 rayleigh automount[333]: could not unmount /net/comms6 Feb 18 11:58:42 rayleigh automount[333]: could not unmount /net/viper Feb 18 11:58:42 rayleigh automount[333]: >> umount: /net: device is busy Feb 18 11:58:44 rayleigh automount[4787]: starting automounter version 3.1.1, path = /net, maptype = file, mapname = /etc/auto.net Feb 18 11:58:44 rayleigh automount[4787]: >> mount: automount(pid4787) already mounted or /net busy Feb 18 11:58:44 rayleigh automount[4787]: >> mount: according to mtab, automount(pid333) is already mounted on /net Feb 18 11:58:44 rayleigh automount[4787]: /net: mount failed! Unfortunately there are other user processes using the mounts so I can't really just kill them all off to restart automount. It would be nice if automount could keep some table like mtab of mounts that were setup by it and then when it is restarted it takes control of them again. I imagine the official line on this is that you don't need to restart an automount process because you can just change the map file and it should take effect immediately. The "autofs reload|restart" only kills or starts mounts which have changed in the auto.master file. I guess I'll either have to kill those processes or permanently mount the directories until the next reboot! Simon -- ----------------------------------------------------------------------- | Simon Liddington | | | E-Mail : sjl96v@ecs.soton.ac.uk | Tel (work) : +44 (0)1703 592422 | ----------------------------------------------------------------------- From owner-autofs@linux.kernel.org Mon Feb 22 06:14:25 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id GAA25837 for ; Mon, 22 Feb 1999 06:14:24 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA07922 for autofs-list; Mon, 22 Feb 1999 05:02:05 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA07918 for ; Mon, 22 Feb 1999 05:02:03 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id FAA04902 for ; Mon, 22 Feb 1999 05:02:02 -0800 Received: from korell.transmeta.com (hpa@korell.transmeta.com [10.8.42.6]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id FAA07392 for ; Mon, 22 Feb 1999 05:02:01 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by korell.transmeta.com (8.8.4/8.7.3) id FAA09677 for autofs@linux.kernel.org; Mon, 22 Feb 1999 05:02:01 -0800 Message-Id: <199902221302.FAA09677@korell.transmeta.com> Subject: autofs v4: speak now or forever hold your peace To: autofs@linux.kernel.org (autofs mailing list) Date: Mon, 22 Feb 1999 05:02:01 -0800 (PST) Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Okay, I'm planning to do outline the basic data structures of autofs v4 over the next several days. Some features are already a given: * Multimounts/scaffolding (/net being a special case of this). * Arbitrary mount point topology, without needing the spawn-an-automounter hack. * Mounting-process information passed to daemon. * Most of the filesystem data will live in "ready-to-eat" kernel data structures (inodes, dentries) as opposed to separate backing store. A few frequently requested features that are *not* included, so don't ask: * Direct mounts. These appear to require a significant slowdown to the entire VFS, and I do not believe this is justifiable. * lofs. Not because I don't think it is a good idea if it can be done right, but because it is a feature orthogonal to autofs. * Replicated servers. This is an NFS issue, not autofs. This means it belongs in mount(8) and the NFS kernel code. I completely refuse to hack in a feature into autofs when it is clear it should be provided elsewhere, *especially* when it is a feature like this when it would be more useful if provided elsewhere. After all, why should you be able to mount a replicated server from autofs but not from /etc/fstab?? However, if there is some key feature that I'm missing, please speak up now: please send an email explaining the feature and its justification, and I will try to fit it into the design. Oh, autofs v4 will be targetted for the 2.3 kernels and libc6. 2.3 because I am planning to ask for certain VFS changes, which probably won't be possible in the 2.2 kernel series, and libc6 because I need a thread-safe dynamic linker, which isn't available for libc5. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 06:39:55 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id GAA25872 for ; Mon, 22 Feb 1999 06:39:55 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA08109 for autofs-list; Mon, 22 Feb 1999 05:05:49 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA08094 for ; Mon, 22 Feb 1999 05:05:20 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id NAA06926; Mon, 22 Feb 1999 13:56:38 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id NAA26770; Mon, 22 Feb 1999 13:56:37 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902221036.CAA27025@cesium.transmeta.com> Date: Mon, 22 Feb 1999 13:56:37 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: Automounter does not start except at bootup Cc: autofs@linux.kernel.org, (Simon Liddington) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 22-Feb-99 : >> >> >> I imagine the official line on this is that you don't need to restart >> an automount process because you can just change the map file and it >> should take effect immediately. The "autofs reload|restart" only kills >> or starts mounts which have changed in the auto.master file. >> >> I guess I'll either have to kill those processes or permanently mount >> the directories until the next reboot! >> > > Basically, in autofs v3 there is no way of taking control over an > abandoned automount, and I haven't been planning on that kind of > support for autofs v4 either; it is really hard to do and I see very > little benefit to it. I disagree that this feature has "little benefit". Imagine, you want to move a user from one partition to another. That happens sometimes if you are a sysadmin. Changing NIS maps will not solve the problem. Yep, then you end up rebooting the machines or tell the user, he/she can only log onto machines, which have been not used by him/her. Or is there another simple official procedure, how to deal with these situations and avoid reboot? That is IMOP the major drawback of autofs versus good old AMD, we did use before. AMD could be restarted with no problems. The only reason, we switch to autofs was, that LINUX mount changed over the years and the AMD linux mount code is not maintained any more. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Mon Feb 22 07:12:52 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25884 for ; Mon, 22 Feb 1999 07:12:52 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA11198 for autofs-list; Mon, 22 Feb 1999 06:09:34 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA11176 for ; Mon, 22 Feb 1999 06:09:05 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id PAA07059; Mon, 22 Feb 1999 15:09:04 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id PAA26973; Mon, 22 Feb 1999 15:09:03 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902221344.FAA18047@cesium.transmeta.com> Date: Mon, 22 Feb 1999 15:09:03 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: Automounter does not start except at bootup Cc: (autofs mailing list) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 22-Feb-99 : >> > >> > Basically, in autofs v3 there is no way of taking control over an >> > abandoned automount, and I haven't been planning on that kind of >> > support for autofs v4 either; it is really hard to do and I see very >> > little benefit to it. >> >> I disagree that this feature has "little benefit". Imagine, you want >> to move a user from one partition to another. That happens sometimes >> if you are a sysadmin. Changing NIS maps will not solve the problem. >> Yep, then you end up rebooting the machines or tell the user, he/she >> can only log onto machines, which have been not used by him/her. >> >> Or is there another simple official procedure, how to deal with >> these situations and avoid reboot? > > Yes, you explicitly umount *that directory*. Maybe I do not understand what you mean with "explicitly umount *that directory*" umount /home/user1 on a autofs mountpoint /home does not work. Therefore the directory entry /home/user1 is stuck. Back to amd. Amd provides just a dynamical link layer over a mount point /home. The real process sees /net/host/userspace/user1.( Remember, we had the discussion about getcwd()). Therefore you can destroy the layer, by killing an restarting the automounter, without affecting the processes. > Note you can't have > processes holding the filesystem busy, no matter what you do. It just > *cannot* work. It looks to me, that the reason is part of autofs, the autofs filesystem, runs in the Linux kernel. Bad luck. >> years and the AMD linux mount code is not maintained any more. > > It is, under the name am-utils, as far as I know. Thanks for the hint. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Mon Feb 22 07:13:53 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25888 for ; Mon, 22 Feb 1999 07:13:53 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA11550 for autofs-list; Mon, 22 Feb 1999 06:14:42 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA11547 for ; Mon, 22 Feb 1999 06:14:41 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id GAA05475; Mon, 22 Feb 1999 06:14:30 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id GAA08636; Mon, 22 Feb 1999 06:14:30 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id GAA18353; Mon, 22 Feb 1999 06:14:30 -0800 Message-Id: <199902221414.GAA18353@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from Frithjof Anders at "Feb 22, 99 03:09:03 pm" To: frithjof@fkp.tu-darmstadt.de Date: Mon, 22 Feb 1999 06:14:30 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Maybe I do not understand what you mean with > "explicitly umount *that directory*" > > umount /home/user1 > on a autofs mountpoint /home does not work. Therefore the directory entry > /home/user1 is stuck. : cesium 1 # cd /home/quinlan : cesium 2 # cd / : cesium 3 # ls -l /home total 68 lrwxrwxrwx 1 root root 19 Feb 22 01:18 cesium -> /export/home/cesium/ drwxr-xr-x 178 hpa hpa 45056 Feb 22 06:12 hpa/ drwxrwsr-x 48 quinlan quinlan 24576 Feb 22 02:52 quinlan/ : cesium 4 # umount /home/quinlan : cesium 5 # ls -l /home total 44 lrwxrwxrwx 1 root root 19 Feb 22 01:18 cesium -> /export/home/cesium/ drwxr-xr-x 178 hpa hpa 45056 Feb 22 06:12 hpa/ : cesium 6 # Note that if there are processes holding /home/user1 open, then you have to kill those processes -- otherwise you can't umount the directory, no matter how it's mounted. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 07:15:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25914 for ; Mon, 22 Feb 1999 07:15:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA09898 for autofs-list; Mon, 22 Feb 1999 05:44:55 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA09895 for ; Mon, 22 Feb 1999 05:44:53 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id FAA05193; Mon, 22 Feb 1999 05:44:28 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id FAA07782; Mon, 22 Feb 1999 05:44:28 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id FAA18047; Mon, 22 Feb 1999 05:44:28 -0800 Message-Id: <199902221344.FAA18047@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from Frithjof Anders at "Feb 22, 99 01:56:37 pm" To: frithjof@fkp.tu-darmstadt.de Date: Mon, 22 Feb 1999 05:44:27 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > >> > >> I imagine the official line on this is that you don't need to restart > >> an automount process because you can just change the map file and it > >> should take effect immediately. The "autofs reload|restart" only kills > >> or starts mounts which have changed in the auto.master file. > >> > >> I guess I'll either have to kill those processes or permanently mount > >> the directories until the next reboot! > >> > > > > Basically, in autofs v3 there is no way of taking control over an > > abandoned automount, and I haven't been planning on that kind of > > support for autofs v4 either; it is really hard to do and I see very > > little benefit to it. > > I disagree that this feature has "little benefit". Imagine, you want > to move a user from one partition to another. That happens sometimes > if you are a sysadmin. Changing NIS maps will not solve the problem. > Yep, then you end up rebooting the machines or tell the user, he/she > can only log onto machines, which have been not used by him/her. > > Or is there another simple official procedure, how to deal with > these situations and avoid reboot? Yes, you explicitly umount *that directory*. Note you can't have processes holding the filesystem busy, no matter what you do. It just *cannot* work. > That is IMOP the major drawback of autofs versus good old AMD, we > did use before. AMD could be restarted with no problems. The only > reason, we switch to autofs was, that LINUX mount changed over the > years and the AMD linux mount code is not maintained any more. It is, under the name am-utils, as far as I know. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 07:15:31 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25918 for ; Mon, 22 Feb 1999 07:15:31 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA11682 for autofs-list; Mon, 22 Feb 1999 06:16:54 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA11679 for ; Mon, 22 Feb 1999 06:16:53 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id GAA05508; Mon, 22 Feb 1999 06:16:51 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id GAA08688; Mon, 22 Feb 1999 06:16:50 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id GAA18370; Mon, 22 Feb 1999 06:16:50 -0800 Message-Id: <199902221416.GAA18370@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from "Brad M. Garcia" at "Feb 22, 99 09:08:46 am" To: bgarcia@fore.com (Brad M. Garcia) Date: Mon, 22 Feb 1999 06:16:50 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On Mon, 22 Feb 1999, Frithjof Anders wrote: > > > > Thus spoke H. Peter Anvin on 22-Feb-99 : > > > > > > Basically, in autofs v3 there is no way of taking control over an > > > abandoned automount, and I haven't been planning on that kind of > > > support for autofs v4 either; it is really hard to do and I see very > > > little benefit to it. > > > > I disagree that this feature has "little benefit". Imagine, you want to > > move a user from one partition to another. That happens sometimes if you > > are a sysadmin. Changing NIS maps will not solve the problem. > > Yep, then you end up rebooting the machines or tell the user, he/she can only > > log onto machines, which have been not used by him/her. > > I have to agree with Frithjof. I have had other reasons for restarting > autofs as well (upgrading autofs for one), and it is a pain to make sure > that nobody on the machine is currently using an autofs-controlled > directory. > > I understand if you don't wish to code it yourself for v4, but I would > like to convince you that such a feature would be very beneficial. Let me rephrase: it is not worth the *large* amount of overhead that it would cause. It is somewhat like asking that you should be able to cut your head off and put it on your next-door neighbour; in between you have real problems... -hpa From owner-autofs@linux.kernel.org Mon Feb 22 07:30:34 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25929 for ; Mon, 22 Feb 1999 07:30:34 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA11457 for autofs-list; Mon, 22 Feb 1999 06:13:03 -0800 Received: from mailgate.fore.com (mailgate.fore.com [169.144.68.6]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA11443 for ; Mon, 22 Feb 1999 06:13:01 -0800 Received: from mailman.fore.com (mailman.fore.com [169.144.2.12]) by mailgate.fore.com (8.9.1/8.9.1) with ESMTP id JAA08875; Mon, 22 Feb 1999 09:06:08 -0500 (EST) Received: from sol.eng.fore.com (sol [169.144.155.73]) by mailman.fore.com (8.8.8/8.8.8) with ESMTP id JAA22887; Mon, 22 Feb 1999 09:06:37 -0500 (EST) Received: from spud.eng.fore.com (bgarcia@spud [169.144.87.22]) by sol.eng.fore.com (8.8.8/8.8.8) with ESMTP id JAA03855; Mon, 22 Feb 1999 09:06:08 -0500 (EST) Received: from localhost (bgarcia@localhost) by spud.eng.fore.com (8.8.7/8.8.7) with ESMTP id JAA17041; Mon, 22 Feb 1999 09:08:46 -0500 X-Authentication-Warning: spud.eng.fore.com: bgarcia owned process doing -bs Date: Mon, 22 Feb 1999 09:08:46 -0500 (EST) From: "Brad M. Garcia" X-Sender: bgarcia@spud.eng.fore.com To: Frithjof Anders cc: "H. Peter Anvin" , autofs@linux.kernel.org, sjl96v@ecs.soton.ac.uk Subject: Re: Automounter does not start except at bootup In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Mon, 22 Feb 1999, Frithjof Anders wrote: > > Thus spoke H. Peter Anvin on 22-Feb-99 : > > > > Basically, in autofs v3 there is no way of taking control over an > > abandoned automount, and I haven't been planning on that kind of > > support for autofs v4 either; it is really hard to do and I see very > > little benefit to it. > > I disagree that this feature has "little benefit". Imagine, you want to > move a user from one partition to another. That happens sometimes if you > are a sysadmin. Changing NIS maps will not solve the problem. > Yep, then you end up rebooting the machines or tell the user, he/she can only > log onto machines, which have been not used by him/her. I have to agree with Frithjof. I have had other reasons for restarting autofs as well (upgrading autofs for one), and it is a pain to make sure that nobody on the machine is currently using an autofs-controlled directory. I understand if you don't wish to code it yourself for v4, but I would like to convince you that such a feature would be very beneficial. Brad Garcia ___/ __ / __ / ___/ "Being the Linux of digital media __/ / / / _/ __/ would be a very good life." _/ ____/ _/ _| ____/ - Jean-Louis Gassee, CEO of Be, Inc. From owner-autofs@linux.kernel.org Mon Feb 22 07:35:14 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25933 for ; Mon, 22 Feb 1999 07:35:13 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA12203 for autofs-list; Mon, 22 Feb 1999 06:26:33 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA12200 for ; Mon, 22 Feb 1999 06:26:32 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id GAA05578 for ; Mon, 22 Feb 1999 06:26:31 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id GAA08816 for ; Mon, 22 Feb 1999 06:26:31 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id GAA18535 for autofs@linux.kernel.org; Mon, 22 Feb 1999 06:26:31 -0800 Message-Id: <199902221426.GAA18535@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace (fwd) To: autofs@linux.kernel.org (autofs mailing list) Date: Mon, 22 Feb 1999 06:26:31 -0800 (PST) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Why not ? Any way to get that feature working would be _very_ > helpfull, and as far as I can see, kernel support for that feature > seems to be years away. > > > However, if there is some key feature that I'm missing, please speak > > up now: please send an email explaining the feature and its > > justification, and I will try to fit it into the design. > > Will individual timeouts for different keys be implemented ? So you can > have for example 1 minute for the floppy/cd and say 5 minutes for a server > mount. Also a command for trying to unmount a special key may be helpfull > (for changing of floppys/cd's). > That's already available; it's called umount. As far as the different timeouts is concerned; I will consider it, but if it ends up being expensive data-structure wise I will drop it like a hot potato. With a single timeout, one can just use a simple priority queue, which is very cheap to maintain (just move 4 pointers on each access.) -hpa From owner-autofs@linux.kernel.org Mon Feb 22 07:45:48 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25950 for ; Mon, 22 Feb 1999 07:45:48 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA12422 for autofs-list; Mon, 22 Feb 1999 06:31:55 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA12417 for ; Mon, 22 Feb 1999 06:31:46 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id PAA07111; Mon, 22 Feb 1999 15:31:47 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id PAA27067; Mon, 22 Feb 1999 15:31:47 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902221412.GAA18292@cesium.transmeta.com> Date: Mon, 22 Feb 1999 15:31:47 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" , autofs@linux.kernel.org Subject: Re: Automounter does not start except at bootup Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 22-Feb-99 : >> >> Maybe I do not understand what you mean with >> "explicitly umount *that directory*" >> >> umount /home/user1 >> on a autofs mountpoint /home does not work. Therefore the directory entry >> /home/user1 is stuck. > > Why doesn't it work? It is supposed to. Sorry it does now. My mistake. A lost process hang on it. >> >> It looks to me, that the reason is part of autofs, the autofs >> filesystem, runs in the Linux kernel. Bad luck. > > Well, I'll correct myself: you can make it work if you're willing to > take a two order of magnitude slowdown. I disagree. I would implement the nfs mounts of autofs like autofs handles local links. That means, instead of mounting the remote directory directly onto /home it would mount it on /aufofs/host/user.... and then make a link from /home/user to /aufofs/host/user. That means, the access will be as fast as Linux can handle symbolic links. If that is slow, then autofs is slow. But how fast is NFS anyway. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Mon Feb 22 07:49:07 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25956 for ; Mon, 22 Feb 1999 07:49:07 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA12592 for autofs-list; Mon, 22 Feb 1999 06:34:49 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA12589 for ; Mon, 22 Feb 1999 06:34:47 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id GAA05652; Mon, 22 Feb 1999 06:34:33 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id GAA08920; Mon, 22 Feb 1999 06:34:33 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id GAA18654; Mon, 22 Feb 1999 06:34:32 -0800 Message-Id: <199902221434.GAA18654@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from Frithjof Anders at "Feb 22, 99 03:31:47 pm" To: frithjof@fkp.tu-darmstadt.de Date: Mon, 22 Feb 1999 06:34:32 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > >> It looks to me, that the reason is part of autofs, the autofs > >> filesystem, runs in the Linux kernel. Bad luck. > > > > Well, I'll correct myself: you can make it work if you're willing to > > take a two order of magnitude slowdown. > > I disagree. I would implement the nfs mounts of autofs like autofs > handles local links. That means, instead of mounting the remote > directory directly onto /home it would mount it on > /aufofs/host/user.... and then make a link from /home/user to > /aufofs/host/user. > > That means, the access will be as fast as Linux can handle symbolic links. If > that is slow, then autofs is slow. But how fast is NFS anyway. But that doesn't solve your problem; you still can't umount the old directory if it is busy. Besides, you get the path problems, which I *definitely* don't want to re-introduce (in fact, people frequently request lofs, which would completely eliminate them.) -hpa From owner-autofs@linux.kernel.org Mon Feb 22 07:55:25 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA25962 for ; Mon, 22 Feb 1999 07:55:25 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA13828 for autofs-list; Mon, 22 Feb 1999 06:57:51 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA13825 for ; Mon, 22 Feb 1999 06:57:45 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id PAA07182; Mon, 22 Feb 1999 15:57:41 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id PAA27286; Mon, 22 Feb 1999 15:57:41 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902221452.GAA18839@cesium.transmeta.com> Date: Mon, 22 Feb 1999 15:57:41 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: Automounter does not start except at bootup Cc: (autofs mailing list) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Peter, Thus spoke H. Peter Anvin on 22-Feb-99 : >> > >> > But that doesn't solve your problem; you still can't umount the old >> > directory if it is busy. >> >> I agree, I cannot umount the old directory. But I did not want to right >> away. >> >> I want to copy user data onto an new location and make it accessable >> immediately to the user using his/her standard home path. At some stage the >> processes on the old directory die anyway, which allows the automounter to >> umount the old directories. > > I see. This actually could be done by allowing you to move the old > directory out of the way; I have to admit I'm somewhat sceptical to > the feature for a variety of reasons, but I'll keep it in the back of > my head. Maybe that is the solution to that problem, to allow a move command onto autofs like mv /home/user /home/user_old > ... > Not taken that way. I just want to emphasize that I specifically am > not willing to turn autofs into a kitchen-sink thing; I want to keep > the design reasonably simple. I perfecly understand that. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Mon Feb 22 08:15:35 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA25992 for ; Mon, 22 Feb 1999 08:15:35 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA14393 for autofs-list; Mon, 22 Feb 1999 07:05:34 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA14387 for ; Mon, 22 Feb 1999 07:05:33 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id HAA05956; Mon, 22 Feb 1999 07:04:45 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id HAA09615; Mon, 22 Feb 1999 07:04:45 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id HAA18952; Mon, 22 Feb 1999 07:04:44 -0800 Message-Id: <199902221504.HAA18952@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Janos Farkas at "Feb 22, 99 03:59:53 pm" To: Janos.Farkas-nouce/priv-#xdfcEDYh2TJD9Z/HjVHf6jEai8e@lk9qw.mail.eon.ml.org (Janos Farkas) Date: Mon, 22 Feb 1999 07:04:44 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On 1999-02-22 at 05:02:01, H. Peter Anvin wrote: > > However, if there is some key feature that I'm missing, please speak > > up now: please send an email explaining the feature and its > > justification, and I will try to fit it into the design. > > Not really a 'design' issue, but a non-daemonize option please? It's > about three lines, but I always forget to submit... Justification, please? > Oh, and maybe per-mount timeouts to finetune resource usage to known > usage patterns. Will be considered, but it is potentially too expensive. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 08:17:47 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA25998 for ; Mon, 22 Feb 1999 08:17:47 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA13575 for autofs-list; Mon, 22 Feb 1999 06:53:01 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA13570 for ; Mon, 22 Feb 1999 06:52:58 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id GAA05835; Mon, 22 Feb 1999 06:52:13 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id GAA09254; Mon, 22 Feb 1999 06:52:12 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id GAA18839; Mon, 22 Feb 1999 06:52:12 -0800 Message-Id: <199902221452.GAA18839@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: from Frithjof Anders at "Feb 22, 99 03:47:40 pm" To: frithjof@fkp.tu-darmstadt.de Date: Mon, 22 Feb 1999 06:52:12 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > > But that doesn't solve your problem; you still can't umount the old > > directory if it is busy. > > I agree, I cannot umount the old directory. But I did not want to right > away. > > I want to copy user data onto an new location and make it accessable > immediately to the user using his/her standard home path. At some stage the > processes on the old directory die anyway, which allows the automounter to > umount the old directories. I see. This actually could be done by allowing you to move the old directory out of the way; I have to admit I'm somewhat sceptical to the feature for a variety of reasons, but I'll keep it in the back of my head. > I just reporting here, in what way I abused the amd. I realy enjoyed that > feature. > > BTW: I would like to emphasize that writing the autofs is a very appreciated > task and I would like to thank you for that. I did not mean to critize you at > all. Not taken that way. I just want to emphasize that I specifically am not willing to turn autofs into a kitchen-sink thing; I want to keep the design reasonably simple. