OpenBSD-4.6/usr.sbin/amd/doc/amdref.texinfo
\input texinfo @c -*-texinfo-*-
@c
@c Copyright (c) 1989 Jan-Simon Pendry
@c Copyright (c) 1989 Imperial College of Science, Technology & Medicine
@c Copyright (c) 1989 The Regents of the University of California.
@c All rights reserved.
@c
@c This code is derived from software contributed to Berkeley by
@c Jan-Simon Pendry at Imperial College, London.
@c
@c Redistribution and use in source and binary forms, with or without
@c modification, are permitted provided that the following conditions
@c are met:
@c 1. Redistributions of source code must retain the above copyright
@c notice, this list of conditions and the following disclaimer.
@c 2. Redistributions in binary form must reproduce the above copyright
@c notice, this list of conditions and the following disclaimer in the
@c documentation and/or other materials provided with the distribution.
@c 3. All advertising materials mentioning features or use of this software
@c must display the following acknowledgement:
@c This product includes software developed by the University of
@c California, Berkeley and its contributors.
@c 4. Neither the name of the University nor the names of its contributors
@c may be used to endorse or promote products derived from this software
@c without specific prior written permission.
@c
@c THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
@c ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
@c IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
@c ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
@c FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
@c DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
@c OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
@c HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
@c LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
@c OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
@c
@c from: @(#)amdref.texinfo 8.1 (Berkeley) 6/6/93
@c $Id: amdref.texinfo,v 1.11 2004/02/19 23:40:38 deraadt Exp $
@c
@setfilename amdref.info
@c @setfilename /usr/local/emacs/info/amd
@tex
\overfullrule=0pt
@end tex
@settitle 4.4 BSD Automounter Reference Manual
@dircategory Miscellaneous
@direntry
* Amdref: (amdref). 4.4 BSD Automounter Reference Manual
@end direntry
@titlepage
@sp 6
@center @titlefont{Amd}
@sp 2
@center @titlefont{The 4.4 BSD Automounter}
@sp 2
@center @titlefont{Reference Manual}
@sp 2
@center @authorfont{Jan-Simon Pendry}
@sp
@center @i{and}
@sp
@center @authorfont{Nick Williams}
@sp 4
@center Last updated March 1991
@center Documentation for software revision 5.3 Alpha
@page
Copyright @copyright{} 1989 Jan-Simon Pendry
@sp -1
Copyright @copyright{} 1989 Imperial College of Science, Technology & Medicine
@sp -1
Copyright @copyright{} 1989 The Regents of the University of California.
@sp 0
All Rights Reserved.
@vskip 1ex
Permission to copy this document, or any portion of it, as
necessary for use of this software is granted provided this
copyright notice and statement of permission are included.
@end titlepage
@page
@ifinfo
@node Top, License, , (DIR)
Amd - The 4.4 BSD Automounter
*****************************
Amd is the 4.4 BSD Automounter. This Info file describes how
to use and understand Amd.
@end ifinfo
@menu
* License:: Explains the terms and conditions for using
and distributing Amd.
* Distrib:: How to get the latest Amd distribution.
* Intro:: An introduction to Automounting concepts.
* Overview:: An overview of Amd.
* Supported Platforms:: Machines and Systems supported by Amd.
* Mount Maps:: Details of mount maps
* Amd Command Line Options:: All the Amd command line options explained.
* Filesystem Types:: The different mount types supported by Amd.
* Run-time Administration:: How to start, stop and control Amd.
* Examples:: Some examples showing how Amd might be used.
* Internals:: Implementation details.
* Acknowledgements & Trademarks:: Legal notes.
Indexes
* Index:: An item for each concept.
@end menu
@iftex
@unnumbered Preface
This manual documents the use of the 4.4 BSD automounter---@i{Amd}.
This is primarily a reference manual. Unfortunately, no tutorial
exists.
This manual comes in two forms: the published form and the Info form.