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 08:42:03 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA26005 for ; Mon, 22 Feb 1999 08:42:03 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA16493 for autofs-list; Mon, 22 Feb 1999 07:40:20 -0800 Received: from mail13.digital.com (mail13.digital.com [192.208.46.30]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA16473 for ; Mon, 22 Feb 1999 07:40:16 -0800 Received: from wasted.zk3.dec.com (bywasted.zk3.dec.com [16.140.96.51]) by mail13.digital.com (8.9.1a/8.9.1/WV2.0d) with ESMTP id KAA23038; Mon, 22 Feb 1999 10:40:14 -0500 (EST) Received: from localhost by wasted.zk3.dec.com (8.8.8/1.1.22.2/08Sep98-0251PM) id KAA0000007842; Mon, 22 Feb 1999 10:40:14 -0500 (EST) From: Eric Werme USG Message-Id: <199902221540.KAA0000007842@wasted.zk3.dec.com> X-Authentication-Warning: wasted.zk3.dec.com: localhost [127.0.0.1] didn't use HELO protocol To: "H. Peter Anvin" cc: autofs@linux.kernel.org (autofs mailing list) Subject: Re: Automounter does not start except at bootup In-reply-to: Your message of "Mon, 22 Feb 1999 05:44:27 PST." <199902221344.FAA18047@cesium.transmeta.com> Date: Mon, 22 Feb 1999 10:40:13 -0500 X-Mts: smtp Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Or is there another simple official procedure, how to deal with > these situations and avoid reboot? Yes, you explicitly umount *that directory*. Note you can't have processes holding the filesystem busy, no matter what you do. It just *cannot* work. Can you rename the "intercept points" so that the new incarnation has a new hierarchy to mount things on? Caveats: Dunix doesn't support autofs so there's lots I don't know. Judging from a discussion in nfs-devel, there's a lot I don't know about Linux's VFS, etc. -Ric Werme -- Eric (Ric) Werme | werme@zk3.dec.com Compaq Computer Corp. | http://www.cyberportal.net/werme From owner-autofs@linux.kernel.org Mon Feb 22 09:02:42 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA26015 for ; Mon, 22 Feb 1999 09:02:42 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id IAA17781 for autofs-list; Mon, 22 Feb 1999 08:01:30 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id IAA17776 for ; Mon, 22 Feb 1999 08:01:29 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id IAA06609; Mon, 22 Feb 1999 08:01:27 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id IAA11177; Mon, 22 Feb 1999 08:01:26 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id IAA19421; Mon, 22 Feb 1999 08:01:26 -0800 Message-Id: <199902221601.IAA19421@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: <199902221540.KAA0000007842@wasted.zk3.dec.com> from Eric Werme USG at "Feb 22, 99 10:40:13 am" To: werme@zk3.dec.com (Eric Werme USG) Date: Mon, 22 Feb 1999 08:01:26 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > Or is there another simple official procedure, how to deal with > > these situations and avoid reboot? > > Yes, you explicitly umount *that directory*. Note you can't have > processes holding the filesystem busy, no matter what you do. It just > *cannot* work. > > Can you rename the "intercept points" so that the new incarnation has > a new hierarchy to mount things on? > > Caveats: Dunix doesn't support autofs so there's lots I don't know. > Judging from a discussion in nfs-devel, there's a lot I don't know about > Linux's VFS, etc. > That was one of the ideas. I'm not too keen on the idea as I believe it will greatly complicate the user-space side of autofs, but I will see if it can be done more resonably. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 09:40:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA26048 for ; Mon, 22 Feb 1999 09:40:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id IAA20016 for autofs-list; Mon, 22 Feb 1999 08:45:43 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id IAA20009 for ; Mon, 22 Feb 1999 08:45:41 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id IAA07293; Mon, 22 Feb 1999 08:45:31 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id IAA13197; Mon, 22 Feb 1999 08:45:31 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id IAA19930; Mon, 22 Feb 1999 08:45:30 -0800 Message-Id: <199902221645.IAA19930@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from James O'Kane at "Feb 22, 99 11:43:05 am" To: jo2y@jprc.com (James O'Kane) Date: Mon, 22 Feb 1999 08:45:30 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On Mon, 22 Feb 1999, H. Peter Anvin wrote: > > * Multimounts/scaffolding (/net being a special case of this). > > If I understand what you mean by /net, could it be done without having to > go to a 2.3 kernel? We would like to use this to replace amd, but we need > reliability of the kernel above all else. Probably not. On a technical level it is a very unpleasant feature. > The other feature we use in amd, and I think is in autofs, but I get > conflicting stories, is symlinking. If the mount point happens to be on > the local machine, create a symlink to the directory instead on paying the > penalty of nfs mounting it. autofs does that automatically. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 10:33:34 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA26085 for ; Mon, 22 Feb 1999 10:33:33 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA22553 for autofs-list; Mon, 22 Feb 1999 09:38:33 -0800 Received: from rayleigh.ecs.soton.ac.uk (IDENT:root@rayleigh.ecs.soton.ac.uk [152.78.66.94]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA22542 for ; Mon, 22 Feb 1999 09:38:17 -0800 Received: from hartley.ecs.soton.ac.uk (sjl@hartley.ecs.soton.ac.uk [152.78.66.91]) by rayleigh.ecs.soton.ac.uk (8.8.7/8.8.7) with ESMTP id KAA15215; Mon, 22 Feb 1999 10:58:13 GMT Date: Mon, 22 Feb 1999 10:58:12 +0000 (GMT) From: Simon Liddington To: "H. Peter Anvin" cc: autofs@linux.kernel.org Subject: Re: Automounter does not start except at bootup In-Reply-To: <199902221036.CAA27025@cesium.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Mon, 22 Feb 1999, H. Peter Anvin wrote: > > > > Unfortunately there are other user processes using the mounts so I > > can't really just kill them all off to restart automount. It would be > > nice if automount could keep some table like mtab of mounts that were > > setup by it and then when it is restarted it takes control of them > > again. > > > > I imagine the official line on this is that you don't need to restart > > an automount process because you can just change the map file and it > > should take effect immediately. The "autofs reload|restart" only kills > > or starts mounts which have changed in the auto.master file. > > > > I guess I'll either have to kill those processes or permanently mount > > the directories until the next reboot! > > > > Basically, in autofs v3 there is no way of taking control over an > abandoned automount, and I haven't been planning on that kind of > support for autofs v4 either; it is really hard to do and I see very > little benefit to it. Fair enough. Now I know the implications of stopping automount I just won't do it again. Easy really isn't it ;-) How about a note in the man pages explaining this? BTW, thanks for automount. I much prefer it to amd. Simon ----------------------------------------------------------------------- | Simon Liddington | | | E-Mail : sjl96v@ecs.soton.ac.uk | Tel (work) : +44 (0)1703 592422 | ----------------------------------------------------------------------- From owner-autofs@linux.kernel.org Mon Feb 22 11:08:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id LAA26097 for ; Mon, 22 Feb 1999 11:08:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id KAA24669 for autofs-list; Mon, 22 Feb 1999 10:17:21 -0800 Received: from chipmunk.cita.utoronto.ca (chipmunk.cita.utoronto.ca [128.100.76.49]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id KAA24665 for ; Mon, 22 Feb 1999 10:17:19 -0800 Received: from localhost (pogosyan@localhost) by chipmunk.cita.utoronto.ca (8.6.13/8.6.12) with SMTP id NAA00270; Mon, 22 Feb 1999 13:17:17 -0500 Date: Mon, 22 Feb 1999 13:17:16 -0500 (EST) From: Dmitry Pogosyan To: "H. Peter Anvin" cc: autofs mailing list Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902221302.FAA09677@korell.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: I'm quite new to autofs, but speaking about features I'd like to see something like a browsing mode AMD has, so one can list mounts available for a given mounting points, without actually mounting them. On a side note: what I heard being called as 'mount attack' is at the moment the main problem with automounter on my home machine. Namely, looking at the content of directory, which contains symlinks to my mount keys (like a: -> /mnt/a:, where /mnt is the mount point), with any file manager (I guess except simple 'ls') mounts all directories (including unavailable, like floppy drive without floppy, brrr....). My understanding is that whatever uses stat() call (as ls -l does ??) bound to get the filesystem mounted, but still, maybe there is some simple solution, which would allow safe use of file managers with automount running ? ------------------------------------------------------------ On Mon, 22 Feb 1999, H. Peter Anvin wrote: > Okay, I'm planning to do outline the basic data structures of autofs > v4 over the next several days. Some features are already a given: > > * Multimounts/scaffolding (/net being a special case of this). > * Arbitrary mount point topology, without needing the > spawn-an-automounter hack. > * Mounting-process information passed to daemon. > * Most of the filesystem data will live in "ready-to-eat" kernel data > structures (inodes, dentries) as opposed to separate backing store. > ..................... > > However, if there is some key feature that I'm missing, please speak > up now: please send an email explaining the feature and its > justification, and I will try to fit it into the design. > .................... > -hpa > > From owner-autofs@linux.kernel.org Mon Feb 22 12:41:55 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA26189 for ; Mon, 22 Feb 1999 12:41:55 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id LAA29136 for autofs-list; Mon, 22 Feb 1999 11:55:37 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id LAA29131 for ; Mon, 22 Feb 1999 11:55:33 -0800 Received: from robin.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id LAA16912 for ; Mon, 22 Feb 1999 11:55:26 -0800 X-BlackMail: robin.frogtown.com, robin.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 11:55:26(PST) on February 22, 1999 Date: Mon, 22 Feb 1999 11:52:54 -0800 (PST) From: Michael Newlyn Blake X-Sender: mblake@scooter.corp.frogtown.com To: The autofs list Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902221302.FAA09677@korell.transmeta.com> Message-ID: X-Url: http://www.frogtown.com/~mblake MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > * Direct mounts. These appear to require a significant slowdown to the > entire VFS, and I do not believe this is justifiable. I disagree here... this is an compatability issue. Whether it makes sense or not, most other implimentations of autofs support this and people take it for granted. People will continue to see the Linux implimentation of autofs as buggy or a cheap hack until features like this are taken seriously. Now that doesn't mean that you should necessarily impliment it right away, or drop everything to get it working, but don't dismiss it. It's important. Take your time and get it right, but write it off. > * Replicated servers. This is an NFS issue, not autofs. This means > it belongs in mount(8) and the NFS kernel code. I completely refuse Again, a compatability issue. autofs implimentations that support this do so (I believe) in the autofs code, not the NFS code. There's a lot to be said for doing things right the first time, but there's also a lot to be said for keeping the underlying mechanisms structured as similarly as possable. It makes troubleshooting much easier, especially in a multi-platform envorinment. > should you be able to mount a replicated server from autofs but not > from /etc/fstab?? Because when I put something in fstab, I expect to declare it explicitly and not to change unless I tell it to. When I want something to happen automatically, I use an automounter. > Oh, autofs v4 will be targetted for the 2.3 kernels and libc6. 2.3 > because I am planning to ask for certain VFS changes, which probably Very cool. ----- H. L. Mencken's Law: Those who can -- do. Those who can't -- teach. Martin's Extension: Those who cannot teach -- administrate. From owner-autofs@linux.kernel.org Mon Feb 22 15:18:13 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA26297 for ; Mon, 22 Feb 1999 15:18:13 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA04101 for autofs-list; Mon, 22 Feb 1999 14:44:20 -0800 Received: from tequila.cs.yale.edu (TEQUILA.CS.YALE.EDU [128.36.229.152]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA04098 for ; Mon, 22 Feb 1999 14:44:17 -0800 Received: from tequila.cs.yale.edu (localhost [127.0.0.1]) by tequila.cs.yale.edu (8.8.7/8.8.7) with SMTP id RAA31012 for ; Mon, 22 Feb 1999 17:44:11 -0500 To: autofs@linux.kernel.org From: Stefan Monnier Newsgroups: lists.linux.autofs Subject: Re: Automounter does not start except at bootup References: <199902221036.CAA27025@cesium.transmeta.com> Date: 22 Feb 1999 17:44:05 -0500 Message-ID: <5laey6awju.fsf@tequila.cs.yale.edu> Lines: 19 X-Newsreader: Gnus v5.5/Emacs 20.3 Path: tequila.cs.yale.edu NNTP-Posting-Host: tequila.cs.yale.edu X-Trace: 22 Feb 1999 17:44:06 -0500, tequila.cs.yale.edu Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: >>>>> "H" == H Peter Anvin writes: > Basically, in autofs v3 there is no way of taking control over an > abandoned automount, and I haven't been planning on that kind of > support for autofs v4 either; it is really hard to do and I see very > little benefit to it. Damn! I admit I haven't had any trouble with autofs crashing or being killed randomly, but I always hate it when I have to reboot a machine (or close enough: kill all the processes using some of the submounts) just because I somehow killed the automounter. And even tho as I said I haven't seen autofs die, you can't dismiss the possibility. Also, not being able to kill autofs without nasty side effects means that you cannot upgrade autofs without those nasty side effects. Being able to kill/restart daemons seems very important to me. AutoFS is no exception here. I have no idea how hard it might be to provide such a feature, and I'm not going to code it up any time soon, so I'll just stop whining. Stefan From owner-autofs@linux.kernel.org Mon Feb 22 15:36:22 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA26304 for ; Mon, 22 Feb 1999 15:36:22 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA05422 for autofs-list; Mon, 22 Feb 1999 15:05:59 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA05418 for ; Mon, 22 Feb 1999 15:05:58 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA13820; Mon, 22 Feb 1999 15:00:54 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA17620; Mon, 22 Feb 1999 15:00:53 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA24497; Mon, 22 Feb 1999 15:00:53 -0800 Message-Id: <199902222300.PAA24497@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Michael Newlyn Blake at "Feb 22, 99 11:52:54 am" To: mblake@Frogtown.com (Michael Newlyn Blake) Date: Mon, 22 Feb 1999 15:00:53 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > * Direct mounts. These appear to require a significant slowdown to the > > entire VFS, and I do not believe this is justifiable. > > > > I disagree here... this is an compatability issue. Whether it makes > sense or not, most other implimentations of autofs support this and people > take it for granted. People will continue to see the Linux implimentation > of autofs as buggy or a cheap hack until features like this are taken > seriously. Now that doesn't mean that you should necessarily impliment it > right away, or drop everything to get it working, but don't dismiss it. > It's important. Take your time and get it right, but write it off. I have looked at what it would take. Linus said "no way". It won't happen. > > * Replicated servers. This is an NFS issue, not autofs. This means > > it belongs in mount(8) and the NFS kernel code. I completely refuse > > Again, a compatability issue. autofs implimentations that support this > do so (I believe) in the autofs code, not the NFS code. There's a lot to > be said for doing things right the first time, but there's also a lot to > be said for keeping the underlying mechanisms structured as similarly as > possable. It makes troubleshooting much easier, especially in a > multi-platform envorinment. It's just plain wrong to put it in autofs, for a whole bunch of reasons. One of them is that it can't be done right. It won't happen. > > should you be able to mount a replicated server from autofs but not > > from /etc/fstab?? > > Because when I put something in fstab, I expect to declare it explicitly > and not to change unless I tell it to. When I want something to happen > automatically, I use an automounter. > > > > > Oh, autofs v4 will be targetted for the 2.3 kernels and libc6. 2.3 > > because I am planning to ask for certain VFS changes, which probably > > Very cool. > > > ----- > H. L. Mencken's Law: > Those who can -- do. > Those who can't -- teach. > > Martin's Extension: > Those who cannot teach -- administrate. > From owner-autofs@linux.kernel.org Mon Feb 22 15:42:59 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA26308 for ; Mon, 22 Feb 1999 15:42:59 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA05304 for autofs-list; Mon, 22 Feb 1999 15:03:46 -0800 Received: from tequila.cs.yale.edu (TEQUILA.CS.YALE.EDU [128.36.229.152]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA05301 for ; Mon, 22 Feb 1999 15:03:44 -0800 Received: from tequila.cs.yale.edu (localhost [127.0.0.1]) by tequila.cs.yale.edu (8.8.7/8.8.7) with SMTP id SAA31564 for ; Mon, 22 Feb 1999 18:03:43 -0500 To: autofs@linux.kernel.org From: Stefan Monnier Newsgroups: lists.linux.autofs Subject: Re: autofs v4: speak now or forever hold your peace References: <199902221302.FAA09677@korell.transmeta.com> Date: 22 Feb 1999 18:03:38 -0500 Message-ID: <5l90dqavn9.fsf@tequila.cs.yale.edu> Lines: 24 X-Newsreader: Gnus v5.5/Emacs 20.3 Path: tequila.cs.yale.edu NNTP-Posting-Host: tequila.cs.yale.edu X-Trace: 22 Feb 1999 18:03:38 -0500, tequila.cs.yale.edu Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: >> * Direct mounts. These appear to require a significant slowdown to the >> entire VFS, and I do not believe this is justifiable. > I disagree here... this is an compatability issue. Whether it makes I've been wondering about the feasability of understanding direct mounts in the maps, while implementing them via symlinks to an add-hoc autofs mount point. It seems very doable right now with autofs v3 plus a few hacks (implemented via lookup_program.so). Would that be sufficient ? To me direct mounts seem very unimportant (as a sysadm I can very easily change them to normal autofs mounts with no serious detrimental side effects) so they should only be provided if they don't slow down the "normal" case (the case where no direct mounts are used). >> * Replicated servers. This is an NFS issue, not autofs. This means >> it belongs in mount(8) and the NFS kernel code. I completely refuse > Again, a compatability issue. autofs implimentations that support this How close do you need/want the compatibility to be ? Which amounts: what exactly do you use replicated server entries for ? It seems it shouldn't be too hard to spice up mount to do the work. Stefan From owner-autofs@linux.kernel.org Mon Feb 22 15:46:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA26312 for ; Mon, 22 Feb 1999 15:46:27 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA05557 for autofs-list; Mon, 22 Feb 1999 15:09:31 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA05554 for ; Mon, 22 Feb 1999 15:09:30 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA13960; Mon, 22 Feb 1999 15:09:28 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA18302; Mon, 22 Feb 1999 15:09:28 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA24586; Mon, 22 Feb 1999 15:09:27 -0800 Message-Id: <199902222309.PAA24586@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <5l90dqavn9.fsf@tequila.cs.yale.edu> from Stefan Monnier at "Feb 22, 99 06:03:38 pm" To: monnier+lists/linux/autofs/news/@tequila.cs.yale.edu (Stefan Monnier) Date: Mon, 22 Feb 1999 15:09:27 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > >> * Direct mounts. These appear to require a significant slowdown to the > >> entire VFS, and I do not believe this is justifiable. > > I disagree here... this is an compatability issue. Whether it makes > > I've been wondering about the feasability of understanding direct mounts > in the maps, while implementing them via symlinks to an add-hoc autofs mount > point. It seems very doable right now with autofs v3 plus a few hacks > (implemented via lookup_program.so). Would that be sufficient ? That's trivial to do, obviously. -hpa From owner-autofs@linux.kernel.org Mon Feb 22 16:37:46 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA26361 for ; Mon, 22 Feb 1999 16:37:45 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA08910 for autofs-list; Mon, 22 Feb 1999 16:13:58 -0800 Received: from palrel3.hp.com (palrel3.hp.com [156.153.255.226]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA08907 for ; Mon, 22 Feb 1999 16:13:56 -0800 Received: from hpdmd48.boi.hp.com (hpdmd48.boi.hp.com [15.56.8.48]) by palrel3.hp.com (8.8.6 (PHNE_14041)/8.8.5tis) with ESMTP id QAA11102 for ; Mon, 22 Feb 1999 16:14:06 -0800 (PST) Received: from boi.hp.com (anderson.boi.hp.com) by hpdmd48.boi.hp.com with ESMTP (1.37.109.20/15.5+ECS 3.4 ) id AA112898824; Mon, 22 Feb 1999 17:13:44 -0700 Message-Id: <36D1F1CB.FCE7A6B9@boi.hp.com> Date: Tue, 23 Feb 1999 00:09:47 +0000 From: Bill Anderson Organization: Alternative Resources X-Mailer: Mozilla 4.5 [en] (X11; U; Linux 2.2.0-final i686) X-Accept-Language: en Mime-Version: 1.0 To: "autofs@linux.kernel.org" Subject: question regarding maps Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Ok, I have not delved too deep into the documentation, but i have yet to fond what I am looking for (oh, for time to spare ;^). I need to do something like(in auto.whatever): if $HOST=$foo: user $HOST:/home/user if $HOST!=$foo: user $HOST:/auto/home/user Is this possible for the current implementation (v3)? If not, how difficult would it be to put it in for v4? Thanks, Bill Anderson From owner-autofs@linux.kernel.org Mon Feb 22 17:14:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA26384 for ; Mon, 22 Feb 1999 17:14:16 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA10160 for autofs-list; Mon, 22 Feb 1999 16:41:00 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA10157 for ; Mon, 22 Feb 1999 16:40:58 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id QAA08695; Mon, 22 Feb 1999 16:39:46 -0800 From: Alvin Oga Message-Id: <199902230039.QAA08695@leopard.imedia.com> Subject: Re: autofs v4: speak now or forever hold your peace To: hpa@transmeta.com (H. Peter Anvin) Date: Mon, 22 Feb 1999 16:39:46 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902221302.FAA09677@korell.transmeta.com> from "H. Peter Anvin" at Feb 22, 99 05:02:01 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi peter thanx for working on autofs... good stuff for what it does will there be a feature for, if one server is not available, than it will look for the next available server ? ( multi-server environment ) home home1,home2:/home where /home is ONE directory.....on two servers having access to it....say in a dual-host disk controller...a with raid drives... all client linux boxes get's their /home mounted from one of the two home servers... BTW... donno about you folks out there....but I would assume you all use autofs to mount /var/spool/mail, /home, /usr/local, et.al thanx alvin BTW, I assume I can put the following into the autofs-howto feature list ? ( ie since it's now publicly released in the mailing list... ( it's no longer a secret ?? > Okay, I'm planning to do outline the basic data structures of autofs > v4 over the next several days. Some features are already a given: > > * Multimounts/scaffolding (/net being a special case of this). > * Arbitrary mount point topology, without needing the > spawn-an-automounter hack. > * Mounting-process information passed to daemon. > * Most of the filesystem data will live in "ready-to-eat" kernel data > structures (inodes, dentries) as opposed to separate backing store. > > A few frequently requested features that are *not* included, so don't > ask: > > * Direct mounts. These appear to require a significant slowdown to the > entire VFS, and I do not believe this is justifiable. > * lofs. Not because I don't think it is a good idea if it can be done > right, but because it is a feature orthogonal to autofs. > * Replicated servers. This is an NFS issue, not autofs. This means > it belongs in mount(8) and the NFS kernel code. I completely refuse > to hack in a feature into autofs when it is clear it should be > provided elsewhere, *especially* when it is a feature like this when > it would be more useful if provided elsewhere. After all, why > should you be able to mount a replicated server from autofs but not > from /etc/fstab?? > > However, if there is some key feature that I'm missing, please speak > up now: please send an email explaining the feature and its > justification, and I will try to fit it into the design. > > Oh, autofs v4 will be targetted for the 2.3 kernels and libc6. 2.3 > because I am planning to ask for certain VFS changes, which probably > won't be possible in the 2.2 kernel series, and libc6 because I need a > thread-safe dynamic linker, which isn't available for libc5. > > -hpa > From owner-autofs@linux.kernel.org Mon Feb 22 17:16:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA26410 for ; Mon, 22 Feb 1999 17:16:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA10616 for autofs-list; Mon, 22 Feb 1999 16:51:26 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA10612 for ; Mon, 22 Feb 1999 16:51:25 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id QAA09493; Mon, 22 Feb 1999 16:47:56 -0800 From: Alvin Oga Message-Id: <199902230047.QAA09493@leopard.imedia.com> Subject: Re: question regarding maps - scripts To: banderson@boi.hp.com (Bill Anderson) Date: Mon, 22 Feb 1999 16:47:55 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <36D1F1CB.FCE7A6B9@boi.hp.com> from "Bill Anderson" at Feb 23, 99 00:09:47 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya bill try putting that stuff in the autofs startup script /etc/rc.d/init.d/autofs case "$1" in start) # -->>> if [ $HOST = $foo ]; then -->>> cp /etc/autofs/auto.foo /etc/auto.misc -->>> else -->>> cp /etc/autofs/auto.default /etc/auto.misc -->>> fi # # Check if the automounter is already running? if [ ! -f /var/lock/subsys/automount ]; then echo 'Starting automounter: ' getmounts | sh touch /var/lock/subsys/automount fi ;; stop) kill -TERM $(/sbin/pidof /usr/sbin/automount) rm -f /var/lock/subsys/automount ;; have fun alvin > Ok, I have not delved too deep into the documentation, but i have yet to > fond what I am looking for (oh, for time to spare ;^). > > I need to do something like(in auto.whatever): > if $HOST=$foo: user $HOST:/home/user > if $HOST!=$foo: user $HOST:/auto/home/user > > Is this possible for the current implementation (v3)? > If not, how difficult would it be to put it in for v4? > > > Thanks, > > Bill Anderson > From owner-autofs@linux.kernel.org Mon Feb 22 17:56:47 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA26446 for ; Mon, 22 Feb 1999 17:56:47 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id RAA12486 for autofs-list; Mon, 22 Feb 1999 17:29:07 -0800 Received: from tequila.cs.yale.edu (TEQUILA.CS.YALE.EDU [128.36.229.152]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id RAA12483 for ; Mon, 22 Feb 1999 17:29:05 -0800 Received: from tequila.cs.yale.edu (localhost [127.0.0.1]) by tequila.cs.yale.edu (8.8.7/8.8.7) with SMTP id UAA32291 for ; Mon, 22 Feb 1999 20:29:04 -0500 To: autofs@linux.kernel.org From: Stefan Monnier Newsgroups: lists.linux.autofs Subject: Re: question regarding maps References: <36D1F1CB.FCE7A6B9@boi.hp.com> Date: 22 Feb 1999 20:28:58 -0500 Message-ID: <5l678tc3hh.fsf@tequila.cs.yale.edu> Lines: 10 X-Newsreader: Gnus v5.5/Emacs 20.3 Path: tequila.cs.yale.edu NNTP-Posting-Host: tequila.cs.yale.edu X-Trace: 22 Feb 1999 20:28:59 -0500, tequila.cs.yale.edu Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: >>>>> "Bill" == Bill Anderson writes: > I need to do something like(in auto.whatever): > if $HOST=$foo: user $HOST:/home/user > if $HOST!=$foo: user $HOST:/auto/home/user > Is this possible for the current implementation (v3)? With lookup_program.so, yes, of course! Stefan From owner-autofs@linux.kernel.org Mon Feb 22 23:25:33 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id XAA26682 for ; Mon, 22 Feb 1999 23:25:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id WAA08017 for autofs-list; Mon, 22 Feb 1999 22:59:21 -0800 Received: from ewald.mbi.ucla.edu (ewald.mbi.ucla.edu [128.97.39.21]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id WAA08008 for ; Mon, 22 Feb 1999 22:59:18 -0800 Received: from localhost (brian@localhost) by ewald.mbi.ucla.edu (8.9.1/8.9.1) with SMTP id WAA11880; Mon, 22 Feb 1999 22:59:11 -0800 (PST) Date: Mon, 22 Feb 1999 22:59:10 -0800 (PST) From: Brian Vicente To: "H. Peter Anvin" cc: Stefan Monnier , autofs@linux.kernel.org Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902222309.PAA24586@cesium.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Let me try to convince you on the direct mount: I need to mount all the directories under /usr1/data. If I could have a /- in the auto.master pointed to say auto.data /- auto.data then I could automount this one directory with: data machine:/usr1/data and be done with it. Now I have to put the following in the auto.master /data auto.data and then in the auto.data I put dir1 machine:/usr/data/dir1 dir2 machine:/usr/data/dir2 . . The benefit of this would be to not having everyone to know all the sub-directories under /data or having to guess these. Of course if someone can point out a better way to do this I'd welcome it. I've thought about having /data/data/dir1 etc already. That's a hoakie fix. Thanks, Brian " The Great End of Life is not Knowledge but Action." - T.H. Huxley - "If you would persuade, you must appeal to interest rather than intellect." - Benjamin Franklin - On Mon, 22 Feb 1999, H. Peter Anvin wrote: > > >> * Direct mounts. These appear to require a significant slowdown to the > > >> entire VFS, and I do not believe this is justifiable. > > > I disagree here... this is an compatability issue. Whether it makes > > > > I've been wondering about the feasability of understanding direct mounts > > in the maps, while implementing them via symlinks to an add-hoc autofs mount > > point. It seems very doable right now with autofs v3 plus a few hacks > > (implemented via lookup_program.so). Would that be sufficient ? > > That's trivial to do, obviously. > > -hpa > From owner-autofs@linux.kernel.org Tue Feb 23 00:38:06 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id AAA26769 for ; Tue, 23 Feb 1999 00:38:06 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id XAA14010 for autofs-list; Mon, 22 Feb 1999 23:55:46 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id XAA14001 for ; Mon, 22 Feb 1999 23:55:40 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id IAA08774; Tue, 23 Feb 1999 08:55:35 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id IAA09020; Tue, 23 Feb 1999 08:55:35 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Tue, 23 Feb 1999 08:55:35 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Brian Vicente Subject: Re: autofs v4: speak now or forever hold your peace Cc: Stefan Monnier , autofs@linux.kernel.org, "H. Peter Anvin" Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thus spoke Brian Vicente on 23-Feb-99 : > Let me try to convince you on the direct mount: > I need to mount all the directories under /usr1/data. > If I could have a /- in the auto.master pointed to say auto.data > /- auto.data > then I could automount this one directory with: > data machine:/usr1/data > and be done with it. > Now I have to put the following in the auto.master > /data auto.data > and then in the auto.data I put > dir1 machine:/usr/data/dir1 > dir2 machine:/usr/data/dir2 > . Sorry, I do not see the problem. Maybe I do misunderstand, what you want. But I had the same problem, to mount a tree of directories from a machine, without the user guessing + plus the mechnanis should work for any host whithout changing the automounter maps. Here is my solution within the existing autofs. 1) I reorganized all my nfs servers to have mount points like /hostname/data1 /hostname/data2 2) I created an automounter moint point /imports and a map auto.imports * &:/& That was all. A simple universal map, whcih needs no change, if you buy a new machine. We make local temp space for calcuation intensive programs availible that way to all hosts of the cluster. cd /imports/hostname will mount the whole tree of the other machine and thanks to how LINUX handels exports, I even can dynamicaly mount and unmount directories on the server, which are propagated to the clients. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Tue Feb 23 01:04:07 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA26802 for ; Tue, 23 Feb 1999 01:04:07 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA14803 for autofs-list; Tue, 23 Feb 1999 00:05:23 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA14798 for ; Tue, 23 Feb 1999 00:05:21 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id AAA17428; Tue, 23 Feb 1999 00:04:45 -0800 From: Alvin Oga Message-Id: <199902230804.AAA17428@leopard.imedia.com> Subject: Re: autofs v4: speak now or forever hold your peace To: brian@mbi.ucla.edu (Brian Vicente) Date: Tue, 23 Feb 1999 00:04:45 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Brian Vicente" at Feb 22, 99 10:59:10 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi Brian.... > Let me try to convince you on the direct mount: > I need to mount all the directories under /usr1/data. okay... this is what I would do ( assuming I understood you ) and that the machine is exporting /usr1/data client# cat /etc/auto.master /.autofs /etc/auto.data --timeout 60 client# cat /etc/auto.data # Data -fstype=nfs machine:/usr1/data client# cd / ; ln -s .autofs/Data Data - - restart autofs as needed - client# ls -la /Data and I am done...and should see the contents of machine:/usr1/data/{dir1,dir2,etc} > If I could have a /- in the auto.master pointed to say auto.data > /- auto.data > then I could automount this one directory with: > data machine:/usr1/data > and be done with it. okay... there is a couple more steps with autofs....but my process outlined above works for me... as I see everything on any_server:/ > Now I have to put the following in the auto.master > /data auto.data > and then in the auto.data I put > dir1 machine:/usr/data/dir1 > dir2 machine:/usr/data/dir2 > . > . nah that is NOT necessary...you can if you like but it creates a huge mess on both the server and clients > The benefit of this would be to not having everyone to know all the > sub-directories under /data or having to guess these. if you want to hide directories....you can change group permissions and or create different export directories > Of course if someone can point out a better way to do this I'd welcome it. not necessarily better....just the way I do things... have fun alvin From owner-autofs@linux.kernel.org Tue Feb 23 00:59:44 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id AAA26777 for ; Tue, 23 Feb 1999 00:59:43 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA15204 for autofs-list; Tue, 23 Feb 1999 00:09:50 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA15199 for ; Tue, 23 Feb 1999 00:09:47 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id JAA08810; Tue, 23 Feb 1999 09:09:42 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id JAA09033; Tue, 23 Feb 1999 09:09:41 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: Date: Tue, 23 Feb 1999 09:09:41 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Michael Newlyn Blake Subject: Re: autofs v4: speak now or forever hold your peace Cc: The autofs list Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Thus spoke Michael Newlyn Blake on 22-Feb-99 : >> * Direct mounts. These appear to require a significant slowdown to the >> entire VFS, and I do not believe this is justifiable. > > > > I disagree here... this is an compatability issue. Whether it makes > sense or not, most other implimentations of autofs support this and people > take it for granted. People will continue to see the Linux implimentation > of autofs as buggy or a cheap hack until features like this are taken > seriously. Now that doesn't mean that you should necessarily impliment it > right away, or drop everything to get it working, but don't dismiss it. > It's important. Take your time and get it right, but write it off. > I totally agree. A daemon is a service program which should be able to be started and stopped like any other service program by the system administator. Especially under Linux, which is one of most stable operation systems I worked with, it would help to design daemons in a coherent service friendly manner. Maybe it is just me, but it always hurts me, if I have to reboot a machines, just because I scewed it up so much that no clean simple kills will help any more. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Tue Feb 23 03:26:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA27028 for ; Tue, 23 Feb 1999 03:26:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA29201 for autofs-list; Tue, 23 Feb 1999 02:30:22 -0800 Received: from roma.axis.se (roma.axis.se [193.13.178.2]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA29193 for ; Tue, 23 Feb 1999 02:30:18 -0800 Received: from pcronny.axis.se (pcronny.axis.se [172.19.2.160]) by roma.axis.se (8.8.8/8.8.6) with ESMTP id LAA09629 for ; Tue, 23 Feb 1999 11:30:11 +0100 (MET) Received: (from ronny@localhost) by pcronny.axis.se (8.8.7/8.8.4) id LAA00939; Tue, 23 Feb 1999 11:30:11 +0100 Date: Tue, 23 Feb 1999 11:30:11 +0100 Message-Id: <199902231030.LAA00939@pcronny.axis.se> From: Ronny Ranerup To: autofs@linux.kernel.org In-reply-to: (message from Frithjof Anders on Tue, 23 Feb 1999 09:09:41 +0100 (CET)) Subject: Re: autofs v4: speak now or forever hold your peace References: Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: frithjof@fkp.tu-darmstadt.de: >I totally agree. A daemon is a service program which should be able to be >started and stopped like any other service program by the system administator. >Especially under Linux, which is one of most stable operation systems I worked >with, it would help to design daemons in a coherent service friendly manner. I agree with the agreement. But I guess I also understand that it might be difficult... AMD was the singularly most crappy program for Linux and the only reason we had to reboot our machines (several months apart, but anyway). You know the procedure; hmm, can't login, go to machine, see several screens of amd messages on console, try to kill amd, fail, try to unmount all NFS mounts, fail, try to figure out which bloody proc holds the mount, fail, getting damn cold in the server room, getting mad, killing all user processes, killing amd, killing NFS, bla, bla. Rebooting because the bloody kernel refuses to forget some silly old mount... ... Sorry about the ramblings, had to get in off my chest. But you get the picture. I _really_ would like an easier way to shut down the automounter and flush all mount points. There are reasons you want to do it and there always will be reasons to do it. Have already had to go through the shutdown procedure with autofs because of bugs. Felt ominously like amd. Another point for the wishlist; since I guess most of the known Unix universe is using AMD, a script to parse AMD-maps and generate autofs ones would probably be very helpful, as you can see from this list. I wrote one myself but it's just a trifle and won't do for serious distribution. More whining... after fighting with AMD and trying to figure out where the hell those amd-upl-xxx came from I didn't get happier when trying to find autofs, no official web-site, just a ftp archive, already a frightening number of strange autofs distributions. A NEWS file in the distribution where there are almost no differences between version 0.3.14 and 3.1.2. No HOWTO or similar document. No explanation as to how the daemon ties into the kernel, no explanation on which version of the kernel is needed... Yeah, I know, a lot of whining. Sorry. On the positive side, autofs _might_ be less buggy then AMD... ;) /ronny From owner-autofs@linux.kernel.org Tue Feb 23 04:36:50 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id EAA27139 for ; Tue, 23 Feb 1999 04:36:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id DAA00893 for autofs-list; Tue, 23 Feb 1999 03:29:48 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id DAA00887 for ; Tue, 23 Feb 1999 03:29:45 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id DAA24318; Tue, 23 Feb 1999 03:29:01 -0800 From: Alvin Oga Message-Id: <199902231129.DAA24318@leopard.imedia.com> Subject: Re: autofs v4: speak now .... documents To: ronny.ranerup@axis.com (Ronny Ranerup) Date: Tue, 23 Feb 1999 03:29:01 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902231030.LAA00939@pcronny.axis.se> from "Ronny Ranerup" at Feb 23, 99 11:30:11 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi ya ronny > Another point for the wishlist; since I guess most of the known Unix > universe is using AMD, a script to parse AMD-maps and generate autofs > ones would probably be very helpful, as you can see from this list. > I wrote one myself but it's just a trifle and won't do for serious > distribution. can you forward me a copy of it ?? I have collected some conversion scripts that generates autofs files from NIS, ... ( see autofs-HOWTO.html - section 5.3 ) > More whining... after fighting with AMD and trying to figure out where > the hell those amd-upl-xxx came from I didn't get happier when trying ftp://sunsite.unc.edu:/pub/Linux/system/mount amd920824upl67.tar.gz ... etc ... > to find autofs, no official web-site ftp.sunsite.unc.edu/pub/Linux/system/mount/autofs-* ftp.kernel.org:/pub/linux/daemons/autofs/autofs-* > just a ftp archive, already a > frightening number of strange autofs distributions. huh ??? other versions than what Peter is maintaining ??? > A NEWS file in the > distribution where there are almost no differences between version > 0.3.14 and 3.1.2. > No HOWTO or similar document. http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > No explanation as to > how the daemon ties into the kernel, no explanation on which version > of the kernel is needed... see above howto... > Yeah, I know, a lot of whining. Sorry. On the positive side, autofs > _might_ be less buggy then AMD... ;) works an order of magnitude better than amd for simple automounting of /var/spool/mail, /home, /usr/local have fun alvin http://www.Linux-Consulting.com/Amd_AutoFS From owner-autofs@linux.kernel.org Tue Feb 23 06:43:39 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id GAA27245 for ; Tue, 23 Feb 1999 06:43:39 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA13817 for autofs-list; Tue, 23 Feb 1999 05:18:50 -0800 Received: from roma.axis.se (roma.axis.se [193.13.178.2]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA13808 for ; Tue, 23 Feb 1999 05:18:47 -0800 Received: from pcronny.axis.se (pcronny.axis.se [172.19.2.160]) by roma.axis.se (8.8.8/8.8.6) with ESMTP id OAA15445 for ; Tue, 23 Feb 1999 14:18:45 +0100 (MET) Received: (from ronny@localhost) by pcronny.axis.se (8.8.7/8.8.4) id OAA01371; Tue, 23 Feb 1999 14:18:45 +0100 Date: Tue, 23 Feb 1999 14:18:45 +0100 Message-Id: <199902231318.OAA01371@pcronny.axis.se> From: Ronny Ranerup To: autofs@linux.kernel.org In-reply-to: <199902231129.DAA24318@leopard.imedia.com> (message from Alvin Oga on Tue, 23 Feb 1999 03:29:01 -0800 (PST)) Subject: Re: autofs v4: speak now .... documents References: <199902231129.DAA24318@leopard.imedia.com> Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: aoga@imedia.com: >http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html Ahh, excellent, now if this could be included in the distribution or at least a link to it in the man pages of the distribution we would probably make someones life a little easier. I haven't checked all of the HOWTO, but methinks there's no info on how it ties into the kernel and some on the inner workings of the beast... (Not that it necessarily should be in the HOWTO.) And yes there's always TFS, but if I would check the source for every program I wanted to figure out, I'd do nothing else. Linus', and others, attitude of not documenting anything either inside our outside the code I find kind of childish, 'This is mine, mine, mine, and if you don't understand it don't mess with it...' My company (were I work, that is, not own.) is porting Linux to our new CPU and we would enjoy it even more if the comments in it weren't from the beginning of the 90s... and incorrect... and outdated... >ftp.sunsite.unc.edu/pub/Linux/system/mount/autofs-* Yes, I know but I meant one of those new thingies with HTML stuff and the like. The -official-homepage-where-you-find-everything-you-need-to-know-\ and-then-some-without-spending-an-afternoon-scouring-the-net-and-\ just-finding-the-usual-naked-girls- kind of deal... >works an order of magnitude better than amd for simple automounting of >/var/spool/mail, /home, /usr/local I'll take your word on it. I moved a couple of the most used mounts from AMD into /etc/fstab so, as I said, it just crashed a couple of months apart. I've only used autofs for a couple of months now and it has only crashed once yet. And I'm sorry that I can't elaborate much on this specific bug. (I had a disk (S) mounted with autofs on the client (C), and on the server I messed around with mounting another local disk (S2) under S, unmounting it, mounting it somewhere else on the server, making a link from under S to it. Somewhere along the way autofs got screwed up and the S still existed but was empty. If I mounted S in another autofs path it worked but the original path wouldn't work. Had to shut down autofs, unmount everything, and restart it) By the way hpa, please, please, please, more debug options settable at runtime, like in AMD. And not something you should recompile to enable. If they're in there I might have the energy to enable them to see if I can figure out what goes wrong. If I have to recompile, I probably won't do it and the bug will stay in there til someone with more energy and interest finds it... Then again, it didn't help AMD, so perhaps it's no use... >huh ??? other versions than what Peter is maintaining ??? Dunno, I didn't check the contents. I found the ftp-site above before I got that desperate. Damn, my mails do get long. Guess it's the poet in me... /ronny From owner-autofs@linux.kernel.org Tue Feb 23 06:55:38 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id GAA27251 for ; Tue, 23 Feb 1999 06:55:38 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id FAA16168 for autofs-list; Tue, 23 Feb 1999 05:36:43 -0800 Received: from tequila.cs.yale.edu (TEQUILA.CS.YALE.EDU [128.36.229.152]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id FAA16160 for ; Tue, 23 Feb 1999 05:36:39 -0800 Received: from tequila.cs.yale.edu (localhost [127.0.0.1]) by tequila.cs.yale.edu (8.8.7/8.8.7) with SMTP id IAA04356 for ; Tue, 23 Feb 1999 08:36:32 -0500 To: autofs@linux.kernel.org From: Stefan Monnier Newsgroups: lists.linux.autofs Subject: Re: autofs v4: speak now or forever hold your peace References: Date: 23 Feb 1999 08:36:27 -0500 Message-ID: <5l1zjhb5t0.fsf@tequila.cs.yale.edu> Lines: 8 X-Newsreader: Gnus v5.5/Emacs 20.3 Path: tequila.cs.yale.edu NNTP-Posting-Host: tequila.cs.yale.edu X-Trace: 23 Feb 1999 08:36:27 -0500, tequila.cs.yale.edu Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: >>>>> "Frithjof" == Frithjof Anders writes: > thanks to how LINUX handels exports, You mean: `thanks to how Linux's user-lever nfs server handles exports,' The kernel nfsd behaves just like on other Unixes. Stefan From owner-autofs@linux.kernel.org Tue Feb 23 07:18:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id HAA27283 for ; Tue, 23 Feb 1999 07:18:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id GAA20732 for autofs-list; Tue, 23 Feb 1999 06:20:42 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id GAA20721 for ; Tue, 23 Feb 1999 06:20:37 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id GAA27162; Tue, 23 Feb 1999 06:20:05 -0800 From: Alvin Oga Message-Id: <199902231420.GAA27162@leopard.imedia.com> Subject: Re: autofs v4: speak now .... documents To: ronny.ranerup@axis.com (Ronny Ranerup) Date: Tue, 23 Feb 1999 06:20:05 -0800 (PST) Cc: autofs@linux.kernel.org, aoga@leopard.imedia.com (Alvin Oga) In-Reply-To: <199902231318.OAA01371@pcronny.axis.se> from "Ronny Ranerup" at Feb 23, 99 02:18:45 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya ronny > aoga@imedia.com: > >http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > > Ahh, excellent, now if this could be included in the distribution or > at least a link to it in the man pages of the distribution we would > probably make someones life a little easier. I submitted it the folks that maintain the howto's but i did not get any replies or comments/suggestions...donno... > attitude of not documenting anything either inside our outside the > code I find kind of childish, 'This is mine, mine, mine, and if you > don't understand it don't mess with it...' I concur with that approach....also works for "if ain't broke, don't touch it" and also works for "it's all messed and broken, you fix it instead of complaining....applies really well to the other operating system users" > My company (were I work, > that is, not own.) is porting Linux to our new CPU and we would enjoy > it even more if the comments in it weren't from the beginning of the > 90s... and incorrect... and outdated... which docs are outdated ?? the raid stuff was pretty bad...so i started doing some of that documentation too > >ftp.sunsite.unc.edu/pub/Linux/system/mount/autofs-* > > Yes, I know but I meant one of those new thingies with HTML stuff and > the like. The -official-homepage-where-you-find-everything-you-need-to-know-\ > and-then-some-without-spending-an-afternoon-scouring-the-net-and-\ > just-finding-the-usual-naked-girls- kind of deal... tried to do that with the autofs-howto... anything in there that is missing ?? have fun alvin From owner-autofs@linux.kernel.org Tue Feb 23 08:21:31 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA27339 for ; Tue, 23 Feb 1999 08:21:31 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA25423 for autofs-list; Tue, 23 Feb 1999 07:17:18 -0800 Received: from mail13.digital.com (mail13.digital.com [192.208.46.30]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA25418 for ; Tue, 23 Feb 1999 07:17:15 -0800 Received: from wasted.zk3.dec.com (bywasted.zk3.dec.com [16.140.96.51]) by mail13.digital.com (8.9.1a/8.9.1/WV2.0d) with ESMTP id KAA11062; Tue, 23 Feb 1999 10:17:13 -0500 (EST) Received: from localhost by wasted.zk3.dec.com (8.8.8/1.1.22.2/08Sep98-0251PM) id KAA0000009443; Tue, 23 Feb 1999 10:17:11 -0500 (EST) From: Eric Werme USG Message-Id: <199902231517.KAA0000009443@wasted.zk3.dec.com> X-Authentication-Warning: wasted.zk3.dec.com: localhost [127.0.0.1] didn't use HELO protocol To: frithjof@fkp.tu-darmstadt.de cc: Stefan Monnier , autofs@linux.kernel.org, "H. Peter Anvin" Subject: Re: autofs v4: speak now or forever hold your peace In-reply-to: Your message of "Tue, 23 Feb 1999 08:55:35 +0100." Date: Tue, 23 Feb 1999 10:17:11 -0500 X-Mts: smtp Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Sorry, I do not see the problem. Maybe I do misunderstand, what you want. Bu t I had the same problem, to mount a tree of directories from a machine, without the user guessing + plus the mechnanis should work for any host whithout changing the automounter maps. Here is my solution within the existing autof s. 1) I reorganized all my nfs servers to have mount points like /hostname/data1 /hostname/data2 Try coming here for a while and doing that. While "here" uses a subset of the 16.0.0.0 Class A address space, we have more servers than one organization can support, let alone one person! :-) Also, don't forget some of the NFS-served Internet archives like gatekeeper.pa.dec.com (which is not "here", but three time zones away) and one at wustl.edu. They may be good candidates for direct maps and multiple mounts of a subset of what's available. -Ric Werme -- Eric (Ric) Werme | werme@zk3.dec.com Compaq Computer Corp. | http://www.cyberportal.net/werme From owner-autofs@linux.kernel.org Tue Feb 23 08:21:49 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA27343 for ; Tue, 23 Feb 1999 08:21:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA25771 for autofs-list; Tue, 23 Feb 1999 07:20:28 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA25766 for ; Tue, 23 Feb 1999 07:20:26 -0800 Received: from zoot.corp.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id HAA03068 for ; Tue, 23 Feb 1999 07:20:15 -0800 X-BlackMail: robin.frogtown.com, zoot.corp.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 07:20:15(PST) on February 23, 1999 Date: Tue, 23 Feb 1999 07:20:00 -0800 (PST) From: Lori X-Sender: lori@zoot.corp.frogtown.com cc: autofs@linux.kernel.org Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902230804.AAA17428@leopard.imedia.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 23 Feb 1999, Alvin Oga wrote: > client# cat /etc/auto.master > /.autofs /etc/auto.data --timeout 60 > > client# cat /etc/auto.data > # > Data -fstype=nfs machine:/usr1/data > > client# cd / ; ln -s .autofs/Data Data the problem with this method is the clients all have to be modified to utilize this scheme. it's not an out of the box solution. This may not seem to be a problem to some people, but it is if you get a couple hundred boxes or a bunch of engineers who constantly reinstall their systems. The direct mount map can be put in the NIS master auto.master file indicate the other maps and no additional work need be done on the client side. I'm not necessarily advocating doing less work, just that it makes life less complex for those you support if you can just hand them a box or have them reinstall their own system without a bunch of rules and regs before it'll run. When I last set up a bunch of NIS/autofs systems on the network I wanted the clients to be as generic as possible, this was a lab environment and it was important that they be easily reinstalled without a bunch of local hacks - I had to set up two scenarios depending on what OS you were running - solaris or linux. for solaris I set up /usr/dist, and for linux I had to set up /import/dist (among other mounts related to function). This was a workable solution in our environment because there wasn't anything before I set it up, and the solaris systems could use the /import scenario, also. But it was still two different set-ups that were operating system dependant. and therefore more confusion for the end-user. cheers! From owner-autofs@linux.kernel.org Tue Feb 23 08:29:00 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA27350 for ; Tue, 23 Feb 1999 08:29:00 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA25628 for autofs-list; Tue, 23 Feb 1999 07:19:21 -0800 Received: from roma.axis.se (roma.axis.se [193.13.178.2]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA25624 for ; Tue, 23 Feb 1999 07:19:19 -0800 Received: from pcronny.axis.se (pcronny.axis.se [172.19.2.160]) by roma.axis.se (8.8.8/8.8.6) with ESMTP id QAA19493 for ; Tue, 23 Feb 1999 16:19:18 +0100 (MET) Received: (from ronny@localhost) by pcronny.axis.se (8.8.7/8.8.4) id QAA01503; Tue, 23 Feb 1999 16:19:18 +0100 Date: Tue, 23 Feb 1999 16:19:18 +0100 Message-Id: <199902231519.QAA01503@pcronny.axis.se> From: Ronny Ranerup To: autofs@linux.kernel.org In-reply-to: <199902231420.GAA27162@leopard.imedia.com> (message from Alvin Oga on Tue, 23 Feb 1999 06:20:05 -0800 (PST)) Subject: Re: autofs v4: [just barely...] References: <199902231420.GAA27162@leopard.imedia.com> Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Not very autofs specific. Ignore the rest if not interested... aoga@imedia.com >I concur with that approach....also works for "if ain't broke, don't touch it" >and also works for "it's all messed and broken, you fix it instead >of complaining....applies really well to the other operating system users" Hmm, I'm not that fluent in english but I guess you agree with Linus et al. That's fine. Of course, not many companies in this business would agree with that, for obvious reasons. >which docs are outdated ?? Are you kidding? I was messing round in the kernel yesterday to try and figure out how copy-on-write in fork is implemented and the details of the page table handling. The comments in the code are _very_ sparse, and had references to files that haven't existed in Linux for many years. The MMU support in general isn't documented anywhere but in the LDP and they plainly state that they are outdated and that you should buy one of the three kernel books that exist. And the books explain how it works in general (when the books were written that is, and Linux changes a lot faster than the books are printed), but that I already know. The details of the implementation is only documented in the source, as source, and is probably understood by a handful of people on this entire planet. This perhaps works fine for that handful of people, but I think it kind of defeats the meaning of open source. Some comments in, or outside, the code might open it up for a couple of dozen more developers. Of course if you are Linus and still want to filter every bit changed in the kernel, you might want to keep the changes to an absolute minimum and, other then not giving away the source, removing the comments in the fairly obfuscated code probably does the trick just fine... To bring this back to autofs, I just checked the source and well, some comments and documentation wouldn't hurt, would it hpa? Excellent work otherwise, of course. Don't want to put down the effort and result of the people doing this stuff... Well, there are pros and cons to absolutely everything. All I really know is that after a couple of days of digging in the memory handling in the kernel all I want to say is, char l,i;;main(k){for(l;k=i["]^k[k]^i[i]++i["];read('/'/'/',i+++"*5>+?480m{oKK\ \211\221",-('-'-'-'-k),-k));}read(j,i,k){l=+' '+(k-((*(char*)i)-(j<<(j<>(j&j)),j&j&&j);} /ronny From owner-autofs@linux.kernel.org Tue Feb 23 08:37:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id IAA27354 for ; Tue, 23 Feb 1999 08:37:27 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id HAA26469 for autofs-list; Tue, 23 Feb 1999 07:29:26 -0800 Received: from tiktok.cygnus.com (tiktok.cygnus.com [192.80.44.20]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id HAA26464 for ; Tue, 23 Feb 1999 07:29:24 -0800 Received: (from meissner@localhost) by tiktok.cygnus.com (8.8.7/8.8.5) id KAA09154; Tue, 23 Feb 1999 10:29:07 -0500 Date: Tue, 23 Feb 1999 10:29:07 -0500 From: Michael Meissner To: "H. Peter Anvin" Cc: autofs mailing list Subject: Re: autofs v4: speak now or forever hold your peace Message-ID: <19990223102907.A9100@tiktok.cygnus.com> References: <199902221302.FAA09677@korell.transmeta.