The Info form is for on-line perusal with the INFO program which is
distributed along with GNU Emacs. Both forms contain substantially the
same text and are generated from a common source file, which is
distributed with the @i{Amd} source.
@end iftex
@node License, Distrib, Top, Top
@unnumbered License
@cindex License Information
@i{Amd} is not in the public domain; it is copyrighted and there are
restrictions on its distribution.
Redistribution and use in source and binary forms are permitted provided
that: (1) source distributions retain this entire copyright notice and
comment, and (2) distributions including binaries display the following
acknowledgement: ``This product includes software developed by The
University of California, Berkeley and its Contributors'' in the
documentation or other materials provided with the distribution and in
all advertising materials mentioning features or use of this software.
neither the name of the University nor the names of its Contributors may
be used to endorse or promote products derived from this software
without specific prior written permission.
THIS SOFTWARE IS PROVIDED ``AS IS'' AND WITHOUT ANY EXPRESS OR IMPLIED
WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
@node Distrib, Intro, License, Top
@unnumbered Source Distribution
@cindex Source code distribution
@cindex Obtaining the source code
If you have access to the Internet, you can get the latest distribution
version of @i{Amd} from host @file{usc.edu} using anonymous FTP. Move to
the directory @file{/pub/amd} on that host and fetch the file @file{amd.tar.Z}.
If you are in the UK, you can get the latest distribution version of
@i{Amd} from the UKnet info-server. Start by sending email to
@file{info-server@@doc.ic.ac.uk}.
Sites on the UK JANET network can get the latest distribution by using
anonymous NIFTP to fetch the file @samp{<AMD>amd.tar.Z} from host
@samp{uk.ac.imperial.doc.src}.
Revision 5.2 was part of the 4.3 BSD Reno distribution.
Revision 5.3bsdnet, a late alpha version of 5.3, was part
of the BSD network version 2 distribution
@unnumberedsec Bug Reports
@cindex Bug reports
Send all bug reports to @file{jsp@@doc.ic.ac.uk} quoting the details of
the release and your configuration. These can be obtained by running
the command @samp{amd -v}.
@unnumberedsec Mailing List
@cindex Mailing list
There is a mailing list for people interested in keeping uptodate with
developments. To subscribe, send a note to @file{amd-workers-request@@acl.lanl.gov}.
@node Intro, Overview, Distrib, Top
@unnumbered Introduction
@cindex Introduction
An @dfn{automounter} maintains a cache of mounted filesystems.
Filesystems are mounted on demand when they are first referenced,
and unmounted after a period of inactivity.
@i{Amd} may be used as a replacement for Sun's automounter. The choice
of which filesystem to mount can be controlled dynamically with
@dfn{selectors}. Selectors allow decisions of the form ``hostname is
@var{this},'' or ``architecture is not @var{that}.'' Selectors may be
combined arbitrarily. @i{Amd} also supports a variety of filesystem
types, including NFS, UFS and the novel @dfn{program} filesystem. The
combination of selectors and multiple filesystem types allows identical
configuration files to be used on all machines so reducing the
administrative overhead.
@i{Amd} ensures that it will not hang if a remote server goes down.
Moreover, @i{Amd} can determine when a remote server has become
inaccessible and then mount replacement filesystems as and when they
become available.
@i{Amd} contains no proprietary source code and has been ported to
numerous flavours of Unix.
@node Overview, Supported Platforms, Intro, Top
@chapter Overview
@i{Amd} maintains a cache of mounted filesystems. Filesystems are
@dfn{demand-mounted} when they are first referenced, and unmounted after
a period of inactivity. @i{Amd} may be used as a replacement for Sun's
@b{automount}(8) program. It contains no proprietary source code and
has been ported to numerous flavours of Unix. @xref{Supported Operating
Systems}.@refill
@i{Amd} was designed as the basis for experimenting with filesystem
layout and management. Although @i{Amd} has many direct applications it
is loaded with additional features which have little practical use. At
some point the infrequently used components may be removed to streamline
the production system.