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95i In-Reply-To: <199902221302.FAA09677@korell.transmeta.com>; from H. Peter Anvin on Mon, Feb 22, 1999 at 05:02:01AM -0800 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Mon, Feb 22, 1999 at 05:02:01AM -0800, H. Peter Anvin wrote: > Okay, I'm planning to do outline the basic data structures of autofs > v4 over the next several days. Some features are already a given: > > * Multimounts/scaffolding (/net being a special case of this). > * Arbitrary mount point topology, without needing the > spawn-an-automounter hack. > * Mounting-process information passed to daemon. > * Most of the filesystem data will live in "ready-to-eat" kernel data > structures (inodes, dentries) as opposed to separate backing store. One thing that amd has that would be nice if autofs has is symlink mounts. If for instance, you have two systems, fred and barney. Fred exports /pebbles and /dino, while barney exports /bambam, you might want to have a centralized mount point for all disks: /mnt/pebbles /mnt/bambam /mnt/dino If you were on fred, /mnt/pebbles would be a symlink to /pebbles, /mnt/dino would be a symlink to /dino (yeah, direct mounts would be great, but they won't be provided), and /mnt/bambam would be automounted from barney. Even nicer if you could share autofs maps between systems like amd did (under amd, you could say if the system name was fred, do symlinks, otherwise do nfs mounts). -- Michael Meissner, Cygnus Solutions (Massachusetts office) 4th floor, 955 Massachusetts Avenue, Cambridge, MA 02139, USA meissner@cygnus.com, 617-354-5416 (office), 617-354-7161 (fax) From owner-autofs@linux.kernel.org Tue Feb 23 09:07:52 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA27366 for ; Tue, 23 Feb 1999 09:07:52 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id IAA31616 for autofs-list; Tue, 23 Feb 1999 08:23:09 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id IAA31609 for ; Tue, 23 Feb 1999 08:23:04 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id RAA10090; Tue, 23 Feb 1999 17:23:05 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id RAA30328; Tue, 23 Feb 1999 17:23:05 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <19990223102907.A9100@tiktok.cygnus.com> Date: Tue, 23 Feb 1999 17:23:05 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Michael Meissner Subject: Re: autofs v4: speak now or forever hold your peace Cc: autofs mailing list Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke Michael Meissner on 23-Feb-99 : > On Mon, Feb 22, 1999 at 05:02:01AM -0800, H. Peter Anvin wrote: >> Okay, I'm planning to do outline the basic data structures of autofs >> v4 over the next several days. Some features are already a given: >> >> * Multimounts/scaffolding (/net being a special case of this). >> * Arbitrary mount point topology, without needing the >> spawn-an-automounter hack. >> * Mounting-process information passed to daemon. >> * Most of the filesystem data will live in "ready-to-eat" kernel data >> structures (inodes, dentries) as opposed to separate backing store. > > One thing that amd has that would be nice if autofs has is symlink mounts. > If for instance, you have two systems, fred and barney. Fred exports > /pebbles > and /dino, while barney exports /bambam, you might want to have a centralized > mount point for all disks: > > /mnt/pebbles > /mnt/bambam > /mnt/dino > > If you were on fred, /mnt/pebbles would be a symlink to /pebbles, /mnt/dino That work already :) "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Tue Feb 23 12:45:57 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA27518 for ; Tue, 23 Feb 1999 12:45:57 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA17692 for autofs-list; Tue, 23 Feb 1999 12:11:21 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA17686 for ; Tue, 23 Feb 1999 12:11:18 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id MAA28152; Tue, 23 Feb 1999 12:10:52 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id MAA22208; Tue, 23 Feb 1999 12:10:52 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id MAA22325; Tue, 23 Feb 1999 12:10:52 -0800 Message-Id: <199902232010.MAA22325@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Lori at "Feb 23, 99 07:20:00 am" To: lori@Frogtown.com (Lori) Date: Tue, 23 Feb 1999 12:10:52 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > On Tue, 23 Feb 1999, Alvin Oga wrote: > > > client# cat /etc/auto.master > > /.autofs /etc/auto.data --timeout 60 > > > > client# cat /etc/auto.data > > # > > Data -fstype=nfs machine:/usr1/data > > > > client# cd / ; ln -s .autofs/Data Data > > the problem with this method is the clients all have to be modified to > utilize this scheme. it's not an out of the box solution. This may not > seem to be a problem to some people, but it is if you get a couple hundred > boxes or a bunch of engineers who constantly reinstall their systems. The > direct mount map can be put in the NIS master auto.master file indicate > the other maps and no additional work need be done on the client side. I'm > not necessarily advocating doing less work, just that it makes life less > complex for those you support if you can just hand them a box or have them > reinstall their own system without a bunch of rules and regs before it'll > run. > > When I last set up a bunch of NIS/autofs systems on the network I wanted > the clients to be as generic as possible, this was a lab environment and > it was important that they be easily reinstalled without a bunch of local > hacks - I had to set up two scenarios depending on what OS you were > running - solaris or linux. for solaris I set up /usr/dist, and for linux > I had to set up /import/dist (among other mounts related to function). > This was a workable solution in our environment because there wasn't > anything before I set it up, and the solaris systems could use the /import > scenario, also. But it was still two different set-ups that were operating > system dependant. and therefore more confusion for the end-user. > What I might try to do, is when parsing auto.master to replace all direct mounts with a link into /.a or something, and make /.a an indirect mount. -hpa From owner-autofs@linux.kernel.org Tue Feb 23 12:46:49 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA27522 for ; Tue, 23 Feb 1999 12:46:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA17483 for autofs-list; Tue, 23 Feb 1999 12:08:35 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA17478 for ; Tue, 23 Feb 1999 12:08:32 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id MAA28043; Tue, 23 Feb 1999 12:03:22 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id MAA21855; Tue, 23 Feb 1999 12:03:21 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id MAA22263; Tue, 23 Feb 1999 12:03:21 -0800 Message-Id: <199902232003.MAA22263@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Frithjof Anders at "Feb 23, 99 09:09:41 am" To: frithjof@fkp.tu-darmstadt.de Date: Tue, 23 Feb 1999 12:03:21 -0800 (PST) Cc: mblake@Frogtown.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > I totally agree. A daemon is a service program which should be able > to be started and stopped like any other service program by the > system administator. Especially under Linux, which is one of most > stable operation systems I worked with, it would help to design > daemons in a coherent service friendly manner. > > > Maybe it is just me, but it always hurts me, if I have to reboot a machines, > just because I scewed it up so much that no clean simple kills will help any > more. Note you can certainly stop autofs, but you might have to clean up manually before you can restart. The only other option would be for autofs to refuse to shut down if any of its filesystems are in use, like it does if it gets SIGUSR1 (I think...) -hpa From owner-autofs@linux.kernel.org Tue Feb 23 12:50:12 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA27526 for ; Tue, 23 Feb 1999 12:50:12 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA17175 for autofs-list; Tue, 23 Feb 1999 12:02:18 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA17170 for ; Tue, 23 Feb 1999 12:02:15 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id MAA28006; Tue, 23 Feb 1999 12:01:42 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id MAA21510; Tue, 23 Feb 1999 12:01:42 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id MAA22249; Tue, 23 Feb 1999 12:01:42 -0800 Message-Id: <199902232001.MAA22249@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Brian Vicente at "Feb 22, 99 10:59:10 pm" To: brian@mbi.ucla.edu (Brian Vicente) Date: Tue, 23 Feb 1999 12:01:41 -0800 (PST) Cc: hpa@transmeta.com, monnier+lists/linux/autofs/news/@tequila.cs.yale.edu, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Let me try to convince you on the direct mount: > I need to mount all the directories under /usr1/data. > If I could have a /- in the auto.master pointed to say auto.data > /- auto.data > then I could automount this one directory with: > data machine:/usr1/data > and be done with it. > Now I have to put the following in the auto.master > /data auto.data > and then in the auto.data I put > dir1 machine:/usr/data/dir1 > dir2 machine:/usr/data/dir2 > . > . > The benefit of this would be to not having everyone to know all the > sub-directories under /data or having to guess these. > Of course if someone can point out a better way to do this I'd welcome it. > I've thought about having /data/data/dir1 etc already. That's a hoakie > fix. > Thanks, > Brian It won't happen. Get over it. -hpa From owner-autofs@linux.kernel.org Tue Feb 23 12:55:45 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA27536 for ; Tue, 23 Feb 1999 12:55:45 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA17600 for autofs-list; Tue, 23 Feb 1999 12:10:08 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA17572 for ; Tue, 23 Feb 1999 12:10:01 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id MAA28139; Tue, 23 Feb 1999 12:09:58 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id MAA22165; Tue, 23 Feb 1999 12:09:57 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id MAA22298; Tue, 23 Feb 1999 12:09:57 -0800 Message-Id: <199902232009.MAA22298@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <19990223102907.A9100@tiktok.cygnus.com> from Michael Meissner at "Feb 23, 99 10:29:07 am" To: meissner@cygnus.com (Michael Meissner) Date: Tue, 23 Feb 1999 12:09:57 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > One thing that amd has that would be nice if autofs has is symlink > mounts. If for instance, you have two systems, fred and barney. > Fred exports /pebbles and /dino, while barney exports /bambam, you > might want to have a centralized mount point for all disks: > > /mnt/pebbles > /mnt/bambam > /mnt/dino > > If you were on fred, /mnt/pebbles would be a symlink to /pebbles, > /mnt/dino would be a symlink to /dino (yeah, direct mounts would be > great, but they won't be provided), and /mnt/bambam would be > automounted from barney. Even nicer if you could share autofs maps > between systems like amd did (under amd, you could say if the system > name was fred, do symlinks, otherwise do nfs mounts). autofs does that automatically. If a mount identified as NFS is on the local host, it will use a symlink. -hpa From owner-autofs@linux.kernel.org Tue Feb 23 12:57:06 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id MAA27540 for ; Tue, 23 Feb 1999 12:57:06 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA18154 for autofs-list; Tue, 23 Feb 1999 12:17:59 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA18150 for ; Tue, 23 Feb 1999 12:17:57 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id MAA28261; Tue, 23 Feb 1999 12:17:20 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id MAA22638; Tue, 23 Feb 1999 12:17:15 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id MAA22378; Tue, 23 Feb 1999 12:17:15 -0800 Message-Id: <199902232017.MAA22378@cesium.transmeta.com> Subject: Re: autofs v4: speak now .... documents In-Reply-To: <199902231318.OAA01371@pcronny.axis.se> from Ronny Ranerup at "Feb 23, 99 02:18:45 pm" To: ronny.ranerup@axis.com (Ronny Ranerup) Date: Tue, 23 Feb 1999 12:17:15 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > aoga@imedia.com: > >http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > > Ahh, excellent, now if this could be included in the distribution or > at least a link to it in the man pages of the distribution we would > probably make someones life a little easier. > > I haven't checked all of the HOWTO, but methinks there's no info on > how it ties into the kernel and some on the inner workings of the > beast... (Not that it necessarily should be in the HOWTO.) And yes > there's always TFS, but if I would check the source for every program > I wanted to figure out, I'd do nothing else. Linus', and others, > attitude of not documenting anything either inside our outside the > code I find kind of childish, 'This is mine, mine, mine, and if you > don't understand it don't mess with it...' My company (were I work, > that is, not own.) is porting Linux to our new CPU and we would enjoy > it even more if the comments in it weren't from the beginning of the > 90s... and incorrect... and outdated... > Quite frankly, if I were to write a kernel tutorial on every kernel subsystem I hacked on, I wouldn't be able to do it. autofs wouldn't exist. It's interesting that you're porting Linux to a new CPU (Axis... that would be a new SPARC?) and a good indication how far we've come, but you have to realize that how far we've come is in many ways an indication of how we got there! One difference with Linux, too, is that there are lots of people to ask. -hpa From owner-autofs@linux.kernel.org Tue Feb 23 13:34:16 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA27573 for ; Tue, 23 Feb 1999 13:34:16 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id MAA21689 for autofs-list; Tue, 23 Feb 1999 12:59:55 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id MAA21684 for ; Tue, 23 Feb 1999 12:59:52 -0800 Received: from animal.corp.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id MAA08426 for ; Tue, 23 Feb 1999 12:59:20 -0800 X-BlackMail: robin.frogtown.com, animal.corp.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 12:59:20(PST) on February 23, 1999 Date: Tue, 23 Feb 1999 12:59:25 -0800 (PST) From: Michael Newlyn Blake X-Sender: mblake@animal.corp.frogtown.com To: The autofs list Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902232010.MAA22325@cesium.transmeta.com> Message-ID: X-Url: http://www.frogtown.com/~mblake MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 23 Feb 1999, H. Peter Anvin wrote: > What I might try to do, is when parsing auto.master to replace all > direct mounts with a link into /.a or something, and make /.a an > indirect mount. We do something similar here, but we have a network of about 6 machines. Not that hard to maintain. In a company with say... tens of thousand machines, this is just unreasonable. The question is, how commited are you to making Linux easy to use out of the box? I understand that this isn't easy to impliment, and it's your baby. Right and wrong are highly subjective terms, but you get to decide what's right for autofs. I do, however, ask you to consider the importance of making Linux more usable out of the box. Sun (I don't mean to keep using Sun as an example, but I'm most familiar with it... and I work there...) put in the effort to make these features in autofs available for these reasons, and it has paid off remarkably well. In a properly configured environment, a machine can be brought up on the net and have immediata access to all of the shared NFS resources it needs without intervention by the system administrator. If this is to be a Unix platform for corporate end users as well as geeks and other unix proficient folks, ease of use to needs to be taken into consideration as well as difficulty of implimentation. Now I'm not saying drop everything and do it now... but please don't rule this feature out. It's needed... Thanks. ----- Hark ye, Clinker, you are a most notorious offender. You stand convicted of sickness, hunger, wretchedness, and want. -- Tobias Smollet From owner-autofs@linux.kernel.org Tue Feb 23 13:48:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA27579 for ; Tue, 23 Feb 1999 13:48:16 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA22795 for autofs-list; Tue, 23 Feb 1999 13:14:50 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA22792 for ; Tue, 23 Feb 1999 13:14:49 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id NAA28985; Tue, 23 Feb 1999 13:13:55 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id NAA27135; Tue, 23 Feb 1999 13:13:54 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id NAA15845; Tue, 23 Feb 1999 13:13:54 -0800 Message-Id: <199902232113.NAA15845@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Michael Newlyn Blake at "Feb 23, 99 12:59:25 pm" To: mblake@Frogtown.com (Michael Newlyn Blake) Date: Tue, 23 Feb 1999 13:13:54 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On Tue, 23 Feb 1999, H. Peter Anvin wrote: > > What I might try to do, is when parsing auto.master to replace all > > direct mounts with a link into /.a or something, and make /.a an > > indirect mount. > > We do something similar here, but we have a network of about 6 machines. > Not that hard to maintain. In a company with say... tens of thousand > machines, this is just unreasonable. > > The question is, how commited are you to making Linux easy to use out of > the box? I understand that this isn't easy to impliment, and it's your > baby. Right and wrong are highly subjective terms, but you get to decide > what's right for autofs. > > I do, however, ask you to consider the importance of making Linux more > usable out of the box. Sun (I don't mean to keep using Sun as an > example, but I'm most familiar with it... and I work there...) put in the > effort to make these features in autofs available for these reasons, and > it has paid off remarkably well. In a properly configured environment, a > machine can be brought up on the net and have immediata access to all of > the shared NFS resources it needs without intervention by the system > administrator. > > If this is to be a Unix platform for corporate end users as well as > geeks and other unix proficient folks, ease of use to needs to be taken > into consideration as well as difficulty of implimentation. Now I'm not > saying drop everything and do it now... but please don't rule this feature > out. It's needed... > IT WON'T HAPPEN. STOP BITCHING ABOUT IT. I said I'd do the transformation above automatically -- that can easily be done -- but true direct mounts WILL NOT HAPPEN. STOP BUGGING ME ABOUT IT *NOW*!!! -hpa From owner-autofs@linux.kernel.org Tue Feb 23 15:46:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA27688 for ; Tue, 23 Feb 1999 15:46:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA30992 for autofs-list; Tue, 23 Feb 1999 15:10:46 -0800 Received: from fireant.ma.utexas.edu (fireant.ma.utexas.edu [128.83.133.215]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA30989 for ; Tue, 23 Feb 1999 15:10:44 -0800 Received: from leonhard.ma.utexas.edu (leonhard.ma.utexas.edu [128.83.133.242]) by fireant.ma.utexas.edu (8.9.1a/8.9.1) with ESMTP id RAA28395; Tue, 23 Feb 1999 17:10:42 -0600 (CST) Date: Tue, 23 Feb 1999 17:10:43 -0600 (CST) From: Patrick Goetz X-Sender: pgoetz@leonhard.ma.utexas.edu To: "H. Peter Anvin" cc: autofs@linux.kernel.org Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902232113.NAA15845@cesium.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 23 Feb 1999, H. Peter Anvin wrote: > > > > IT WON'T HAPPEN. STOP BITCHING ABOUT IT. > > I said I'd do the transformation above automatically -- that can > easily be done -- but true direct mounts WILL NOT HAPPEN. STOP > BUGGING ME ABOUT IT *NOW*!!! > Dang, no need to shout - you did ask for comments, if I'm not mistaken... From owner-autofs@linux.kernel.org Tue Feb 23 15:47:19 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA27692 for ; Tue, 23 Feb 1999 15:47:18 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA31211 for autofs-list; Tue, 23 Feb 1999 15:13:33 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA31204 for ; Tue, 23 Feb 1999 15:13:32 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA31003; Tue, 23 Feb 1999 15:12:59 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA12771; Tue, 23 Feb 1999 15:12:59 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA17998; Tue, 23 Feb 1999 15:12:59 -0800 Message-Id: <199902232312.PAA17998@cesium.transmeta.com> Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Patrick Goetz at "Feb 23, 99 05:10:43 pm" To: pgoetz@fireant.ma.utexas.edu (Patrick Goetz) Date: Tue, 23 Feb 1999 15:12:59 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On Tue, 23 Feb 1999, H. Peter Anvin wrote: > > > > > > > IT WON'T HAPPEN. STOP BITCHING ABOUT IT. > > > > I said I'd do the transformation above automatically -- that can > > easily be done -- but true direct mounts WILL NOT HAPPEN. STOP > > BUGGING ME ABOUT IT *NOW*!!! > > > > Dang, no need to shout - you did ask for comments, if I'm not mistaken... > I did, but I also said about this specific feature "it won't happen, so don't ask." I *ALSO* explained it wasn't my choice, which some people seemed to choose to ignore. -hpa From owner-autofs@linux.kernel.org Tue Feb 23 16:08:51 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27715 for ; Tue, 23 Feb 1999 16:08:51 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA00238 for autofs-list; Tue, 23 Feb 1999 15:35:52 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA00232 for ; Tue, 23 Feb 1999 15:35:49 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA10468; Tue, 23 Feb 1999 15:35:30 -0800 From: Alvin Oga Message-Id: <199902232335.PAA10468@leopard.imedia.com> Subject: Re: autofs v4: speak now - several hundred linux boxes To: lori@Frogtown.com (Lori) Date: Tue, 23 Feb 1999 15:35:30 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Lori" at Feb 23, 99 07:20:00 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi lori > On Tue, 23 Feb 1999, Alvin Oga wrote: > > > client# cat /etc/auto.master > > /.autofs /etc/auto.data --timeout 60 > > > > client# cat /etc/auto.data > > # > > Data -fstype=nfs machine:/usr1/data > > > > client# cd / ; ln -s .autofs/Data Data > > the problem with this method is the clients all have to be modified to > utilize this scheme. it's not an out of the box solution. This may not > seem to be a problem to some people, but it is if you get a couple hundred > boxes or a bunch of engineers who constantly reinstall their systems. The I think that the above method works even if we have 1,000 linux boxes... each linux box will have to have the binaries installed at some time... during that time.... all the necessary changes can be performed... - - root-boot mode or single user or root-rescue mode etc..whatever that get - you access ot the NFS network or to a cdrom or flash disk or ??? - once your empty virgin linux disk/machine can find the "Master copy" you can now run "install.pl" which will do all the magic of copying /etc/passwd, /etc/auto.master, /etc/rc.d/* etc etc etc... and do a few more symlinks for /usr/local and /home and /var/spool/mail etc - all done with an install script....no problem... but I think we are talking about the same thing as your /import/dist and /usr/dist stuff... just need to add one more "install script program" ?? have fun alvin > direct mount map can be put in the NIS master auto.master file indicate > the other maps and no additional work need be done on the client side. I'm > not necessarily advocating doing less work, just that it makes life less > complex for those you support if you can just hand them a box or have them > reinstall their own system without a bunch of rules and regs before it'll > run. > > When I last set up a bunch of NIS/autofs systems on the network I wanted > the clients to be as generic as possible, this was a lab environment and > it was important that they be easily reinstalled without a bunch of local > hacks - I had to set up two scenarios depending on what OS you were > running - solaris or linux. for solaris I set up /usr/dist, and for linux > I had to set up /import/dist (among other mounts related to function). > This was a workable solution in our environment because there wasn't > anything before I set it up, and the solaris systems could use the /import > scenario, also. But it was still two different set-ups that were operating > system dependant. and therefore more confusion for the end-user. > > > cheers! > From owner-autofs@linux.kernel.org Tue Feb 23 16:16:08 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27744 for ; Tue, 23 Feb 1999 16:16:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA01005 for autofs-list; Tue, 23 Feb 1999 15:45:05 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA00991 for ; Tue, 23 Feb 1999 15:45:02 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA10523; Tue, 23 Feb 1999 15:44:54 -0800 From: Alvin Oga Message-Id: <199902232344.PAA10523@leopard.imedia.com> Subject: Re: autofs v4: speak now or shutdown or not To: hpa@transmeta.com (H. Peter Anvin) Date: Tue, 23 Feb 1999 15:44:54 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902232003.MAA22263@cesium.transmeta.com> from "H. Peter Anvin" at Feb 23, 99 12:03:21 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya peter... yes... if you could...please use some kind of detection schem where autofs will refuse to shutdown and just flag that /home is in use.... instead of going down anyway...and not be able to restart to remount the other dirs that the users now wants but can't get since autofs won't come back up without killing all the old processes... thank you alvin > > I totally agree. A daemon is a service program which should be able > > to be started and stopped like any other service program by the > > system administator. Especially under Linux, which is one of most > > stable operation systems I worked with, it would help to design > > daemons in a coherent service friendly manner. > > > > > > Maybe it is just me, but it always hurts me, if I have to reboot a machines, > > just because I scewed it up so much that no clean simple kills will help any > > more. > > Note you can certainly stop autofs, but you might have to clean up > manually before you can restart. The only other option would be for > autofs to refuse to shut down if any of its filesystems are in use, > like it does if it gets SIGUSR1 (I think...) > > -hpa > From owner-autofs@linux.kernel.org Tue Feb 23 16:17:36 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27748 for ; Tue, 23 Feb 1999 16:17:36 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA01382 for autofs-list; Tue, 23 Feb 1999 15:49:56 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA01379 for ; Tue, 23 Feb 1999 15:49:55 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA31651; Tue, 23 Feb 1999 15:49:39 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA16925; Tue, 23 Feb 1999 15:49:38 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA18309; Tue, 23 Feb 1999 15:49:38 -0800 Message-Id: <199902232349.PAA18309@cesium.transmeta.com> Subject: Re: autofs v4: speak now or shutdown or not In-Reply-To: <199902232344.PAA10523@leopard.imedia.com> from Alvin Oga at "Feb 23, 99 03:44:54 pm" To: aoga@imedia.com (Alvin Oga) Date: Tue, 23 Feb 1999 15:49:38 -0800 (PST) Cc: hpa@transmeta.com, aoga@leopard.imedia.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > hi ya peter... > > yes... if you could...please use some kind of detection schem > where autofs will refuse to shutdown and just flag that /home > is in use.... instead of going down anyway...and not be able > to restart to remount the other dirs that the users now wants > but can't get since autofs won't come back up without killing > all the old processes... > Sure; currently that's what autofs uses if you send it SIGUSR1; would it make sense if I made it behave that way for SIGTERM as well? -hpa From owner-autofs@linux.kernel.org Tue Feb 23 16:18:00 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27752 for ; Tue, 23 Feb 1999 16:17:59 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA00671 for autofs-list; Tue, 23 Feb 1999 15:41:54 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA00667 for ; Tue, 23 Feb 1999 15:41:52 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA10502; Tue, 23 Feb 1999 15:41:47 -0800 From: Alvin Oga Message-Id: <199902232341.PAA10502@leopard.imedia.com> Subject: Re: autofs v4: speak now - common /mnt or /Net To: frithjof@fkp.tu-darmstadt.de Date: Tue, 23 Feb 1999 15:41:47 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Frithjof Anders" at Feb 23, 99 05:23:05 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi michael > > One thing that amd has that would be nice if autofs has is symlink mounts. > > If for instance, you have two systems, fred and barney. Fred exports > > /pebbles > > and /dino, while barney exports /bambam, you might want to have a centralized > > mount point for all disks: > > > > /mnt/pebbles > > /mnt/bambam > > /mnt/dino > > > > If you were on fred, /mnt/pebbles would be a symlink to /pebbles, /mnt/dino > > That work already :) yes... I use it.... to do /Net/All_hosts aka /n/All_hosts etc... cd /Net ; ln -s /.autofs/RemoteHost_or_Dir1 . cd /Net ; ln -s /.autofs/RemoteHost_or_Dir2 . cd /Net ; ln -s / hostname I use it for local and remote backups .... since local disk backup does not help if the cpu/system goes nuts and destroys all local disks have fun alvin From owner-autofs@linux.kernel.org Tue Feb 23 16:34:20 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27760 for ; Tue, 23 Feb 1999 16:34:20 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA01982 for autofs-list; Tue, 23 Feb 1999 15:58:10 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA01967 for ; Tue, 23 Feb 1999 15:58:07 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA10596; Tue, 23 Feb 1999 15:58:05 -0800 From: Alvin Oga Message-Id: <199902232358.PAA10596@leopard.imedia.com> Subject: Re: autofs v4: speak now or shutdown or not To: hpa@transmeta.com (H. Peter Anvin) Date: Tue, 23 Feb 1999 15:58:05 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902232349.PAA18309@cesium.transmeta.com> from "H. Peter Anvin" at Feb 23, 99 03:49:38 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya peter > > yes... if you could...please use some kind of detection schem > > where autofs will refuse to shutdown and just flag that /home > > is in use.... instead of going down anyway...and not be able > > to restart to remount the other dirs that the users now wants > > but can't get since autofs won't come back up without killing > > all the old processes... > > > > Sure; currently that's what autofs uses if you send it SIGUSR1; would > it make sense if I made it behave that way for SIGTERM as well? Am not 100% sure...but doesn't SIGTERM kill the process ?? All I'm wondering is autofs flags that it cannot umount a used mount point and merrily continues it's way out....and cannot get autofs restarted... - so am, wondering, can we give it an option to force it to continue and shutdown like it does now, knowing that it ( autofs) won't come back up .. the other option is to NOT shutdown if it finds "used mountpoints" - I have this problem with /usr/local and don't really want to shutdown the machine... and similarly if /var/spool/mail was unmounted and autofs shutdowns when no one was reading mail... and we're stuck with whatever used mountpoints we have... no /var/spool/mail... which is not critical .. just annoying.. - if I play with /home....than all users are kicked off anyway and the machine rebooted to test the changes... thanx alvin From owner-autofs@linux.kernel.org Tue Feb 23 16:41:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27764 for ; Tue, 23 Feb 1999 16:41:27 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA02663 for autofs-list; Tue, 23 Feb 1999 16:04:38 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA02660 for ; Tue, 23 Feb 1999 16:04:36 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id QAA31910; Tue, 23 Feb 1999 16:04:35 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id QAA18982; Tue, 23 Feb 1999 16:04:35 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id QAA18454; Tue, 23 Feb 1999 16:04:34 -0800 Message-Id: <199902240004.QAA18454@cesium.transmeta.com> Subject: Re: autofs v4: speak now or shutdown or not In-Reply-To: <199902232358.PAA10596@leopard.imedia.com> from Alvin Oga at "Feb 23, 99 03:58:05 pm" To: aoga@imedia.com (Alvin Oga) Date: Tue, 23 Feb 1999 16:04:34 -0800 (PST) Cc: hpa@transmeta.com, aoga@leopard.imedia.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > yes... if you could...please use some kind of detection schem > > > where autofs will refuse to shutdown and just flag that /home > > > is in use.... instead of going down anyway...and not be able > > > to restart to remount the other dirs that the users now wants > > > but can't get since autofs won't come back up without killing > > > all the old processes... > > > > > > > Sure; currently that's what autofs uses if you send it SIGUSR1; would > > it make sense if I made it behave that way for SIGTERM as well? > > Am not 100% sure...but doesn't SIGTERM kill the process ?? SIGTERM is the "normal" kill (not kill -9, which is SIGKILL). Software can, and does, trap SIGTERM for a clean shutdown. In the case of autofs, it umounts what it cans and then puts autofs into catatonic mode. > All I'm wondering is autofs flags that it cannot umount a used mount point > and merrily continues it's way out....and cannot get autofs restarted... > > - so am, wondering, can we give it an option to force it to continue > and shutdown like it does now, knowing that it ( autofs) won't come back up > .. the other option is to NOT shutdown if it finds "used mountpoints" You already have this option: SIGTERM (and SIGQUIT) does the former, and SIGUSR2 (not SIGUSR1 as I said before) does the latter. Perhaps it would make more sense to have SIGTERM do the latter (I presume SIGTERM is what the rc scripts send) and only have SIGQUIT do the former? > - I have this problem with /usr/local and don't really want to > shutdown the machine... and similarly if /var/spool/mail was > unmounted and autofs shutdowns when no one was reading mail... > and we're stuck with whatever used mountpoints we have... > no /var/spool/mail... which is not critical .. just annoying.. > > - if I play with /home....than all users are kicked off anyway > and the machine rebooted to test the changes... -hpa From owner-autofs@linux.kernel.org Tue Feb 23 16:50:50 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA27770 for ; Tue, 23 Feb 1999 16:50:50 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA03645 for autofs-list; Tue, 23 Feb 1999 16:20:45 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA03640 for ; Tue, 23 Feb 1999 16:20:44 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id QAA10697; Tue, 23 Feb 1999 16:20:42 -0800 From: Alvin Oga Message-Id: <199902240020.QAA10697@leopard.imedia.com> Subject: Re: autofs v4: speak now or shutdown or not To: hpa@transmeta.com (H. Peter Anvin) Date: Tue, 23 Feb 1999 16:20:42 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902240004.QAA18454@cesium.transmeta.com> from "H. Peter Anvin" at Feb 23, 99 04:04:34 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya peter okay... I see it now... I just checked /etc/rc.d/init.d/autofs ( yup redhat box ) ... stop) # # change -TERM to -USR2 here ? # --->> kill -TERM $(/sbin/pidof /usr/sbin/automount) rm -f /var/lock/subsys/automount ;; reload|restart) .... while read pid tt stat time command; do echo "$command" >>$TMP2 if ! grep -q "^$command" $TMP2; then --->> kill -USR2 $pid echo "Stop $command" fi done ) .... # # I'd add this to it # force_kill) echo "AutoFS Going down...and might NOT come back up..." sleep 10 ; --->> kill -TERM $(/sbin/pidof /usr/sbin/automount) rm -f /var/lock/subsys/automount ;; guess we can change those options around ?? so as it is... autofs stop will kill autofs... ( which is what I usually do ) and "autofs reload" will try to reload only the changes to auto.master/auto.misc guess one could experiment with changing -TERM to -USR2 to stop it from shutdown completely and not being able to restart ? thanx alvin > > > > yes... if you could...please use some kind of detection schem > > > > where autofs will refuse to shutdown and just flag that /home > > > > is in use.... instead of going down anyway...and not be able > > > > to restart to remount the other dirs that the users now wants > > > > but can't get since autofs won't come back up without killing > > > > all the old processes... > > > > > > > > > > Sure; currently that's what autofs uses if you send it SIGUSR1; would > > > it make sense if I made it behave that way for SIGTERM as well? > > > > Am not 100% sure...but doesn't SIGTERM kill the process ?? > > SIGTERM is the "normal" kill (not kill -9, which is SIGKILL). > Software can, and does, trap SIGTERM for a clean shutdown. In the > case of autofs, it umounts what it cans and then puts autofs into > catatonic mode. > > > All I'm wondering is autofs flags that it cannot umount a used mount point > > and merrily continues it's way out....and cannot get autofs restarted... > > > > - so am, wondering, can we give it an option to force it to continue > > and shutdown like it does now, knowing that it ( autofs) won't come back up > > .. the other option is to NOT shutdown if it finds "used mountpoints" > > You already have this option: SIGTERM (and SIGQUIT) does the former, > and SIGUSR2 (not SIGUSR1 as I said before) does the latter. > > Perhaps it would make more sense to have SIGTERM do the latter (I > presume SIGTERM is what the rc scripts send) and only have SIGQUIT do > the former? > > > - I have this problem with /usr/local and don't really want to > > shutdown the machine... and similarly if /var/spool/mail was > > unmounted and autofs shutdowns when no one was reading mail... > > and we're stuck with whatever used mountpoints we have... > > no /var/spool/mail... which is not critical .. just annoying.. > > > > - if I play with /home....than all users are kicked off anyway > > and the machine rebooted to test the changes... > > -hpa > From owner-autofs@linux.kernel.org Tue Feb 23 17:10:58 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA27786 for ; Tue, 23 Feb 1999 17:10:58 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA04880 for autofs-list; Tue, 23 Feb 1999 16:39:04 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA04872 for ; Tue, 23 Feb 1999 16:39:02 -0800 Received: from robin.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id QAA11802 for ; Tue, 23 Feb 1999 16:38:48 -0800 X-BlackMail: robin.frogtown.com, robin.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 16:38:48(PST) on February 23, 1999 Date: Tue, 23 Feb 1999 16:38:56 -0800 (PST) From: Lori X-Sender: lori@scooter.corp.frogtown.com cc: autofs mailing list Subject: Re: autofs v4: speak now .... documents In-Reply-To: <199902232017.MAA22378@cesium.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: maybe in addition to this legion of kernel hackers and the like there should be an associated legion of doc writers? so that you don't get engineers writing intelligible documentation, free them up for more kernel hacking - or is that what the linux documentation project is supposed to do? just stirring the flames... On Tue, 23 Feb 1999, H. Peter Anvin wrote: > > > > aoga@imedia.com: > > >http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > > > > Ahh, excellent, now if this could be included in the distribution or > > at least a link to it in the man pages of the distribution we would > > probably make someones life a little easier. > > > > I haven't checked all of the HOWTO, but methinks there's no info on > > how it ties into the kernel and some on the inner workings of the > > beast... (Not that it necessarily should be in the HOWTO.) And yes > > there's always TFS, but if I would check the source for every program > > I wanted to figure out, I'd do nothing else. Linus', and others, > > attitude of not documenting anything either inside our outside the > > code I find kind of childish, 'This is mine, mine, mine, and if you > > don't understand it don't mess with it...' My company (were I work, > > that is, not own.) is porting Linux to our new CPU and we would enjoy > > it even more if the comments in it weren't from the beginning of the > > 90s... and incorrect... and outdated... > > > > Quite frankly, if I were to write a kernel tutorial on every kernel > subsystem I hacked on, I wouldn't be able to do it. autofs wouldn't > exist. It's interesting that you're porting Linux to a new CPU > (Axis... that would be a new SPARC?) and a good indication how far > we've come, but you have to realize that how far we've come is in many > ways an indication of how we got there! > > One difference with Linux, too, is that there are lots of people to > ask. > > -hpa > From owner-autofs@linux.kernel.org Wed Feb 24 01:39:11 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA28450 for ; Wed, 24 Feb 1999 01:39:11 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA00701 for autofs-list; Wed, 24 Feb 1999 00:54:39 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA00698 for ; Wed, 24 Feb 1999 00:54:35 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id JAA11579; Wed, 24 Feb 1999 09:54:38 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id JAA00414; Wed, 24 Feb 1999 09:54:38 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902232341.PAA10502@leopard.imedia.com> Date: Wed, 24 Feb 1999 09:54:38 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Alvin Oga Subject: Re: autofs v4: speak now - common /mnt or /Net Cc: (Alvin Oga) , autofs@linux.kernel.org Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Hi Alvin, Thus spoke Alvin Oga on 23-Feb-99 : > > hi michael > >> > One thing that amd has that would be nice if autofs has is symlink mounts. >> > If for instance, you have two systems, fred and barney. Fred exports >> > /pebbles >> > and /dino, while barney exports /bambam, you might want to have a >> > centralized >> > mount point for all disks: >> > >> > /mnt/pebbles >> > /mnt/bambam >> > /mnt/dino >> > >> > If you were on fred, /mnt/pebbles would be a symlink to /pebbles, >> > /mnt/dino >> >> That work already :) > > yes... I use it.... to do /Net/All_hosts aka /n/All_hosts etc... > > cd /Net ; ln -s /.autofs/RemoteHost_or_Dir1 . > cd /Net ; ln -s /.autofs/RemoteHost_or_Dir2 . > cd /Net ; ln -s / hostname > > I use it for local and remote backups .... since local disk backup > does not help if the cpu/system goes nuts and destroys all local disks But there is a problem with that concept: you have make your symbolic links by hand to your /.autofs. Thats, what we do not want, we meaning people who run thousends of cooperative machines or people like me, who do physics for living and are not paid as sys adm. I would like to build my cluster with simple means almost "out of the box" as another poster demanded. I am flexible. I am willing to organize the filesystem structure accouring to to tools, we have. But at the very end we must not forget: Computer are just tools not their own purpose. have fun, too Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 24 01:46:49 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA28456 for ; Wed, 24 Feb 1999 01:46:49 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA01765 for autofs-list; Wed, 24 Feb 1999 01:10:09 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA01760 for ; Wed, 24 Feb 1999 01:10:04 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id KAA11674; Wed, 24 Feb 1999 10:10:05 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id KAA00436; Wed, 24 Feb 1999 10:10:05 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199812100951.BAA01000@cesium.transmeta.com> Date: Wed, 24 Feb 1999 10:10:05 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: shutdown autofs Cc: (autofs mailing list) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Peter, I would like to come back to the shutdown of the automounter question. I remembered what the real problem in my case always was and still is. It is not the umount of mounted directories but removal of symbolic links when the disk is local. How do I remove a symbolic link by hand from an autofs managed moint point? rm does not work. How do I tell the automounter to unlink symbolic links locally? Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 24 01:59:35 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA28464 for ; Wed, 24 Feb 1999 01:59:35 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA02037 for autofs-list; Wed, 24 Feb 1999 01:14:22 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA02034 for ; Wed, 24 Feb 1999 01:14:21 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id BAA05454; Wed, 24 Feb 1999 01:14:03 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id BAA24577; Wed, 24 Feb 1999 01:14:02 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id BAA30267; Wed, 24 Feb 1999 01:14:01 -0800 Message-Id: <199902240914.BAA30267@cesium.transmeta.com> Subject: Re: shutdown autofs In-Reply-To: from Frithjof Anders at "Feb 24, 99 10:10:05 am" To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 01:14:01 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Peter, > > I would like to come back to the shutdown of the automounter question. > > I remembered what the real problem in my case always was and still > is. It is not the umount of mounted directories but removal of > symbolic links when the disk is local. > > How do I remove a symbolic link by hand from an autofs managed moint point? > rm does not work. How do I tell the automounter to unlink symbolic links > locally? > You're right, it doesn't work, and it probably should. Let me see what I can do about it. -hpa From owner-autofs@linux.kernel.org Wed Feb 24 02:03:23 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA28470 for ; Wed, 24 Feb 1999 02:03:23 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA01825 for autofs-list; Wed, 24 Feb 1999 01:11:00 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA01822 for ; Wed, 24 Feb 1999 01:10:59 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id BAA23059; Wed, 24 Feb 1999 01:10:56 -0800 From: Alvin Oga Message-Id: <199902240910.BAA23059@leopard.imedia.com> Subject: Re: autofs v4: speak now - common /mnt or /Net To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 01:10:56 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Frithjof Anders" at Feb 24, 99 09:54:38 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi firthjof > >> > One thing that amd has that would be nice if autofs has is symlink mounts. > >> > If for instance, you have two systems, fred and barney. Fred exports > >> > /pebbles > >> > and /dino, while barney exports /bambam, you might want to have a > >> > centralized > >> > mount point for all disks: > >> > > >> > /mnt/pebbles > >> > /mnt/bambam > >> > /mnt/dino > >> > > >> > If you were on fred, /mnt/pebbles would be a symlink to /pebbles, > >> > /mnt/dino > >> > >> That work already :) > > > > yes... I use it.... to do /Net/All_hosts aka /n/All_hosts etc... > > > > cd /Net ; ln -s /.autofs/RemoteHost_or_Dir1 . > > cd /Net ; ln -s /.autofs/RemoteHost_or_Dir2 . > > cd /Net ; ln -s / hostname > > > But there is a problem with that concept: you have make your symbolic links by > hand to your /.autofs. yes... okay... I suppose we can make a simple script that creates the proper /etc/auto.misc and /etc/auto.master and create the above symlinks... I can get the list of symlinks from what files ??? > Thats, what we do not want, we meaning people who run thousends of > cooperative machines or people like me, who do physics for living and are not > paid as sys adm. I would like to build my cluster with simple means almost "out > of the box" as another poster demanded. yup...to me, it means just run a bunch of setup scripts.. have fun alvin From owner-autofs@linux.kernel.org Wed Feb 24 02:20:28 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA28499 for ; Wed, 24 Feb 1999 02:20:27 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA02885 for autofs-list; Wed, 24 Feb 1999 01:31:31 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA02882 for ; Wed, 24 Feb 1999 01:31:27 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id KAA11732; Wed, 24 Feb 1999 10:31:33 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id KAA00478; Wed, 24 Feb 1999 10:31:32 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902240914.BAA30267@cesium.transmeta.com> Date: Wed, 24 Feb 1999 10:31:32 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: shutdown autofs Cc: (autofs mailing list) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 24-Feb-99 : >> Peter, >> >> I would like to come back to the shutdown of the automounter question. >> >> I remembered what the real problem in my case always was and still >> is. It is not the umount of mounted directories but removal of >> symbolic links when the disk is local. >> >> How do I remove a symbolic link by hand from an autofs managed moint point? >> rm does not work. How do I tell the automounter to unlink symbolic links >> locally? >> > > You're right, it doesn't work, and it probably should. Let me see > what I can do about it. That would be really great! Making that work would have an impact how you can setup your autofs in large cluster. There was a discussion about the lack of sublinks of autofs which lets to may mounts of the same filesystem, one mount for each user. People had to recompile the linux kernel with more allowed mounts. Not a good solution IMOP. My orginal solution was, to make /home a autofs which manages only symbolic links to an /import automounter mount point. It worked find: each filesystem was mounted only once. But since the symlink were not removable, I was screw once a user has to be moved, or I did a typo at installing a new user. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 24 02:36:03 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA28506 for ; Wed, 24 Feb 1999 02:36:02 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA03167 for autofs-list; Wed, 24 Feb 1999 01:36:46 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA03162 for ; Wed, 24 Feb 1999 01:36:44 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id BAA05644; Wed, 24 Feb 1999 01:36:32 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id BAA25180; Wed, 24 Feb 1999 01:36:31 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id BAA30439; Wed, 24 Feb 1999 01:36:25 -0800 Message-Id: <199902240936.BAA30439@cesium.transmeta.com> Subject: Re: shutdown autofs In-Reply-To: from Frithjof Anders at "Feb 24, 99 10:31:32 am" To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 01:36:25 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > You're right, it doesn't work, and it probably should. Let me see > > what I can do about it. > > That would be really great! Making that work would have an impact how you can > setup your autofs in large cluster. There was a discussion about the lack of > sublinks of autofs which lets to may mounts of the same filesystem, one mount > for each user. People had to recompile the linux kernel with more allowed > mounts. Not a good solution IMOP. > > My orginal solution was, to make /home a autofs which manages only symbolic > links to an /import automounter mount point. It worked find: each filesystem > was mounted only once. But since the symlink were not removable, I was screw > once a user has to be moved, or I did a typo at installing a new user. This I'd rather solve by Solaris-style pseudomounts, again due to the namespace issue (it may not be an issue for you, but it is a cause of frequent complaints.) Solaris allows you to specify a root for the kernel to mount, and the kernel will internally maintain an NFS mount for multiple users, exposed only at a higher level. Probably doesn't solve the #superblocks issue, though, but that's better handled by allocating superblocks dynamically anyway. I tend to believe klugy solutions just make the problem worse by reducing the pressure to fix the real problem... -hpa From owner-autofs@linux.kernel.org Wed Feb 24 02:38:51 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA28510 for ; Wed, 24 Feb 1999 02:38:51 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA03275 for autofs-list; Wed, 24 Feb 1999 01:38:41 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA03272 for ; Wed, 24 Feb 1999 01:38:40 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id BAA05656; Wed, 24 Feb 1999 01:38:29 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id BAA25209; Wed, 24 Feb 1999 01:38:28 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id BAA30460; Wed, 24 Feb 1999 01:38:28 -0800 Message-Id: <199902240938.BAA30460@cesium.transmeta.com> Subject: Re: shutdown autofs In-Reply-To: from Frithjof Anders at "Feb 24, 99 10:31:32 am" To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 01:38:28 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > Thus spoke H. Peter Anvin on 24-Feb-99 : > >> Peter, > >> > >> I would like to come back to the shutdown of the automounter question. > >> > >> I remembered what the real problem in my case always was and still > >> is. It is not the umount of mounted directories but removal of > >> symbolic links when the disk is local. > >> > >> How do I remove a symbolic link by hand from an autofs managed moint point? > >> rm does not work. How do I tell the automounter to unlink symbolic links > >> locally? > >> > > You're right, it doesn't work, and it probably should. Let me see > > what I can do about it. > > That would be really great! Making that work would have an impact how you can > setup your autofs in large cluster. There was a discussion about the lack of > sublinks of autofs which lets to may mounts of the same filesystem, one mount > for each user. People had to recompile the linux kernel with more allowed > mounts. Not a good solution IMOP. > > My orginal solution was, to make /home a autofs which manages only symbolic > links to an /import automounter mount point. It worked find: each filesystem > was mounted only once. But since the symlink were not removable, I was screw > once a user has to be moved, or I did a typo at installing a new user. > Oh by the way... in the meantime you can send SIGUSR1 to automount. SIGUSR1 means "umount all you can but don't exit"; since symlinks are always removable they will be removed by the daemon. -hpa From owner-autofs@linux.kernel.org Wed Feb 24 03:23:13 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA28547 for ; Wed, 24 Feb 1999 03:23:13 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA06565 for autofs-list; Wed, 24 Feb 1999 02:34:34 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA06553 for ; Wed, 24 Feb 1999 02:34:22 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id LAA12049; Wed, 24 Feb 1999 11:34:11 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id LAA00941; Wed, 24 Feb 1999 11:34:10 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902240936.BAA30439@cesium.transmeta.com> Date: Wed, 24 Feb 1999 11:34:10 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: shutdown autofs Cc: autofs@linux.kernel.org Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 24-Feb-99 : >> >> My orginal solution was, to make /home a autofs which manages only symbolic >> links to an /import automounter mount point. It worked find: each filesystem >> was mounted only once. But since the symlink were not removable, I was screw >> once a user has to be moved, or I did a typo at installing a new user. > > This I'd rather solve by Solaris-style pseudomounts, again due to the > namespace issue (it may not be an issue for you, but it is a cause of > frequent complaints.) Solaris allows you to specify a root for the > kernel to mount, and the kernel will internally maintain an NFS mount > for multiple users, exposed only at a higher level. Probably doesn't > solve the #superblocks issue, though, but that's better handled by > allocating superblocks dynamically anyway. > > I tend to believe klugy solutions just make the problem worse by > reducing the pressure to fix the real problem... On the other hand the AIX automounter allows you to do exactly what I want via shared mounts Sharing Mounts The hostname is the server name where the file system will mount, pathname is the directory path to mount, and subdir, when supplied, is where the symbolic link is made to a sub-directory. This can be used to prevent duplicate mounts when multiple directories in the same remote file system can be accessed. For example, consider the following map for /home: able prime:/home/prime:able baker prime:/home/prime:baker When a user attempts to access a file in /home/able, the automount command mounts prime:/home/prime, but creates a symbolic link called /home/able to the able sub-directory in the temporarily mounted file system. If a user immediately tries to access a file in /home/baker, the automount command creates a symbolic link that points to the baker sub-directory. The /home/prime directory would already be mounted. Why do you think, putting extra code in the kernel is better than allowing a solution on deamon level? By no means I want to trigger a discussion of a monolitic kernel vesus the mach approach. The monolitic won, but learned how to handle modules ;-) And I tend to believe, that, as always in live, there is not only one unique solution to a particular problem. Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Wed Feb 24 03:24:13 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id DAA28551 for ; Wed, 24 Feb 1999 03:24:13 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id CAA05284 for autofs-list; Wed, 24 Feb 1999 02:12:32 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id CAA05281 for ; Wed, 24 Feb 1999 02:12:30 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id CAA05970 for ; Wed, 24 Feb 1999 02:12:29 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id CAA25900 for ; Wed, 24 Feb 1999 02:12:29 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id CAA04277 for autofs@linux.kernel.org; Wed, 24 Feb 1999 02:12:28 -0800 Message-Id: <199902241012.CAA04277@cesium.transmeta.com> Subject: Patch to allow root to rm symlinks To: autofs@linux.kernel.org (autofs mailing list) Date: Wed, 24 Feb 1999 02:12:28 -0800 (PST) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: This patch against 2.2.2 should allow root to rm autofs symlinks. Please test it and let me know if it causes problems; since this is an enhancement rather than a bugfix I need to make a pretty strong case to let it past Linus at this time. It also changes some of the error codes to be more posixly correct. It turns out "permission denied" is EACCES, not EPERM (operation not permitted), which is what I had used in most places. -hpa diff -ur stock/linux-2.2.2/fs/autofs/root.c linux-2.2.2/fs/autofs/root.c --- stock/linux-2.2.2/fs/autofs/root.c Wed Feb 24 02:07:11 1999 +++ linux-2.2.2/fs/autofs/root.c Wed Feb 24 01:54:35 1999 @@ -281,7 +281,7 @@ autofs_say(dentry->d_name.name,dentry->d_name.len); if ( !autofs_oz_mode(sbi) ) - return -EPERM; + return -EACCES; if ( dentry->d_name.len > NAME_MAX ) return -ENAMETOOLONG; @@ -349,17 +349,20 @@ struct autofs_dir_ent *ent; unsigned int n; - if ( !autofs_oz_mode(sbi) ) - return -EPERM; + /* This allows root to remove symlinks */ + if ( !autofs_oz_mode(sbi) && !capable(CAP_SYS_ADMIN) ) + return -EACCES; ent = autofs_hash_lookup(dh, &dentry->d_name); if ( !ent ) return -ENOENT; n = ent->ino - AUTOFS_FIRST_SYMLINK; - if ( n >= AUTOFS_MAX_SYMLINKS || !test_bit(n,sbi->symlink_bitmap) ) - return -EINVAL; /* Not a symlink inode, can't unlink */ - + if ( n >= AUTOFS_MAX_SYMLINKS ) + return -EISDIR; /* It's a directory, dummy */ + if ( !test_bit(n,sbi->symlink_bitmap) ) + return -EINVAL; /* Nonexistent symlink? Shouldn't happen */ + dentry->d_time = (unsigned long)(struct autofs_dirhash *)NULL; autofs_hash_delete(ent); clear_bit(n,sbi->symlink_bitmap); @@ -376,7 +379,7 @@ struct autofs_dir_ent *ent; if ( !autofs_oz_mode(sbi) ) - return -EPERM; + return -EACCES; ent = autofs_hash_lookup(dh, &dentry->d_name); if ( !ent ) @@ -405,7 +408,7 @@ ino_t ino; if ( !autofs_oz_mode(sbi) ) - return -EPERM; + return -EACCES; if ( dentry->d_name.len > NAME_MAX ) return -ENAMETOOLONG; From owner-autofs@linux.kernel.org Wed Feb 24 09:08:58 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA28827 for ; Wed, 24 Feb 1999 09:08:58 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id IAA25318 for autofs-list; Wed, 24 Feb 1999 08:03:55 -0800 Received: from fireant.ma.utexas.edu (fireant.ma.utexas.edu [128.83.133.215]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id IAA25315 for ; Wed, 24 Feb 1999 08:03:54 -0800 Received: from leonhard.ma.utexas.edu (leonhard.ma.utexas.edu [128.83.133.242]) by fireant.ma.utexas.edu (8.9.1a/8.9.1) with ESMTP id KAA11847; Wed, 24 Feb 1999 10:03:43 -0600 (CST) Date: Wed, 24 Feb 1999 10:03:45 -0600 (CST) From: Patrick Goetz X-Sender: pgoetz@leonhard.ma.utexas.edu To: Lori cc: autofs mailing list Subject: Re: autofs v4: speak now .... documents In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 23 Feb 1999, Lori wrote: > > > > maybe in addition to this legion of kernel hackers and the like there > should be an associated legion of doc writers? so that you don't get > engineers writing intelligible documentation, free them up for more kernel > hacking - or is that what the linux documentation project is supposed to > do? > > just stirring the flames... > I totally agree with you. Here is the problem, though. Hackers write code because it is fun. Nobody would be slaving over a hot terminal wondering whether to use SIGTERM or SIGQUIT for free if they didn't enjoy doing it. Writing documentation, on the other hand, is not fun, at least not for most people, so it is harder to find people who will commit a couple of hundred (or more) hours of free time to a documentation project. On the other hand, my experience has been that the free documentation which comes with linux is an order of magnitude better than that which comes with Solaris (at least - so far I've found Answerbook to be 100% useless), for example, so I guess that corporations are even more reluctant to PAY someone to write documentation than the average joe schmoe is to write it for free. From owner-autofs@linux.kernel.org Wed Feb 24 09:15:05 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id JAA28853 for ; Wed, 24 Feb 1999 09:15:05 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id IAA25277 for autofs-list; Wed, 24 Feb 1999 08:02:46 -0800 Received: from news.uni-kl.de (IDENT:mmdf@news.uni-kl.de [131.246.137.51]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id IAA25270 for ; Wed, 24 Feb 1999 08:02:37 -0800 Received: from mv.uni-kl.de ( mmdf@mail.mv.uni-kl.de [131.246.91.98] ) by news.uni-kl.de id aa09701 for ; 24 Feb 1999 17:03 MEZ Received: from mail.mv.uni-kl.de by mail.mv.uni-kl.de id aa07535; 24 Feb 99 17:11 MET Message-ID: <36D42290.9BEFE837@mv.uni-kl.de> Date: Wed, 24 Feb 1999 17:02:24 +0100 From: Percy Stoehr X-Mailer: Mozilla 4.5 [de] (WinNT; I) X-Accept-Language: de,en MIME-Version: 1.0 To: Autofs mailinglist Subject: Re: Automounter does not start except at bootup References: <199902221036.CAA27025@cesium.transmeta.com> <5laey6awju.fsf@tequila.cs.yale.edu> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > Damn! I admit I haven't had any trouble with autofs crashing or being killed > randomly, but I always hate it when I have to reboot a machine (or close > enough: kill all the processes using some of the submounts) just because > I somehow killed the automounter. And even tho as I said I haven't seen autofs > die, you can't dismiss the possibility. > Also, not being able to kill autofs without nasty side effects means that you > cannot upgrade autofs without those nasty side effects. > Being able to kill/restart daemons seems very important to me. AutoFS is no > exception here. I have no idea how hard it might be to provide such a feature, > and I'm not going to code it up any time soon, so I'll just stop whining. Yes ! In my oppinion it is very important to have an ability to kill every daemon. I remember in earlier Kernel-Versions ive used a timeout for my cdrom of 5 seconds. Some processes are using my cdrom and sometimes it happens that the cdrom keeps being locked. I dont know why this happened, and i couldnt reproduce it. Maybe it is fixed. Unmounting doesnt work cause unmount believes the path is being used ("busy"), but fuser outputs no users. Restarting of the daemon fails. The daemon was killed but could not be started again, cause the system believes the directory is mounted by the old daemon. The only way get my cdrom out, is rebooting the system ;). And i really hate rebooting my system. And i would have been really happy if a were able to start a new daemon taking over the control of the old one. Percy From owner-autofs@linux.kernel.org Wed Feb 24 10:01:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA28884 for ; Wed, 24 Feb 1999 10:01:08 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA28433 for autofs-list; Wed, 24 Feb 1999 09:02:45 -0800 Received: from tequila.cs.yale.edu (TEQUILA.CS.YALE.EDU [128.36.229.152]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA28428 for ; Wed, 24 Feb 1999 09:02:42 -0800 Received: from tequila.cs.yale.edu (localhost [127.0.0.1]) by tequila.cs.yale.edu (8.8.7/8.8.7) with SMTP id MAA13832 for ; Wed, 24 Feb 1999 12:02:36 -0500 To: autofs@linux.kernel.org From: Stefan Monnier Newsgroups: lists.linux.autofs Subject: Re: Automounter does not start except at bootup References: <199902221036.CAA27025@cesium.transmeta.com> <5laey6awju.fsf@tequila.cs.yale.edu> <36D42290.9BEFE837@mv.uni-kl.de> Date: 24 Feb 1999 12:02:31 -0500 Message-ID: <5lpv6zag60.fsf@tequila.cs.yale.edu> Lines: 19 X-Newsreader: Gnus v5.5/Emacs 20.3 Path: tequila.cs.yale.edu NNTP-Posting-Host: tequila.cs.yale.edu X-Trace: 24 Feb 1999 12:02:31 -0500, tequila.cs.yale.edu Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: >>>>> "Percy" == Percy Stoehr writes: > for my cdrom of 5 seconds. Some processes are using my cdrom and > sometimes it happens that the cdrom keeps being locked. I dont know why > this happened, and i couldnt reproduce it. Maybe it is fixed. Unmounting > doesnt work cause unmount believes the path is being used ("busy"), but > fuser outputs no users. Restarting of the daemon fails. The daemon was > killed but could not be started again, cause the system believes the > directory is mounted by the old daemon. The only way get my cdrom out, > is rebooting the system ;). And i really hate rebooting my system. And i > would have been really happy if a were able to start a new daemon taking > over the control of the old one. Note that the example is not a good reason to ask autofs to be killable. In your case, the problem is that the submount (created by mount via autofs) is somehow screwed up. No matter how many times you kill/restart autofs, you will not solve this problem. Stefan From owner-autofs@linux.kernel.org Wed Feb 24 10:50:55 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA28927 for ; Wed, 24 Feb 1999 10:50:55 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id JAA31368 for autofs-list; Wed, 24 Feb 1999 09:55:29 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id JAA31359 for ; Wed, 24 Feb 1999 09:55:23 -0800 Received: from robin.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id JAA27706 for ; Wed, 24 Feb 1999 09:55:05 -0800 X-BlackMail: robin.frogtown.com, robin.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 09:55:05(PST) on February 24, 1999 Date: Wed, 24 Feb 1999 09:55:12 -0800 (PST) From: Lori X-Sender: lori@scooter.corp.frogtown.com cc: autofs@linux.kernel.org Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: <199902232010.MAA22325@cesium.transmeta.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, 23 Feb 1999, H. Peter Anvin wrote: > > What I might try to do, is when parsing auto.master to replace all > direct mounts with a link into /.a or something, and make /.a an > indirect mount. > Hi - I probably missed the point - but are you implying that you could put a hook into autofs that would create this scenario when it encounters a direct key - ie: if you auto.master has a /- auto.direct-or-something you can have the clients automatically create the /.direct-hack -> /usr/direct-or-something ? I think that would be brilliant. It would fake the direct map automatically without having to modify the client by hand or include a specific site start up/install script for this type of thing. Since the /- convention is already there would it be easy enough to then create the symlink to direct mount fake for the client side? let's see - the auto.direct file, for example, might say something like: /usr/dist server1:/export/dist /tools server2:/export/tools and the action would be something like: symlinks= /usr/dist -> /.direct-usr-dist /tools -> /.direct-tools mountpoints= server1:/export/dist /.dir-usr-dist server2:/export/tools /.dir-tools Suppose you could end up with quite a few funny mount points. But it seems to me like a reasonable and automatic work around, if this is what you had in mind. I don't intend to keep pushing the point, you've made yourself quite clear on this issue - I just wanted to clarify if this is what you were telling me. cheers! From owner-autofs@linux.kernel.org Wed Feb 24 10:58:47 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id KAA28933 for ; Wed, 24 Feb 1999 10:58:47 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id KAA00171 for autofs-list; Wed, 24 Feb 1999 10:22:12 -0800 Received: from lithp.net (IDENT:root@lithp.net [209.220.40.228]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id KAA00166 for ; Wed, 24 Feb 1999 10:22:10 -0800 Received: from animal.corp.frogtown.com (root@localhost) by lithp.net (8.8.7/8.8.7) with ESMTP id KAA28155 for ; Wed, 24 Feb 1999 10:21:55 -0800 X-BlackMail: robin.frogtown.com, animal.corp.frogtown.com, , 209.220.40.226 X-Authenticated-Timestamp: 10:21:55(PST) on February 24, 1999 Date: Wed, 24 Feb 1999 10:21:54 -0800 (PST) From: Michael Newlyn Blake X-Sender: mblake@animal.corp.frogtown.com To: The autofs list Subject: Re: autofs v4: speak now .... documents In-Reply-To: Message-ID: X-Url: http://www.frogtown.com/~mblake MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: (trimed the to/cc line back down to just the list) On Wed, 24 Feb 1999, Patrick Goetz wrote: > Writing documentation, on the other hand, is not fun, at least not for > most people, so it is harder to find people who will commit a couple of > hundred (or more) hours of free time to a documentation project. Actually, I disagree here. I have met hundreds of system administrators in my almost 10 years as a sysadmin. Most of them aren't particularly good coders, nor have they needed to be. All you really need to be a sysadmin is the ability to understand how the fundamental mechanisms of the system work and interact with eachother, the ability to think creatively, and to be able to explain what's going on. Hmmm... sounds a lot like a tech writer to me. ;) I'm sure a lot of us would appreciate the opportunity to be able to actually contribute something to these projects. And since we have skills almost idealy suited for playing with thing and breaking them, then being able to explain how we broke them, and writing up the way things should work, we would make excelent non-coding testers and doc writers. Not that I have a whole lot of free time, but who does? I would still like to try. (Maybe I should just check out the ldp site and see if they already have a place for people to volumnteer for something like this, but it might be best if people can offer their services to the projects that atually matter to them.) > which comes with linux is an order of magnitude better than that which > comes with Solaris (at least - so far I've found Answerbook to be 100% > useless), for example, so I guess that corporations are even more Now, let's be fair.. 95% useless. There have been times when it gave imformation almost specific enough that I was able to infer the right answer from the crap it contained. Not often but... Cheers. ----- In case of fire, stand in the hall and shout "Fire!" -- The Kidner Report From owner-autofs@linux.kernel.org Wed Feb 24 11:00:39 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id LAA28937 for ; Wed, 24 Feb 1999 11:00:39 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id KAA32179 for autofs-list; Wed, 24 Feb 1999 10:09:12 -0800 Received: from tiktok.cygnus.com (tiktok.cygnus.com [192.80.44.20]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id KAA32176 for ; Wed, 24 Feb 1999 10:09:11 -0800 Received: (from meissner@localhost) by tiktok.cygnus.com (8.8.7/8.8.5) id NAA02620; Wed, 24 Feb 1999 13:09:00 -0500 Date: Wed, 24 Feb 1999 13:09:00 -0500 From: Michael Meissner To: Frithjof Anders Cc: Michael Meissner , autofs mailing list Subject: Re: autofs v4: speak now or forever hold your peace Message-ID: <19990224130900.A2606@tiktok.cygnus.com> References: <19990223102907.A9100@tiktok.cygnus.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95i In-Reply-To: ; from Frithjof Anders on Tue, Feb 23, 1999 at 05:23:05PM +0100 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Tue, Feb 23, 1999 at 05:23:05PM +0100, Frithjof Anders wrote: (my post about symlink mounts to the current host deleted) > That work already :) Learn something every day. However, it is not mentioned in any of the man pages I checked, nor in the /usr/doc/autofs* files. -- Michael Meissner, Cygnus Solutions PO Box 98, Ayer Massachusetts, USA 01432-0098 meissner@cygnus.com From owner-autofs@linux.kernel.org Wed Feb 24 13:53:17 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id NAA29042 for ; Wed, 24 Feb 1999 13:53:17 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA09304 for autofs-list; Wed, 24 Feb 1999 13:05:41 -0800 Received: from reddevil.CS.Berkeley.EDU (IDENT:mangoman@reddevil.CS.Berkeley.EDU [128.32.131.199]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA09299 for ; Wed, 24 Feb 1999 13:05:38 -0800 Received: (from mangoman@localhost) by reddevil.CS.Berkeley.EDU (8.8.7/8.8.7) id MAA00689 for autofs@linux.kernel.org; Wed, 24 Feb 1999 12:53:30 -0800 Message-ID: <19990224125330.A660@reddevil.CS.Berkeley.EDU> Date: Wed, 24 Feb 1999 12:53:30 -0800 From: Jeff Mangewala To: autofs@linux.kernel.org Subject: direct map Reply-To: mangoman@CS.Berkeley.EDU Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.2 Organization: sometimes X-OS: Linux 2.2.1 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Hello, I'm running Redhat 5.2 I upgraded to kernel v2.2.1 and installed all of the updates from: ftp://metalab.unc.edu/pub/Linux/distributions/redhat/updates/5.2/kernel-2.2/i386 Then I compiled autofs 3.1.3 % automount -v Linux automount version 3.1.3 If I have a /etc/auto.master like this: /var/spool /etc/auto.mail and a /etc/auto.mail like this: mail -rw,fstype=nfs,nolock,actimeo=0 mailspool:/export/spool/mail Then I can't see anything in /var/spool (like mqueue) One way I've fixed this problem is: /etc/auto.master: /misc /etc/auto.misc --timeout 60 /etc/auto.misc: mail -rw,fstype=nfs,nolock,actimeo=0 mailspool:/export/spool/mail ln -s /misc/mail /var/spool/mail Is there a better way? Is there a plan to put direct maps into autofs? I think "amd" might be able to do what I want, but I've heard of a bunch of improvements going on in autofs...I've been unable to find good documentation on autofs. Is amd gonna go away??? I would appreciate pointers to all things autofs in Linux. Is there an archive of this mailing list? Thanks, Jeff Mangewala From owner-autofs@linux.kernel.org Wed Feb 24 14:17:30 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA29076 for ; Wed, 24 Feb 1999 14:17:29 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id NAA11548 for autofs-list; Wed, 24 Feb 1999 13:42:38 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id NAA11545 for ; Wed, 24 Feb 1999 13:42:36 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id NAA14682; Wed, 24 Feb 1999 13:42:21 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id NAA03214; Wed, 24 Feb 1999 13:42:21 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id NAA02542; Wed, 24 Feb 1999 13:42:21 -0800 Message-Id: <199902242142.NAA02542@cesium.transmeta.com> Subject: Re: shutdown autofs In-Reply-To: from Frithjof Anders at "Feb 24, 99 11:34:10 am" To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 13:42:21 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > On the other hand the AIX automounter allows you to do exactly what I want > via shared mounts AIX doing it is, if anything, a good reason *NOT* to... > Why do you think, putting extra code in the kernel is better than > allowing a solution on deamon level? By no means I want to trigger a > discussion of a monolitic kernel vesus the mach approach. The monolitic won, > but learned how to handle modules ;-) What I'm saying is that namespace issues matter. Apparently it doesn't matter to *you*, but you're probably in the minority on this one. -hpa From owner-autofs@linux.kernel.org Wed Feb 24 14:42:06 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id OAA29083 for ; Wed, 24 Feb 1999 14:42:06 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA12777 for autofs-list; Wed, 24 Feb 1999 14:02:45 -0800 Received: from chipmunk.cita.utoronto.ca (chipmunk.cita.utoronto.ca [128.100.76.49]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id OAA12774 for ; Wed, 24 Feb 1999 14:02:43 -0800 Received: from localhost (pogosyan@localhost) by chipmunk.cita.utoronto.ca (8.6.13/8.6.12) with SMTP id RAA02367; Wed, 24 Feb 1999 17:02:40 -0500 Date: Wed, 24 Feb 1999 17:02:39 -0500 (EST) From: Dmitry Pogosyan To: mangoman@CS.Berkeley.EDU cc: autofs@linux.kernel.org Subject: Re: direct map In-Reply-To: <19990224125330.A660@reddevil.CS.Berkeley.EDU> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Am I right, that with autofs you can see only directories which are mounted at the time you check mount point (like ls /var/spool). AMD has a 'browser' option which allowed to see everything which is defined to be mounted without actually mounting them ? I'd like this be possible with autofs. On Wed, 24 Feb 1999, Jeff Mangewala wrote: > Hello, > > > I'm running Redhat 5.2 > I upgraded to kernel v2.2.1 and installed all of the updates from: > ftp://metalab.unc.edu/pub/Linux/distributions/redhat/updates/5.2/kernel-2.2/i386 > > Then I compiled autofs 3.1.3 > % automount -v > Linux automount version 3.1.3 > > If I have a /etc/auto.master like this: > /var/spool /etc/auto.mail > > and a /etc/auto.mail like this: > mail -rw,fstype=nfs,nolock,actimeo=0 mailspool:/export/spool/mail > > Then I can't see anything in /var/spool (like mqueue) > > One way I've fixed this problem is: > > /etc/auto.master: > /misc /etc/auto.misc --timeout 60 > > /etc/auto.misc: > mail -rw,fstype=nfs,nolock,actimeo=0 mailspool:/export/spool/mail > > ln -s /misc/mail /var/spool/mail > > > > Is there a better way? > > Is there a plan to put direct maps into autofs? > > > I think "amd" might be able to do what I want, but I've heard > of a bunch of improvements going on in autofs...I've been unable > to find good documentation on autofs. > > Is amd gonna go away??? > > I would appreciate pointers to all things autofs in Linux. > > Is there an archive of this mailing list? > > > Thanks, > > Jeff Mangewala > From owner-autofs@linux.kernel.org Wed Feb 24 15:11:06 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA29099 for ; Wed, 24 Feb 1999 15:11:06 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA14397 for autofs-list; Wed, 24 Feb 1999 14:24:53 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA14394 for ; Wed, 24 Feb 1999 14:24:52 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id OAA15340; Wed, 24 Feb 1999 14:23:49 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id OAA06166; Wed, 24 Feb 1999 14:23:49 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id OAA02940; Wed, 24 Feb 1999 14:23:49 -0800 Message-Id: <199902242223.OAA02940@cesium.transmeta.com> Subject: Re: direct map In-Reply-To: from Dmitry Pogosyan at "Feb 24, 99 05:02:39 pm" To: pogosyan@cita.utoronto.ca (Dmitry Pogosyan) Date: Wed, 24 Feb 1999 14:23:49 -0800 (PST) Cc: mangoman@CS.Berkeley.EDU, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > Am I right, that with autofs you can see only directories which are > mounted at the time you check mount point (like ls /var/spool). > AMD has a 'browser' option which allowed to see everything which > is defined to be mounted without actually mounting them ? > I'd like this be possible with autofs. > It's technically very difficult to do, but it is something that is obviously useful, so I'm trying to figure out how to do it in autofs v4. The problem is that almost everything that browses does both a readdir() and a stat(); the readdir() is easy but keeping the stat() from mounting the filesystem is really, really hard... -hpa From owner-autofs@linux.kernel.org Wed Feb 24 15:12:53 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA29103 for ; Wed, 24 Feb 1999 15:12:52 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA15670 for autofs-list; Wed, 24 Feb 1999 14:43:34 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA15666 for ; Wed, 24 Feb 1999 14:43:33 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id OAA13878; Wed, 24 Feb 1999 14:43:07 -0800 From: Alvin Oga Message-Id: <199902242243.OAA13878@leopard.imedia.com> Subject: Re: Automounter does not start except at bootup To: stoehr@mv.uni-kl.de (Percy Stoehr) Date: Wed, 24 Feb 1999 14:43:07 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <36D42290.9BEFE837@mv.uni-kl.de> from "Percy Stoehr" at Feb 24, 99 05:02:24 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi percy > Maybe it is fixed. autofs-3.1.* is a lot better with 2.0.36 and 2.2.