@c @i{Amd} supports the notion of @dfn{replicated} filesystems by evaluating
@c each member of a list of possible filesystem locations in parallel.
@c @i{Amd} checks that each cached mapping remains valid. Should a mapping be
@c lost -- such as happens when a fileserver goes down -- @i{Amd} automatically
@c selects a replacement should one be available.
@c
@menu
* Fundamentals::
* Filesystems and Volumes::
* Volume Naming::
* Volume Binding::
* Operational Principles::
* Mounting a Volume::
* Automatic Unmounting::
* Keep-alives::
* Non-blocking Operation::
@end menu
@node Fundamentals, Filesystems and Volumes, Overview, Overview
@comment node-name, next, previous, up
@section Fundamentals
@cindex Automounter fundamentals
The fundamental concept behind @i{Amd} is the ability to separate the
name used to refer to a file from the name used to refer to its physical
storage location. This allows the same files to be accessed with the
same name regardless of where in the network the name is used. This is
very different from placing @file{/n/hostname} in front of the pathname
since that includes location dependent information which may change if
files are moved to another machine.
By placing the required mappings in a centrally administered database,
filesystems can be re-organised without requiring changes to
configuration files, shell scripts and so on.
@node Filesystems and Volumes, Volume Naming, Fundamentals, Overview
@comment node-name, next, previous, up
@section Filesystems and Volumes
@cindex Filesystem
@cindex Volume
@cindex Fileserver
@cindex sublink
@i{Amd} views the world as a set of fileservers, each containg one or
more filesystems where each filesystem contains one or more
@dfn{volumes}. Here the term @dfn{volume} is used to refer to a
coherent set of files such as a user's home directory or a @TeX{}
distribution.@refill
In order to access the contents of a volume, @i{Amd} must be told in
which filesystem the volume resides and which host owns the filesystem.
By default the host is assumed to be local and the volume is assumed to
be the entire filesystem. If a filesystem contains more than one
volume, then a @dfn{sublink} is used to refer to the sub-directory
within the filesystem where the volume can be found.
@node Volume Naming, Volume Binding, Filesystems and Volumes, Overview
@comment node-name, next, previous, up
@section Volume Naming
@cindex Volume names
@cindex Network-wide naming
@cindex Replicated volumes
@cindex Duplicated volumes
@cindex Replacement volumes
Volume names are defined to be unique across the entire network. A
volume name is the pathname to the volume's root as known by the users
of that volume. Since this name uniquely identifies the volume
contents, all volumes can be named and accessed from each host, subject
to administrative controls.
Volumes may be replicated or duplicated. Replicated volumes contain
identical copies of the same data and reside at two or more locations in
the network. Each of the replicated volumes can be used
interchangeably. Duplicated volumes each have the same name but contain
different, though functionally identical, data. For example,
@samp{/vol/tex} might be the name of a @TeX{} distribution which varied
for each machine architecture.@refill
@i{Amd} provides facilities to take advantage of both replicated and
duplicated volumes. Configuration options allow a single set of
configuration data to be shared across an entire network by taking
advantage of replicated and duplicated volumes.
@i{Amd} can take advantage of replacement volumes by mounting them as
required should an active fileserver become unavailable.
@node Volume Binding, Operational Principles, Volume Naming, Overview
@comment node-name, next, previous, up
@section Volume Binding
@cindex Volume binding
@cindex Unix namespace
@cindex Namespace
@cindex Binding names to filesystems
Unix implements a namespace of hierarchically mounted filesystems. Two
forms of binding between names and files are provided. A @dfn{hard
link} completes the binding when the name is added to the filesystem. A
@dfn{soft link} delays the binding until the name is accessed. An
@dfn{automounter} adds a further form in which the binding of name to
filesystem is delayed until the name is accessed.@refill
The target volume, in its general form, is a tuple (host, filesystem,
sublink) which can be used to name the physical location of any volume
in the network.
When a target is referenced, @i{Amd} ignores the sublink element and
determines whether the required filesystem is already mounted. This is
done by computing the local mount point for the filesystem and checking
for an existing filesystem mounted at the same place. If such a
filesystem already exists then it is assumed to be functionally
identical to the target filesystem. By default there is a one-to-one
mapping between the pair (host, filesystem) and the local mount point so
this assumption is valid.
@node Operational Principles, Mounting a Volume, Volume Binding, Overview
@comment node-name, next, previous, up
@section Operational Principles
@cindex Operational principles
@i{Amd} operates by introducing new mount points into the namespace.
These are called @dfn{automount} points. The kernel sees these
automount points as NFS filesystems being served by @i{Amd}. Having
attached itself to the namespace, @i{Amd} is now able to control the
view the rest of the system has of those mount points. RPC calls are
received from the kernel one at a time.
When a @dfn{lookup} call is received @i{Amd} checks whether the name is
already known. If it is not, the required volume is mounted. A
symbolic link pointing to the volume root is then returned. Once the
symbolic link is returned, the kernel will send all other requests
direct to the mounted filesystem.
If a volume is not yet mounted, @i{Amd} consults a configuration
@dfn{mount-map} corresponding to the automount point. @i{Amd} then
makes a runtime decision on what and where to mount a filesystem based
on the information obtained from the map.
@i{Amd} does not implement all the NFS requests; only those relevant
to name binding such as @dfn{lookup}, @dfn{readlink} and @dfn{readdir}.
Some other calls are also implemented but most simply return an error
code; for example @dfn{mkdir} always returns ``read-only filesystem''.
@node Mounting a Volume, Automatic Unmounting, Operational Principles, Overview
@comment node-name, next, previous, up
@section Mounting a Volume
@cindex Mounting a volume
@cindex Location lists
@cindex Alternate locations
@cindex Mount retries
@cindex Background mounts
Each automount point has a corresponding mount map. The mount map
contains a list of key--value pairs. The key is the name of the volume
to be mounted. The value is a list of locations describing where the
filesystem is stored in the network. In the source for the map the
value would look like
@display
location1 location2 @dots{} locationN
@end display
@i{Amd} examines each location in turn. Each location may contain
@dfn{selectors} which control whether @i{Amd} can use that location.
For example, the location may be restricted to use by certain hosts.
Those locations which cannot be used are ignored.
@i{Amd} attempts to mount the filesystem described by each remaining
location until a mount succeeds or @i{Amd} can no longer proceed. The
latter can occur in three ways:
@itemize @bullet
@item
If none of the locations could be used, or if all of the locations
caused an error, then the last error is returned.
@item
If a location could be used but was being mounted in the background then
@i{Amd} marks that mount as being ``in progress'' and continues with
the next request; no reply is sent to the kernel.
@item
Lastly, one or more of the mounts may have been @dfn{deferred}. A mount
is deferred if extra information is required before the mount can
proceed. When the information becomes available the mount will take
place, but in the mean time no reply is sent to the kernel. If the
mount is deferred, @i{Amd} continues to try any remaining locations.
@end itemize
Once a volume has been mounted, @i{Amd} establishes a @dfn{volume
mapping} which is used to satisfy subsequent requests.@refill
@node Automatic Unmounting, Keep-alives, Mounting a Volume, Overview
@comment node-name, next, previous, up
@section Automatic Unmounting
To avoid an ever increasing number of filesystem mounts, @i{Amd} removes
volume mappings which have not been used recently. A time-to-live
interval is associated with each mapping and when that expires the
mapping is removed. When the last reference to a filesystem is removed,
that filesystem is unmounted. If the unmount fails, for example the
filesystem is still busy, the mapping is re-instated and its
time-to-live interval is extended. The global default for this grace
period is controlled by the ``-w'' command-line option (@pxref{-w
Option, -w}). It is also possible to set this value on a per-mount
basis (@pxref{opts Option, opts, opts}).@refill
Filesystems can be forcefully timed out using the @i{Amq} command.
@xref{Run-time Administration}.