* ( no more randomly dying autofs like it used to ) > Unmounting > doesnt work cause unmount believes the path is being used ("busy"), but > fuser outputs no users. this fuser problem occurs for manually mounted directories tooo... I claim it is NOT an autofs problea because I manually run mount, NOT autofs.... - yes... I have a few machines where /Backup is not unmountable anymore after the nightly backups... have fun alvin From owner-autofs@linux.kernel.org Wed Feb 24 15:15:27 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA29129 for ; Wed, 24 Feb 1999 15:15:26 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA14787 for autofs-list; Wed, 24 Feb 1999 14:30:23 -0800 Received: from iwr1.iwr.uni-heidelberg.de (iwr1.iwr.uni-heidelberg.de [129.206.104.40]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA14781 for ; Wed, 24 Feb 1999 14:30:21 -0800 Received: from giotto.iwr.uni-heidelberg.de (giotto.iwr.uni-heidelberg.de [129.206.107.82]) by iwr1.iwr.uni-heidelberg.de (8.8.8/8.8.8) with ESMTP id XAA26391; Wed, 24 Feb 1999 23:30:09 +0100 (MET) Received: by giotto.iwr.uni-heidelberg.de (8.8.7/971006.SGImailhost) id WAA26636; Wed, 24 Feb 1999 22:29:53 GMT From: "Hermann Lauer" Message-Id: <9902242329.ZM26634@giotto.iwr.uni-heidelberg.de> Date: Wed, 24 Feb 1999 23:29:53 +0100 In-Reply-To: Ronny Ranerup "Re: autofs v4: speak now or forever hold your peace" (Feb 23, 11:30am) References: <199902231030.LAA00939@pcronny.axis.se> Received: by NeXT.Mailer (1.118.2.RR) X-Mailer: Z-Mail (3.2.3 08feb96 MediaMail) To: Ronny Ranerup , autofs@linux.kernel.org Subject: Re: autofs v4: speak now or forever hold your peace Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Feb 23, 11:30am, Ronny Ranerup wrote: > Subject: Re: autofs v4: speak now or forever hold your peace > > frithjof@fkp.tu-darmstadt.de: > >I totally agree. A daemon is a service program which should be able to be > >started and stopped like any other service program by the system administator. > >Especially under Linux, which is one of most stable operation systems I worked > >with, it would help to design daemons in a coherent service friendly manner. > > I agree with the agreement. But I guess I also understand that it > might be difficult... AMD was the singularly most crappy program for > Linux and the only reason we had to reboot our machines (several > months apart, but anyway). You know the procedure; hmm, can't login, > go to machine, see several screens of amd messages on console, try to > kill amd, fail, try to unmount all NFS mounts, fail, try to figure out > which bloody proc holds the mount, fail, getting damn cold in the > server room, getting mad, killing all user processes, killing amd, > killing NFS, bla, bla. Rebooting because the bloody kernel refuses to > forget some silly old mount... amd could be killed and restarted with not to much problems - the only thing are to unmount the automouter directory itself with "umount" (also works for autofs, if there aren't any submounts) and to use the right option wile restarting amd. BTW.: amd is now found in am-utils, link see below. > Another point for the wishlist; since I guess most of the known Unix > universe is using AMD, a script to parse AMD-maps and generate autofs > ones would probably be very helpful, as you can see from this list. I > wrote one myself but it's just a trifle and won't do for serious > distribution. That would be a very cool script, because on the other unices there's nothing available besides amd (and autofs couldn't be ported, as it lives in the kernel). But maybe it's not so easy - how to convert the "wire" feature and the replicated server feature ? The coolest thing would be an am-utils interpreter for autofs or a modification of am-utils (amd) to use the kernel autofs - I looked at the code, but it's not so easy it could be done in my spare time. > More whining... after fighting with AMD and trying to figure out where > the hell those amd-upl-xxx came from I didn't get happier when trying > to find autofs, no official web-site, just a ftp archive, already a > frightening number of strange autofs distributions. A NEWS file in the > distribution where there are almost no differences between version > 0.3.14 and 3.1.2. No HOWTO or similar document. No explanation as to > how the daemon ties into the kernel, no explanation on which version > of the kernel is needed... For amd (am-utils): http://www.cs.columbia.edu/~ezk/am-utils/ For autofs: Hmm, at least the FAQ is available somewher on the WWW. > Yeah, I know, a lot of whining. Sorry. On the positive side, autofs > _might_ be less buggy then AMD... ;) hpa seems to love simplicity, so you are probably right. Thanks for all work done. Greetings Hermann From owner-autofs@linux.kernel.org Wed Feb 24 15:21:54 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA29152 for ; Wed, 24 Feb 1999 15:21:53 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id OAA15904 for autofs-list; Wed, 24 Feb 1999 14:47:10 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id OAA15900 for ; Wed, 24 Feb 1999 14:47:09 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id OAA13893; Wed, 24 Feb 1999 14:46:36 -0800 From: Alvin Oga Message-Id: <199902242246.OAA13893@leopard.imedia.com> Subject: Re: autofs v4: speak now .... documents To: pgoetz@fireant.ma.utexas.edu (Patrick Goetz) Date: Wed, 24 Feb 1999 14:46:35 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Patrick Goetz" at Feb 24, 99 10:03:45 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya all... yes people write docs when time is available or have a need to do so... and if there is stuff you'd like to add for the autofs howto.... please let me know and I will put it in... thank you alvin http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > > should be an associated legion of doc writers? so that you don't get > > engineers writing intelligible documentation, free them up for more kernel > > hacking - or is that what the linux documentation project is supposed to > > do? > > > > just stirring the flames... > > > > > I totally agree with you. Here is the problem, though. Hackers write code > because it is fun. Nobody would be slaving over a hot terminal wondering > whether to use SIGTERM or SIGQUIT for free if they didn't enjoy doing it. > Writing documentation, on the other hand, is not fun, at least not for > most people, so it is harder to find people who will commit a couple of > hundred (or more) hours of free time to a documentation project. > > On the other hand, my experience has been that the free documentation > which comes with linux is an order of magnitude better than that which > comes with Solaris (at least - so far I've found Answerbook to be 100% > useless), for example, so I guess that corporations are even more > reluctant to PAY someone to write documentation than the average joe > schmoe is to write it for free. > > From owner-autofs@linux.kernel.org Wed Feb 24 15:52:03 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id PAA29255 for ; Wed, 24 Feb 1999 15:52:02 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA16767 for autofs-list; Wed, 24 Feb 1999 15:01:04 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA16763 for ; Wed, 24 Feb 1999 15:01:02 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA14010; Wed, 24 Feb 1999 15:00:22 -0800 From: Alvin Oga Message-Id: <199902242300.PAA14010@leopard.imedia.com> Subject: Re: direct map To: mangoman@CS.Berkeley.EDU Date: Wed, 24 Feb 1999 15:00:22 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <19990224125330.A660@reddevil.CS.Berkeley.EDU> from "Jeff Mangewala" at Feb 24, 99 12:53:30 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi ya jeff > Is there a better way? let me know too... heheehe > I would appreciate pointers to all things autofs in Linux. have some docs at: http://www.linux-consulting.com/Amd_AutoFS/autofs-HOWTO.html ( donno why it's not propagating into the HOWTO dirs at LDP ) > Is there an archive of this mailing list? I keep a copy of all postings since I started playing/using autofs since back in Jan-98 have fun alvin From owner-autofs@linux.kernel.org Wed Feb 24 16:02:39 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA29263 for ; Wed, 24 Feb 1999 16:02:38 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA18199 for autofs-list; Wed, 24 Feb 1999 15:24:20 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA18195 for ; Wed, 24 Feb 1999 15:24:18 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA16506; Wed, 24 Feb 1999 15:24:01 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA12090; Wed, 24 Feb 1999 15:24:00 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA03390; Wed, 24 Feb 1999 15:24:00 -0800 Message-Id: <199902242324.PAA03390@cesium.transmeta.com> Subject: Re: Automounter does not start except at bootup In-Reply-To: <199902242243.OAA13878@leopard.imedia.com> from Alvin Oga at "Feb 24, 99 02:43:07 pm" To: aoga@imedia.com (Alvin Oga) Date: Wed, 24 Feb 1999 15:24:00 -0800 (PST) Cc: stoehr@mv.uni-kl.de, aoga@leopard.imedia.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: > > hi percy > > > Maybe it is fixed. > > autofs-3.1.* is a lot better with 2.0.36 and 2.2.* > ( no more randomly dying autofs like it used to ) > > > Unmounting > > doesnt work cause unmount believes the path is being used ("busy"), but > > fuser outputs no users. > > this fuser problem occurs for manually mounted directories tooo... > I claim it is NOT an autofs problea because I manually run mount, NOT autofs.... > - yes... I have a few machines where /Backup is not unmountable > anymore after the nightly backups... > Yes, fuser seems to have problems. I have seen this too. Perhaps I need to upgrade my procps. -hpa From owner-autofs@linux.kernel.org Wed Feb 24 16:34:18 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA29303 for ; Wed, 24 Feb 1999 16:34:17 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA20824 for autofs-list; Wed, 24 Feb 1999 16:03:42 -0800 Received: from vindaloo.atnf.CSIRO.AU (vindaloo.atnf.CSIRO.AU [130.155.198.47]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id QAA20815 for ; Wed, 24 Feb 1999 16:03:33 -0800 Received: (from rgooch@localhost) by vindaloo.atnf.CSIRO.AU (8.6.12/8.6.12) id LAA01596; Thu, 25 Feb 1999 11:03:15 +1100 Date: Thu, 25 Feb 1999 11:03:15 +1100 Message-Id: <199902250003.LAA01596@vindaloo.atnf.CSIRO.AU> From: Richard Gooch To: autofs@linux.kernel.org Subject: On direct mounts: exploring an idea Notfrom: spamlog@atnf.csiro.au Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Hi, all. A random thought just popped into my head which may be a reasonable compromise for implementing direct mounts. To start, I'll summarise my understand of why the in-kernel solution won't happen, as a reality check. I'll then propose my solution. Say you want a direct map in / and you have "data1" and "data2" as two of the prospective mount points. The in-kernel solution would require the automounter to get its grubby hands on each dentry lookup operation in / which in turn implies that every dentry lookup has to pass through the automounter. Aside: perhaps this can be mitigated by attaching a dentry lookup overriding method for / but then we still pay the penalty of an extra test for every lookup, and for / we have an indirect function call. It's been proposed that an alternative solution is to populate a directory with a direct map with symlinks. Unfortunately, that looks a little ugly, and would cause them all to be mounted, which is not desirable, I presume. OK, my idea (and remember: I did say it just popped into my head:-) is to define a new inode type. Call it a "fake special" ("fakes" for short). For a direct map, the automounter creates fakes in /, so you would have: /data1 and /data2 being fakes. The filesystem (i.e. ext2) would do the normal inode lookup thing as required, but either the FS or the VFS would do something different if a fake is returned: the device number or the name of the fake is passed to the automounter for validation. If the automounter approves, the inode is changed to a normal directory, then the automounter is told to go ahead and do its mounting, and finally the inode is returned. If the automounter does no approve, the inode is discarded and the dentry is dropped. The effect is that there is no dentry for that name. For directory lookups, fakes can be ignored, so won't show /data1 and /data2. This could be done in the FS or in the VFS functions. When a direct mount expires, the automounter signals the kernel to change the mount point back to a fake and drop the dentry. Finally, a method is required to remove fakes. Probably the simplest would be to convert VFS fake inodes to directory inodes when the automounter connection is not available, just like the case when the automounter validates the fake, except that no mounting is performed. You could then have the automounter breaking its connection to the kernel and proceed to unlink(2) the fakes. I readily admit this involves some hackery in the kernel and the automounter, but it has the advantage of looking clean from userspace and not throwing away speed in our fast VFS/dentry layer. The only real overhead I see is the check for fakes after an inode lookup() method is called. This seems pretty cheap to me. This scheme won't slow down cached dentry lookups, which is the principal concern that Linus has, if I'm not mistaken. Anyway, this idea is somewhat half-baked. If there's some fundamental flaw, I'm sure you won't be backwards about being forwards ;-) Regards, Richard.... From owner-autofs@linux.kernel.org Wed Feb 24 16:48:56 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA29309 for ; Wed, 24 Feb 1999 16:48:56 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA21776 for autofs-list; Wed, 24 Feb 1999 16:19:12 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA21773 for ; Wed, 24 Feb 1999 16:19:11 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id QAA17587; Wed, 24 Feb 1999 16:18:08 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id QAA17442; Wed, 24 Feb 1999 16:18:07 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id QAA06390; Wed, 24 Feb 1999 16:18:06 -0800 Message-Id: <199902250018.QAA06390@cesium.transmeta.com> Subject: Re: On direct mounts: exploring an idea In-Reply-To: <199902250003.LAA01596@vindaloo.atnf.CSIRO.AU> from Richard Gooch at "Feb 25, 99 11:03:15 am" To: rgooch@atnf.CSIRO.AU (Richard Gooch) Date: Wed, 24 Feb 1999 16:18:06 -0800 (PST) Cc: autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: > > It's been proposed that an alternative solution is to populate a > directory with a direct map with symlinks. Unfortunately, that looks a > little ugly, and would cause them all to be mounted, which > is not desirable, I presume. > This is a general problem with direct mounts anyway. The technical issue is that the Linux VFS doesn't support "sandwich mounts", i.e. mounting a filesystem on top of the root of another filesystem (in the case of a direct mount, the sandwiched filesystem would be autofs, of course.) (Incidentally, we *know* this is what Solaris uses.) Sandwich mounts would eliminate an important optimization in the VFS, and we don't want that (after all, the Linux filesystem is probably the fastest one on the market, and we *want* that!) However, you bring up a perfectly valid point, and one which I hadn't really considered: it may be possible to use the dentry system to trapdoor a directory without making it a proper mount. This is definitely v2.3 stuff and means autofs-specific hooks in the VFS, but if it can be done cleanly enough Linus just might let me do it. Now you've given me something to think about. I can't say offhand if this is doable, but I can't say offhand that it *isn't*, either... Thanks for adding some technical content to this particular issue, which was starting to degenerate into a few people thinking that if they just whined loudly enough they would get their way... -hpa From owner-autofs@linux.kernel.org Wed Feb 24 17:13:59 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA29321 for ; Wed, 24 Feb 1999 17:13:59 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA22637 for autofs-list; Wed, 24 Feb 1999 16:36:57 -0800 Received: from vindaloo.atnf.CSIRO.AU (vindaloo.atnf.CSIRO.AU [130.155.198.47]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id QAA22633 for ; Wed, 24 Feb 1999 16:36:53 -0800 Received: (from rgooch@localhost) by vindaloo.atnf.CSIRO.AU (8.6.12/8.6.12) id LAA02193; Thu, 25 Feb 1999 11:36:45 +1100 Date: Thu, 25 Feb 1999 11:36:45 +1100 Message-Id: <199902250036.LAA02193@vindaloo.atnf.CSIRO.AU> From: Richard Gooch To: "H. Peter Anvin" Cc: autofs@linux.kernel.org Subject: Re: On direct mounts: exploring an idea In-Reply-To: <199902250018.QAA06390@cesium.transmeta.com> References: <199902250003.LAA01596@vindaloo.atnf.CSIRO.AU> <199902250018.QAA06390@cesium.transmeta.com> Notfrom: spamlog@atnf.csiro.au Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: H. Peter Anvin writes: > > > > It's been proposed that an alternative solution is to populate a > > directory with a direct map with symlinks. Unfortunately, that looks a > > little ugly, and would cause them all to be mounted, which > > is not desirable, I presume. > > This is a general problem with direct mounts anyway. > > The technical issue is that the Linux VFS doesn't support "sandwich > mounts", i.e. mounting a filesystem on top of the root of another > filesystem (in the case of a direct mount, the sandwiched filesystem > would be autofs, of course.) (Incidentally, we *know* this is what > Solaris uses.) I don't understand how sandwich mounts come into the discussion, at least not when using the fakes I was talking about. What am I missing? I suppose you're referring to the normal way the automounter works, which is to mount an autofs on top of a mount point, and then have submount points under that. But I don't understand why you *need* to mount an autofs with the scheme I outlined. > Sandwich mounts would eliminate an important optimization in the > VFS, and we don't want that (after all, the Linux filesystem is > probably the fastest one on the market, and we *want* that!) Absolutely. When you have a system that is so fast that it *feels* faster almost immediately after logging in (cf. Solaris), you know you've got something good. > However, you bring up a perfectly valid point, and one which I > hadn't really considered: it may be possible to use the dentry > system to trapdoor a directory without making it a proper mount. Ah. So you're actually considering adding an optional lookup method to dentries? I discarded that idea because it would add an extra test to each lookup (at each level). > This is definitely v2.3 stuff and means autofs-specific hooks in the > VFS, but if it can be done cleanly enough Linus just might let me do > it. Now you've given me something to think about. I can't say > offhand if this is doable, but I can't say offhand that it *isn't*, > either... OK. But could you please explain why the fake idea is no good? I like it (despite a certain ugliness to it) because it doesn't slow down cached dentry lookups *at all*. > Thanks for adding some technical content to this particular issue, > which was starting to degenerate into a few people thinking that if > they just whined loudly enough they would get their way... Thanks. I guess I managed to see both sides of the argument. I use autofs in an environment where we have NIS mount maps from Solaris boxes, and I had to hack the startup script to ignore direct maps. It would be good to support them. It's not essential to me, but I can see how it is for others. On the other hand, performance matters a lot to me (it's a wonderful selling point too), so I figured we needed a way out of the impasse. I've done my best work when arguing with someone (wanting some or other feature) just as stubborn as me :-) Eventually, an approach pops out that gives them what they need but doesn't compromise my design. Sigh. Those were good days. Regards, Richard.... From owner-autofs@linux.kernel.org Wed Feb 24 19:00:44 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA29417 for ; Wed, 24 Feb 1999 19:00:44 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id SAA28651 for autofs-list; Wed, 24 Feb 1999 18:27:25 -0800 Received: from fks15.eastnet.gatech.edu (fks15.eastnet.gatech.edu [128.61.104.15]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id SAA28647 for ; Wed, 24 Feb 1999 18:27:22 -0800 Received: from j by fks15.eastnet.gatech.edu with local (Exim 2.05 #1 (Debian)) id 10FqWO-00008R-00; Wed, 24 Feb 1999 21:27:28 -0500 Date: Wed, 24 Feb 1999 21:27:28 -0500 From: Jason Lunz To: autofs@linux.kernel.org Subject: BUG: in 2.2.2, autofs mount of SCSI CD-ROM on aha1542 card locks up machine Message-ID: <19990224212728.A500@cc.gatech.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.3i Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Well, here's a nice bug for you all. I'd appreciate knowing what went wrong if/when it gets figured out. Feel free to contact me for more info. Please let me know if there's a better place I should I have sent this. without further ado: the standardized bug report. [1.] in 2.2.2, autofs mount of SCSI CD-ROM on aha1542 card locks up machine [2.] I've got both autofs and support for my Adaptec aha1542 SCSI card compiled into the kernel. Now, trying to automount the CD-ROM locks up my machine. I can still switch between VC's, but I can't log in on any of them. When telnetting to the machine, I get a connection but nothing else...it's like inetd/tcpd accepts but doesn't hand off to in.telnetd. This wasn't a problem in the 2.2.1 kernel. [3.] keywords: autofs, scsi, aha1542 [4.] Linux version 2.2.2 (j@fks15) (gcc version 2.7.2.3) #1 Tue Feb 23 20:43:49 EST 1999 [5.] automounting the cd-rom produces output like the following: $ ls /vol/cd Detected scsi CD-ROM at scsi0, channel 0, id 4, lun 0 aha1542c: Trying device reset for target 4 bug: kernel timer added twice at c017ec44 I don't know if 'c017ec44' is a kernel address, but if it is, the relevant portion of my 'nm vmlinux | sort' output is: c017e8ac T scsi_eh_action_done c017e8ec T scsi_sense_valid c017e908 T scsi_eh_retry_command c017e954 T scsi_request_sense c017eaa0 T scsi_test_unit_ready c017ebec T scsi_sleep_done c017ec00 T scsi_sleep c017ec64 T scsi_send_eh_cmnd c017ed80 T scsi_unit_is_ready c017edc4 T scsi_eh_finish_command c017ede4 T scsi_try_to_abort_command c017ee24 T scsi_try_bus_device_reset c017ee6c T scsi_try_bus_reset (I guess that'd make sense...problems with timers in the 'scsi_sleep' function. If I were any more motivated, I might try to learn kernel coding :). [6.] above [7.] Environment - ??? [7.1.] Software (add the output of the ver_linux script here) -- Versions installed: (if some fields are empty or looks -- unusual then possibly you have very old versions) Linux fks15 2.2.2 #1 Tue Feb 23 20:43:49 EST 1999 i586 unknown Kernel modules 2.1.121 Gnu C 2.7.2.3 Binutils 2.9.1.0.19 Linux C Library 2.0.7 Dynamic linker ldd: version 1.9.10 Procps 1.2.9 Mount 2.9g Net-tools (1998-03-02) Kbd 0.96 Sh-utils 1.16 [7.2.] Processor information (from /proc/cpuinfo): processor : 0 vendor_id : GenuineIntel cpu family : 5 model : 4 model name : Pentium MMX stepping : 3 cpu MHz : 233.226225 fdiv_bug : no hlt_bug : no sep_bug : no f00f_bug : yes fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr mce cx8 mmx bogomips : 465.31 [7.3.] Module information (from /proc/modules): I've got lots of modules compiled, but none have to be loaded for this to happen [7.4.] SCSI information (from /proc/scsi/scsi) Attached devices: Host: scsi0 Channel: 00 Id: 00 Lun: 00 Vendor: CONNER Model: CP30540 545MB3.5 Rev: B0BC Type: Direct-Access ANSI SCSI revision: 02 Host: scsi0 Channel: 00 Id: 01 Lun: 00 Vendor: QUANTUM Model: LPS540S Rev: 590S Type: Direct-Access ANSI SCSI revision: 02 Host: scsi0 Channel: 00 Id: 02 Lun: 00 Vendor: CONNER Model: CFA540S Rev: 0EAD Type: Direct-Access ANSI SCSI revision: 02 Host: scsi0 Channel: 00 Id: 04 Lun: 00 Vendor: NEC Model: CD-ROM DRIVE:501 Rev: 2.2 Type: CD-ROM ANSI SCSI revision: 02 [7.5.] Other information that might be relevant to the problem (please look in /proc and include all information that you think to be relevant): I've got a Pro Audio Spectrum 16 card, which doesn't play nice with the aha1542 card, but the worst problem that's ever come of that has been problems playing sound. I haven't done anything with sound in the 2.2.2 kernel anyway. thanks for your help, J From owner-autofs@linux.kernel.org Wed Feb 24 19:35:28 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id TAA29457 for ; Wed, 24 Feb 1999 19:35:28 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id TAA30288 for autofs-list; Wed, 24 Feb 1999 19:00:36 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id TAA30285 for ; Wed, 24 Feb 1999 19:00:34 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id TAA20022; Wed, 24 Feb 1999 19:00:00 -0800 From: Alvin Oga Message-Id: <199902250300.TAA20022@leopard.imedia.com> Subject: Re: BUG: in 2.2.2, autofs mount of SCSI CD-ROM on aha1542 card locks up machine To: j@cc.gatech.edu (Jason Lunz) Date: Wed, 24 Feb 1999 19:00:00 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <19990224212728.A500@cc.gatech.edu> from "Jason Lunz" at Feb 24, 99 09:27:28 pm Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: hi ya jason when using linux-2.2.*.... are yoou also running compatible binaries and executables ?? I am running autofs on 2.2.1 also and is not an issue but have not tried under 2.2.2 eg: try using egcs-2.90.29 instead ??? - had problems compiling has with gcc-2.7.2.3.... > Linux fks15 2.2.2 #1 Tue Feb 23 20:43:49 EST 1999 i586 unknown > Kernel modules 2.1.121 > Gnu C 2.7.2.3 > Binutils 2.9.1.0.19 > Linux C Library 2.0.7 > Dynamic linker ldd: version 1.9.10 > Procps 1.2.9 > Mount 2.9g > Net-tools (1998-03-02) > Kbd 0.96 > Sh-utils 1.16 okay.... looks good from the little I know... thanx alvin http://www.Linux-Consulting.com/linux-2.2/upgrade* From owner-autofs@linux.kernel.org Wed Feb 24 22:12:01 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id WAA29543 for ; Wed, 24 Feb 1999 22:12:01 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id VAA05030 for autofs-list; Wed, 24 Feb 1999 21:27:07 -0800 Received: from fuzzy.foxtaur.com (tnt-dal-stat-47.dallas.net [209.44.42.47]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id VAA05025 for ; Wed, 24 Feb 1999 21:27:04 -0800 Received: (from fox@localhost) by fuzzy.foxtaur.com (8.7.5/8.6.11) id XAA05473 for autofs@linux.kernel.org; Wed, 24 Feb 1999 23:22:19 -0600 Date: Wed, 24 Feb 1999 23:22:19 -0600 From: Fuzzy Fox To: autofs@linux.kernel.org Subject: Re: direct map Message-ID: <19990224232219.B5180@fuzzy.foxtaur.com> Mail-Followup-To: autofs@linux.kernel.org References: <199902242223.OAA02940@cesium.transmeta.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.1i In-Reply-To: <199902242223.OAA02940@cesium.transmeta.com>; from H. Peter Anvin on Wed, Feb 24, 1999 at 02:23:49PM -0800 Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: H. Peter Anvin wrote: > > The problem is that almost everything that browses does both a > readdir() and a stat(); the readdir() is easy but keeping the stat() > from mounting the filesystem is really, really hard... In my experience with other automounters, that's how they work. If you stat() the mount-point, the filesystem is mounted. I don't see it as a problem; just the cost of doing business. :) -- fox@dallas.net (Fuzzy Fox) || "Nothing takes the taste out of peanut sometimes known as David DeSimone || butter quite like unrequited love." http://www.dallas.net/~fox/ || -- Charlie Brown From owner-autofs@linux.kernel.org Wed Feb 24 22:16:10 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id WAA29569 for ; Wed, 24 Feb 1999 22:16:10 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id VAA05606 for autofs-list; Wed, 24 Feb 1999 21:38:22 -0800 Received: from vindaloo.atnf.CSIRO.AU (vindaloo.atnf.CSIRO.AU [130.155.198.47]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id VAA05603 for ; Wed, 24 Feb 1999 21:38:17 -0800 Received: (from rgooch@localhost) by vindaloo.atnf.CSIRO.AU (8.6.12/8.6.12) id QAA04777; Thu, 25 Feb 1999 16:38:06 +1100 Date: Thu, 25 Feb 1999 16:38:06 +1100 Message-Id: <199902250538.QAA04777@vindaloo.atnf.CSIRO.AU> From: Richard Gooch To: Fuzzy Fox Cc: autofs@linux.kernel.org Subject: Re: direct map In-Reply-To: <19990224232219.B5180@fuzzy.foxtaur.com> References: <199902242223.OAA02940@cesium.transmeta.com> <19990224232219.B5180@fuzzy.foxtaur.com> Notfrom: spamlog@atnf.csiro.au Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: Fuzzy Fox writes: > H. Peter Anvin wrote: > > > > The problem is that almost everything that browses does both a > > readdir() and a stat(); the readdir() is easy but keeping the stat() > > from mounting the filesystem is really, really hard... > > In my experience with other automounters, that's how they work. If you > stat() the mount-point, the filesystem is mounted. I don't see it as a > problem; just the cost of doing business. :) There may be no need to worry. This trickery with stat(2) is only an issue if the mount point appears in a readdir(2), which in turn would happen if you hacked around with symlinks. If "fakes" are used, we can get around that. Alternatively, if overloading the dentry lookup method is used, it may also not be a problem. Regards, Richard.... From owner-autofs@linux.kernel.org Wed Feb 24 22:42:38 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id WAA29578 for ; Wed, 24 Feb 1999 22:42:37 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id VAA06281 for autofs-list; Wed, 24 Feb 1999 21:51:13 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id VAA06278 for ; Wed, 24 Feb 1999 21:51:11 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id VAA21778; Wed, 24 Feb 1999 21:50:04 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id VAA18998; Wed, 24 Feb 1999 21:50:03 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id VAA31379; Wed, 24 Feb 1999 21:50:00 -0800 Message-Id: <199902250550.VAA31379@cesium.transmeta.com> Subject: Re: direct map In-Reply-To: <199902250538.QAA04777@vindaloo.atnf.CSIRO.AU> from Richard Gooch at "Feb 25, 99 04:38:06 pm" To: rgooch@atnf.CSIRO.AU (Richard Gooch) Date: Wed, 24 Feb 1999 21:50:00 -0800 (PST) Cc: fox@dallas.net, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: RO X-Status: > Fuzzy Fox writes: > > H. Peter Anvin wrote: > > > > > > The problem is that almost everything that browses does both a > > > readdir() and a stat(); the readdir() is easy but keeping the stat() > > > from mounting the filesystem is really, really hard... > > > > In my experience with other automounters, that's how they work. If you > > stat() the mount-point, the filesystem is mounted. I don't see it as a > > problem; just the cost of doing business. :) > > There may be no need to worry. This trickery with stat(2) is only an > issue if the mount point appears in a readdir(2), which in turn would > happen if you hacked around with symlinks. If "fakes" are used, we can > get around that. Alternatively, if overloading the dentry lookup > method is used, it may also not be a problem. > Well, it's definitely something to think about. -hpa From owner-autofs@linux.kernel.org Thu Feb 25 00:17:12 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id AAA29710 for ; Thu, 25 Feb 1999 00:17:11 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id XAA10498 for autofs-list; Wed, 24 Feb 1999 23:17:19 -0800 Received: from namib.north.de (root@namib.North.DE [192.109.226.18]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id XAA10493 for ; Wed, 24 Feb 1999 23:17:16 -0800 Received: from artcom0.UUCP (uucp@localhost) by namib.north.de (8.8.8/8.8.8/Debian/GNU) with UUCP id HAA23049; Thu, 25 Feb 1999 07:57:26 +0100 Received: by artcom0.north.de (Smail3.2 #1) id m10Fbon-000txiC; Wed, 24 Feb 1999 11:45:29 +0100 (MET) Message-Id: From: pf@artcom0.north.de (Peter Funk) Subject: Re: autofs v4: speak now or forever hold your peace In-Reply-To: from Frithjof Anders at "Feb 23, 99 08:55:35 am" To: frithjof@fkp.tu-darmstadt.de Date: Wed, 24 Feb 1999 11:45:28 +0100 (MET) Cc: autofs@linux.kernel.org, abbe@artcom-gmbh.de Organization: privat (Mitglied im IN e.V.) Phone: +49 4222 950270 X-Fax: +49 4222 950260 X-Commercial-ReplyTo: pf@artcom-gmbh.de (Peter Funk) X-Mailer: ELM [version 2.4ME+ PL37 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Hello! Frithjof Anders wrote: > Sorry, I do not see the problem. Maybe I do misunderstand, what you want. But I > had the same problem, to mount a tree of directories from a machine, without > the user guessing + plus the mechnanis should work for any host whithout > changing the automounter maps. Here is my solution within the existing autofs. > > 1) I reorganized all my nfs servers to have mount points like > /hostname/data1 > /hostname/data2 > > 2) I created an automounter moint point /imports > and a map auto.imports > * &:/& > > That was all. A simple universal map, whcih needs no change, if you buy a new > machine. We make local temp space for calcuation intensive programs availible > that way to all hosts of the cluster. That works indeed fine, if you have _ONLY_ Linux NFS servers in your net. Unfortunately we have some servers still running Solaris or IRIX. These machines also have more than one single /-filesystem. With the approach above you can't access the additional filesystems on the Solaris/IRIX boxes. So I have to wait what will happen first: 1. I can convince our people to replace IRIX 6.5/Solaris with MIPS/Sparc Linux on our Origin 2000 Server and Sparc systems. (Relatively unlikely :-( ) 2. Those Servers aren't used any more, since our new Dual P-II PCs run every of our applications at least twice as fast under Linux. (More likely) 3. HPA finishes autofs v4.x which may (will?) support this difficult stuff. (Don't know how likely this is ... hint hint hint ;-) > cd /imports/hostname will mount the whole tree of the other machine and > thanks to how LINUX handels exports, I even can dynamicaly mount and > unmount directories on the server, which are propagated to the clients. Indeed this works fine. As long as you don't need to access data on a None-Linux NFS server... :-( Regards, Peter -- Peter Funk, Oldenburger Str.86, 27777 Ganderkesee, Tel: 04222 9502 70, Fax: -60 From owner-autofs@linux.kernel.org Thu Feb 25 01:00:16 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA29765 for ; Thu, 25 Feb 1999 01:00:16 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA12458 for autofs-list; Thu, 25 Feb 1999 00:05:11 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA12452 for ; Thu, 25 Feb 1999 00:05:06 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id JAA14307; Thu, 25 Feb 1999 09:05:05 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id JAA09145; Thu, 25 Feb 1999 09:05:04 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902242246.OAA13893@leopard.imedia.com> Date: Thu, 25 Feb 1999 09:05:04 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: Alvin Oga Subject: Re: autofs v4: speak now .... documents Cc: autofs@linux.kernel.org, (Alvin Oga) , (Patrick Goetz) Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke Alvin Oga on 24-Feb-99 : > > hi ya all... > > yes people write docs when time is available or have a need to do so... > > and if there is stuff you'd like to add for the autofs howto.... > please let me know and I will put it in... > > thank you > alvin > http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html Since Alvin is doing such a nice job to write a documentation to autofs and Peter is doing an even greater job in writing clean and fast autofs code, why shouldn´t these effords be joined by packing Alvin´s *.html files, or at least a latex version of the manual into the autofs distribution? Peter, is that possible? Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Thu Feb 25 01:07:52 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA29788 for ; Thu, 25 Feb 1999 01:07:52 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA13023 for autofs-list; Thu, 25 Feb 1999 00:18:31 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA13019 for ; Thu, 25 Feb 1999 00:18:27 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id AAA23058; Thu, 25 Feb 1999 00:18:08 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id AAA26164; Thu, 25 Feb 1999 00:18:08 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id AAA32430; Thu, 25 Feb 1999 00:18:08 -0800 Message-Id: <199902250818.AAA32430@cesium.transmeta.com> Subject: Re: autofs v4: speak now .... documents In-Reply-To: from Frithjof Anders at "Feb 25, 99 09:05:04 am" To: frithjof@fkp.tu-darmstadt.de Date: Thu, 25 Feb 1999 00:18:08 -0800 (PST) Cc: aoga@imedia.com, autofs@linux.kernel.org, aoga@leopard.imedia.com, pgoetz@fireant.ma.utexas.edu X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > Thus spoke Alvin Oga on 24-Feb-99 : > > > > hi ya all... > > > > yes people write docs when time is available or have a need to do so... > > > > and if there is stuff you'd like to add for the autofs howto.... > > please let me know and I will put it in... > > > > thank you > > alvin > > http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html > > Since Alvin is doing such a nice job to write a documentation to autofs > and Peter is doing an even greater job in writing clean and fast autofs > code, why shouldn´t these effords be joined by packing Alvin´s *.html > files, or at least a latex version of the manual into the autofs distribution? > > Peter, is that possible? > Yes; I'm going to put in into the next release. -hpa From owner-autofs@linux.kernel.org Thu Feb 25 01:28:12 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA29819 for ; Thu, 25 Feb 1999 01:28:11 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA13644 for autofs-list; Thu, 25 Feb 1999 00:32:56 -0800 Received: from lec.fkp.physik.tu-darmstadt.de (root@lec.fkp.physik.tu-darmstadt.de [130.83.85.210]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA13640 for ; Thu, 25 Feb 1999 00:32:50 -0800 Received: from cobra.fkp.physik.tu-darmstadt.de (frithjof@cobra.fkp.physik.tu-darmstadt.de [130.83.85.212]) by lec.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) with ESMTP id JAA14359; Thu, 25 Feb 1999 09:31:10 +0100 Received: (from frithjof@localhost) by cobra.fkp.physik.tu-darmstadt.de (8.9.0/8.8.7) id JAA09183; Thu, 25 Feb 1999 09:31:09 +0100 Message-ID: X-Mailer: XFMail 1.2 [p0] on Linux X-Priority: 3 (Normal) Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 In-Reply-To: <199902242142.NAA02542@cesium.transmeta.com> Date: Thu, 25 Feb 1999 09:31:09 +0100 (CET) Reply-To: frithjof@fkp.tu-darmstadt.de From: Frithjof Anders To: "H. Peter Anvin" Subject: Re: shutdown autofs Cc: autofs@linux.kernel.org Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: Thus spoke H. Peter Anvin on 24-Feb-99 : >> On the other hand the AIX automounter allows you to do exactly what I want >> via shared mounts > > AIX doing it is, if anything, a good reason *NOT* to... > Is that an argument or an opinion? I see that most of you guys out there come from the SUN world. I have no experience with Solaris as sys admin, just as user but that was a devastating one. The server chrashed once a week, the NFS went down once a day in the institution, I worked as a postdoc in the US. I ended up doing my calculations on a Pentium 120 MHZ laptop, since a) the laptop had the same speed for my calculations as the SUN server and under LINUX was more reliable :) - it has a nice suspend button, so I do not have to reboot :) Before I worked with DEC Ultrix. This OS had some design flaws for running large applications and my workstation crashed regularly while opening xterms. Back home I was sysadm for a AIX cluster. It was a pain in the butt to admistrate, no user friendlyness, but stability, stability, stability. Uptimes of 3-4 months were normal and reboots usually only for system upgrades and buxfixes neccesary. And our workstations were always 2-3 times faster then SUN´s of other physics departments, I also had a chance to run my code on. >From me stability and computational speed is important, since some of my programs use to run for a week. Excuse my personal remarks, but different people have different experiences and different needs. That makes live fun :) Frithjof "If you see someone without a smile, give him one of yours" Frithjof Anders Institut fuer Festkoerperphysik Technische Universitaet Darmstadt Hochschulstr. 6 64289 Darmstadt, GERMANY Tel +49 (6151) 16-5235 email: frithjof@fkp.tu-darmstadt.de FAX +49 (6151) 16-3681 From owner-autofs@linux.kernel.org Thu Feb 25 01:36:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id BAA29823 for ; Thu, 25 Feb 1999 01:36:07 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id AAA14015 for autofs-list; Thu, 25 Feb 1999 00:41:20 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id AAA14012 for ; Thu, 25 Feb 1999 00:41:19 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id AAA23280; Thu, 25 Feb 1999 00:41:05 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id AAA26638; Thu, 25 Feb 1999 00:41:04 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id AAA32585; Thu, 25 Feb 1999 00:41:04 -0800 Message-Id: <199902250841.AAA32585@cesium.transmeta.com> Subject: Re: shutdown autofs In-Reply-To: from Frithjof Anders at "Feb 25, 99 09:31:09 am" To: frithjof@fkp.tu-darmstadt.de Date: Thu, 25 Feb 1999 00:41:04 -0800 (PST) Cc: hpa@transmeta.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > Thus spoke H. Peter Anvin on 24-Feb-99 : > >> On the other hand the AIX automounter allows you to do exactly what I want > >> via shared mounts > > > > AIX doing it is, if anything, a good reason *NOT* to... > Is that an argument or an opinion? It's a wisecrack :) However, the handful of AIX machines we have on site are definitely nothing but trouble. -hpa From owner-autofs@linux.kernel.org Thu Feb 25 02:04:22 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA29835 for ; Thu, 25 Feb 1999 02:04:22 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA15906 for autofs-list; Thu, 25 Feb 1999 01:17:25 -0800 Received: from namib.north.de (root@namib.North.DE [192.109.226.18]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA15895 for ; Thu, 25 Feb 1999 01:17:12 -0800 Received: from artcom0.UUCP (uucp@localhost) by namib.north.de (8.8.8/8.8.8/Debian/GNU) with UUCP id JAA24387 for linux.kernel.org!autofs; Thu, 25 Feb 1999 09:57:59 +0100 Received: by artcom0.north.de (Smail3.2 #1) id m10FuuY-000twmC; Thu, 25 Feb 1999 08:08:42 +0100 (MET) Message-Id: From: pf@artcom0.north.de (Peter Funk) Subject: Re: autofs v4: speak now or shutdown or not In-Reply-To: <199902232349.PAA18309@cesium.transmeta.com> from "H. Peter Anvin" at "Feb 23, 99 03:49:38 pm" To: autofs@linux.kernel.org Date: Thu, 25 Feb 1999 08:08:41 +0100 (MET) Organization: privat (Mitglied im IN e.V.) Phone: +49 4222 950270 X-Fax: +49 4222 950260 X-Commercial-ReplyTo: pf@artcom-gmbh.de (Peter Funk) X-Mailer: ELM [version 2.4ME+ PL37 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: H. Peter Anvin schrieb: > > yes... if you could...please use some kind of detection schem > > where autofs will refuse to shutdown and just flag that /home > > is in use.... instead of going down anyway...and not be able > > to restart to remount the other dirs that the users now wants > > but can't get since autofs won't come back up without killing > > all the old processes... > > Sure; currently that's what autofs uses if you send it SIGUSR1; would > it make sense if I made it behave that way for SIGTERM as well? No, that doesn't make sense. 'SIGTERM' should terminate automount; that is what a termination signal is supposed to do. May be that people want to change the behaviour of their /etc/rc.d/autofs resp. (here on SuSE Linux) their /sbin/init.d/autofs shell-script used to start/restart/stop the autofs ssub system? The version distributed with SuSE contains two special functions for the SuSE and Debian distributions. What about Red Hat, Caldera and others? Regards from germany, Peter -- Peter Funk, Oldenburger Str.86, 27777 Ganderkesee, Tel: 04222 9502 70, Fax: -60 From owner-autofs@linux.kernel.org Thu Feb 25 02:06:38 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id CAA29839 for ; Thu, 25 Feb 1999 02:06:38 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id BAA15902 for autofs-list; Thu, 25 Feb 1999 01:17:21 -0800 Received: from cyberelk.demon.co.uk (mail@cyberelk.demon.co.uk [194.222.36.72]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id BAA15896 for ; Thu, 25 Feb 1999 01:17:13 -0800 Received: from tim by cyberelk.demon.co.uk with local (Exim 1.92 #1 (Debian)) id 10Fwn7-0003O5-00; Thu, 25 Feb 1999 09:09:09 +0000 Date: Thu, 25 Feb 1999 09:09:08 +0000 (GMT) From: Tim Waugh X-Sender: tim@cyberelk.elk.co.uk To: Jason Lunz cc: autofs@linux.kernel.org Subject: Re: BUG: in 2.2.2, autofs mount of SCSI CD-ROM on aha1542 card locks up machine In-Reply-To: <19990224212728.A500@cc.gatech.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: On Wed, 24 Feb 1999, Jason Lunz wrote: > $ ls /vol/cd > Detected scsi CD-ROM at scsi0, channel 0, id 4, lun 0 > aha1542c: Trying device reset for target 4 > bug: kernel timer added twice at c017ec44 The kernel timer being added twice is a known bug, and I think it's fixed in 2.2.2-ac2 (ftp.linux.org.uk). Tim. */ From owner-autofs@linux.kernel.org Thu Feb 25 16:33:58 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA30460 for ; Thu, 25 Feb 1999 16:33:58 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA22322 for autofs-list; Thu, 25 Feb 1999 15:54:09 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA22310 for ; Thu, 25 Feb 1999 15:54:06 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id PAA15330; Thu, 25 Feb 1999 15:54:03 -0800 From: Alvin Oga Message-Id: <199902252354.PAA15330@leopard.imedia.com> Subject: Re: autofs v4: speak now .... documents To: hpa@transmeta.com (H. Peter Anvin) Date: Thu, 25 Feb 1999 15:54:03 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: <199902250818.AAA32430@cesium.transmeta.com> from "H. Peter Anvin" at Feb 25, 99 00:18:08 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi ya all > > > > Since Alvin is doing such a nice job to write a documentation to autofs > > and Peter is doing an even greater job in writing clean and fast autofs > > code, why shouldn´t these effords be joined by packing Alvin´s *.html > > files, or at least a latex version of the manual into the autofs distribution? > > > > Peter, is that possible? > > > > Yes; I'm going to put in into the next release. So that the new changes gets included, please let me know what you folks want added/or changed/patched etc... the other docs, references are in that same directory thank you alvin http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html From owner-autofs@linux.kernel.org Thu Feb 25 16:34:02 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA30464 for ; Thu, 25 Feb 1999 16:34:02 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA22860 for autofs-list; Thu, 25 Feb 1999 16:04:05 -0800 Received: from leopard.imedia.com (leopard.imedia.com [206.3.97.120]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA22855 for ; Thu, 25 Feb 1999 16:04:03 -0800 Received: (from aoga@localhost) by leopard.imedia.com (8.8.7/8.7.3) id QAA15390; Thu, 25 Feb 1999 16:03:52 -0800 From: Alvin Oga Message-Id: <199902260003.QAA15390@leopard.imedia.com> Subject: Re: autofs v4: speak now or shutdown or not To: pf@artcom0.north.de (Peter Funk) Date: Thu, 25 Feb 1999 16:03:52 -0800 (PST) Cc: aoga@leopard.imedia.com (Alvin Oga), autofs@linux.kernel.org In-Reply-To: from "Peter Funk" at Feb 25, 99 08:08:41 am Content-Type: text Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: hi peter f. > > > yes... if you could...please use some kind of detection schem > > > where autofs will refuse to shutdown and just flag that /home > > > is in use.... instead of going down anyway...and not be able > > > to restart to remount the other dirs that the users now wants > > > but can't get since autofs won't come back up without killing > > > all the old processes... > > > > Sure; currently that's what autofs uses if you send it SIGUSR1; would > > it make sense if I made it behave that way for SIGTERM as well? > > No, that doesn't make sense. 'SIGTERM' should terminate automount; > that is what a termination signal is supposed to do. yes.... sigterm should kill the process... am looking for another sig to send it to try to kill but don't if it is used...since autofs does not restart if it did not "cleanly die" > May be that people want to change the behaviour of their > /etc/rc.d/autofs resp. (here on SuSE Linux) their /sbin/init.d/autofs > shell-script used to start/restart/stop the autofs ssub system? > The version distributed with SuSE contains two special functions > for the SuSE and Debian distributions. What about Red Hat, Caldera > and others? attached below for fun is the rh-5.x /etc/rc.d/init.d/autofs script ( too big, guess I'll put it up on the website instead... if you care to see it ) http://www.Linux-Consulting.com/Amd_AutoFS/init.d_autofs note that it uses SIGTERM ( stop ) and SIGUSR2 ( reload ) have fun alvin From owner-autofs@linux.kernel.org Thu Feb 25 16:34:21 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA30468 for ; Thu, 25 Feb 1999 16:34:20 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id PAA22507 for autofs-list; Thu, 25 Feb 1999 15:59:12 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id PAA22503 for ; Thu, 25 Feb 1999 15:59:11 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id PAA03143; Thu, 25 Feb 1999 15:59:10 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id PAA19439; Thu, 25 Feb 1999 15:59:10 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id PAA23519; Thu, 25 Feb 1999 15:59:09 -0800 Message-Id: <199902252359.PAA23519@cesium.transmeta.com> Subject: Re: autofs v4: speak now .... documents In-Reply-To: <199902252354.PAA15330@leopard.imedia.com> from Alvin Oga at "Feb 25, 99 03:54:03 pm" To: aoga@imedia.com (Alvin Oga) Date: Thu, 25 Feb 1999 15:59:09 -0800 (PST) Cc: autofs@linux.kernel.org (autofs mailing list) X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > Peter, is that possible? > > > > > > > Yes; I'm going to put in into the next release. > > So that the new changes gets included, > please let me know what you folks want added/or > changed/patched etc... > > the other docs, references are in that same directory > > thank you > alvin > http://www.Linux-Consulting.com/Amd_AutoFS/autofs-HOWTO.html Will do. For what it's worth, I discussed a few options for supporting direct mounts (based on Richard's suggestion) with Linus today, and I might actually be able to get a VFS change into 2.3 that would let us support direct mounts; we're evaluating a few methods (including one option of supporting sandwich mounts that hadn't been considered before.) Please don't get your hopes too far up, but there is at least a possibility that direct mounts might be supportable after all. -hpa From owner-autofs@linux.kernel.org Thu Feb 25 16:36:29 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA30475 for ; Thu, 25 Feb 1999 16:36:29 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA23383 for autofs-list; Thu, 25 Feb 1999 16:12:29 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA23380 for ; Thu, 25 Feb 1999 16:12:28 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id QAA03374; Thu, 25 Feb 1999 16:12:21 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id QAA20751; Thu, 25 Feb 1999 16:12:20 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id QAA23728; Thu, 25 Feb 1999 16:12:20 -0800 Message-Id: <199902260012.QAA23728@cesium.transmeta.com> Subject: Re: autofs v4: speak now or shutdown or not In-Reply-To: <199902260007.QAA23669@cesium.transmeta.com> from "H. Peter Anvin" at "Feb 25, 99 04:07:56 pm" To: hpa@transmeta.com (H. Peter Anvin) Date: Thu, 25 Feb 1999 16:12:20 -0800 (PST) Cc: aoga@imedia.com, pf@artcom0.north.de, aoga@leopard.imedia.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > > > > No, that doesn't make sense. 'SIGTERM' should terminate automount; > > > that is what a termination signal is supposed to do. > > > > yes.... sigterm should kill the process... > > > > am looking for another sig to send it to try to kill but don't if it > > is used...since autofs does not restart if it did not "cleanly die" > > > > Okay, so right now we have: > > SIGKILL - Die instantly (leaves a mess) > SIGTERM - Try to shut down cleanly but die no matter what > SIGUSR2 - Try to shut down cleanly but continue if not possible SIGHUP - Prune as many mounts as possible, and then reinitialize the daemon > SIGUSR1 - Prune as many mounts as possible, but don't > shut down > -hpa From owner-autofs@linux.kernel.org Thu Feb 25 16:38:21 1999 Return-Path: Received: from linux.kernel.org (linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id QAA30479 for ; Thu, 25 Feb 1999 16:38:21 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA23118 for autofs-list; Thu, 25 Feb 1999 16:08:06 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA23106 for ; Thu, 25 Feb 1999 16:08:04 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id QAA03286; Thu, 25 Feb 1999 16:07:57 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id QAA20279; Thu, 25 Feb 1999 16:07:56 -0800 (PST) From: "H. Peter Anvin" Received: (from hpa@localhost) by cesium.transmeta.com (8.8.4/8.7.3) id QAA23669; Thu, 25 Feb 1999 16:07:56 -0800 Message-Id: <199902260007.QAA23669@cesium.transmeta.com> Subject: Re: autofs v4: speak now or shutdown or not In-Reply-To: <199902260003.QAA15390@leopard.imedia.com> from Alvin Oga at "Feb 25, 99 04:03:52 pm" To: aoga@imedia.com (Alvin Oga) Date: Thu, 25 Feb 1999 16:07:56 -0800 (PST) Cc: pf@artcom0.north.de, aoga@leopard.imedia.com, autofs@linux.kernel.org X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: > > > > No, that doesn't make sense. 'SIGTERM' should terminate automount; > > that is what a termination signal is supposed to do. > > yes.... sigterm should kill the process... > > am looking for another sig to send it to try to kill but don't if it > is used...since autofs does not restart if it did not "cleanly die" > Okay, so right now we have: SIGKILL - Die instantly (leaves a mess) SIGTERM - Try to shut down cleanly but die no matter what SIGUSR2 - Try to shut down cleanly but continue if not possible SIGUSR1 - Prune as many mounts as possible, but don't shut down -hpa From owner-autofs@linux.kernel.org Thu Feb 25 17:13:32 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA30499 for ; Thu, 25 Feb 1999 17:13:32 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA25388 for autofs-list; Thu, 25 Feb 1999 16:49:05 -0800 Received: from vindaloo.atnf.CSIRO.AU (vindaloo.atnf.CSIRO.AU [130.155.198.47]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id QAA25383 for ; Thu, 25 Feb 1999 16:49:01 -0800 Received: (from rgooch@localhost) by vindaloo.atnf.CSIRO.AU (8.6.12/8.6.12) id LAA13417; Fri, 26 Feb 1999 11:48:51 +1100 Date: Fri, 26 Feb 1999 11:48:51 +1100 Message-Id: <199902260048.LAA13417@vindaloo.atnf.CSIRO.AU> From: Richard Gooch To: "H. Peter Anvin" Cc: autofs@linux.kernel.org (autofs mailing list) Subject: Re: autofs v4: speak now .... documents In-Reply-To: <199902252359.PAA23519@cesium.transmeta.com> References: <199902252354.PAA15330@leopard.imedia.com> <199902252359.PAA23519@cesium.transmeta.com> Notfrom: spamlog@atnf.csiro.au Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: H. Peter Anvin writes: > Will do. For what it's worth, I discussed a few options for > supporting direct mounts (based on Richard's suggestion) with Linus > today, and I might actually be able to get a VFS change into 2.3 > that would let us support direct mounts; we're evaluating a few > methods (including one option of supporting sandwich mounts that > hadn't been considered before.) Please don't get your hopes too far > up, but there is at least a possibility that direct mounts might be > supportable after all. Could you explain what you have in mind for these sandwich mounts? How would they work? Regards, Richard.... From owner-autofs@linux.kernel.org Thu Feb 25 17:22:08 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA30546 for ; Thu, 25 Feb 1999 17:22:07 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA25819 for autofs-list; Thu, 25 Feb 1999 16:56:11 -0800 Received: from vindaloo.atnf.CSIRO.AU (vindaloo.atnf.CSIRO.AU [130.155.198.47]) by linux.kernel.org (8.9.1/8.9.1) with SMTP id QAA25816 for ; Thu, 25 Feb 1999 16:56:07 -0800 Received: (from rgooch@localhost) by vindaloo.atnf.CSIRO.AU (8.6.12/8.6.12) id LAA13621; Fri, 26 Feb 1999 11:55:58 +1100 Date: Fri, 26 Feb 1999 11:55:58 +1100 Message-Id: <199902260055.LAA13621@vindaloo.atnf.CSIRO.AU> From: Richard Gooch To: "H. Peter Anvin" Cc: autofs@linux.kernel.org Subject: Re: autofs v4: speak now .... documents In-Reply-To: <199902260052.QAA06637@cesium.transmeta.com> References: <199902260048.LAA13417@vindaloo.atnf.CSIRO.AU> <199902260052.QAA06637@cesium.transmeta.com> Notfrom: spamlog@atnf.csiro.au Sender: owner-autofs@linux.kernel.org Precedence: bulk Status: O X-Status: H. Peter Anvin writes: > > H. Peter Anvin writes: > > > Will do. For what it's worth, I discussed a few options for > > > supporting direct mounts (based on Richard's suggestion) with Linus > > > today, and I might actually be able to get a VFS change into 2.3 > > > that would let us support direct mounts; we're evaluating a few > > > methods (including one option of supporting sandwich mounts that > > > hadn't been considered before.) Please don't get your hopes too far > > > up, but there is at least a possibility that direct mounts might be > > > supportable after all. > > > > Could you explain what you have in mind for these sandwich mounts? How > > would they work? > > Basically, I argued with Linus for letting me add another pointer to > the dentry. Instead of d_mounts and d_covers we would have a d_top, > d_bottom and d_down pointers. This means you still don't need any > tests: d_top will always take you to the topmost dentry (like the > current d_mounts) and d_bottom will take you to the bottommost entry > (like the current d_covers). The d_down pointers would form a > linked list of entries to be popped off by umount. Brilliant. I wish I'd thought of it ;-) That's much better than hacking around with fakes. Regards, Richard.... From owner-autofs@linux.kernel.org Thu Feb 25 17:23:05 1999 Return-Path: Received: from linux.kernel.org (majordom@linux.kernel.org [206.184.214.34]) by mail.Linux-Consulting.com (8.8.7/8.8.7) with ESMTP id RAA30551 for ; Thu, 25 Feb 1999 17:23:05 -0800 Received: (from majordom@localhost) by linux.kernel.org (8.9.1/8.9.1) id QAA25617 for autofs-list; Thu, 25 Feb 1999 16:53:02 -0800 Received: from neon.transmeta.com (neon-best.transmeta.com [206.184.214.10]) by linux.kernel.org (8.9.1/8.9.1) with ESMTP id QAA25614 for ; Thu, 25 Feb 1999 16:53:01 -0800 Received: from deepthought.transmeta.com (mailhost.transmeta.com [10.1.1.15]) by neon.transmeta.com (8.9.1/8.9.1) with ESMTP id QAA04015; Thu, 25 Feb 1999 16:52:25 -0800 Received: from cesium.transmeta.com (hpa@cesium.transmeta.com [10.1.2.55]) by deepthought.transmeta.com (8.8.8+spamcan/8.8.5) with ESMTP id QAA238