@node Keep-alives, Non-blocking Operation, Automatic Unmounting, Overview
@comment node-name, next, previous, up
@section Keep-alives
@cindex Keep-alives
@cindex Server crashes
@cindex NFS ping
Use of some filesystem types requires the presence of a server on
another machine. If a machine crashes then it is of no concern to
processes on that machine that the filesystem is unavailable. However,
to processes on a remote host using that machine as a fileserver this
event is important. This situation is most widely recognised when an
NFS server crashes and the behaviour observed on client machines is that
more and more processes hang. In order to provide the possibility of
recovery, @i{Amd} implements a @dfn{keep-alive} interval timer for some
filesystem types. Currently only NFS makes use of this service.
The basis of the NFS keep-alive implementation is the observation that
most sites maintain replicated copies of common system data such as
manual pages, most or all programs, system source code and so on. If
one of those servers goes down it would be reasonable to mount one of
the others as a replacement.
The first part of the process is to keep track of which fileservers are
up and which are down. @i{Amd} does this by sending RPC requests to the
servers' NFS @code{NullProc} and checking whether a reply is returned.
While the server state is uncertain the requests are re-transmitted at
three second intervals and if no reply is received after four attempts
the server is marked down. If a reply is received the fileserver is
marked up and stays in that state for 30 seconds at which time another
NFS ping is sent.
Once a fileserver is marked down, requests continue to be sent every 30
seconds in order to determine when the fileserver comes back up. During
this time any reference through @i{Amd} to the filesystems on that
server fail with the error ``Operation would block''. If a replacement
volume is available then it will be mounted, otherwise the error is
returned to the user.
@c @i{Amd} keeps track of which servers are up and which are down.
@c It does this by sending RPC requests to the servers' NFS {\sc NullProc} and
@c checking whether a reply is returned. If no replies are received after a
@c short period, @i{Amd} marks the fileserver @dfn{down}.
@c RPC requests continue to be sent so that it will notice when a fileserver
@c comes back up.
@c ICMP echo packets \cite{rfc:icmp} are not used because it is the availability
@c of the NFS service that is important, not the existence of a base kernel.
@c Whenever a reference to a fileserver which is down is made via @i{Amd}, an alternate
@c filesystem is mounted if one is available.
@c
Although this action does not protect user files, which are unique on
the network, or processes which do not access files via @i{Amd} or
already have open files on the hung filesystem, it can prevent most new
processes from hanging.
By default, fileserver state is not maintained for NFS/TCP mounts. The
remote fileserver is always assumed to be up.
@c
@c With a suitable combination of filesystem management and mount-maps,
@c machines can be protected against most server downtime. This can be
@c enhanced by allocating boot-servers dynamically which allows a diskless
@c workstation to be quickly restarted if necessary. Once the root filesystem
@c is mounted, @i{Amd} can be started and allowed to mount the remainder of
@c the filesystem from whichever fileservers are available.
@node Non-blocking Operation, , Keep-alives, Overview
@comment node-name, next, previous, up
@section Non-blocking Operation
@cindex Non-blocking operation
@cindex Multiple-threaded server
@cindex RPC retries
Since there is only one instance of @i{Amd} for each automount point,
and usually only one instance on each machine, it is important that it
is always available to service kernel calls. @i{Amd} goes to great
lengths to ensure that it does not block in a system call. As a last
resort @i{Amd} will fork before it attempts a system call that may block
indefinitely, such as mounting an NFS filesystem. Other tasks such as
obtaining filehandle information for an NFS filesystem, are done using a
purpose built non-blocking RPC library which is integrated with
@i{Amd}'s task scheduler. This library is also used to implement NFS
keep-alives (@pxref{Keep-alives}).
Whenever a mount is deferred or backgrounded, @i{Amd} must wait for it
to complete before replying to the kernel. However, this would cause
@i{Amd} to block waiting for a reply to be constructed. Rather than do
this, @i{Amd} simply @dfn{drops} the call under the assumption that the
kernel RPC mechanism will automatically retry the request.
@node Supported Platforms, Mount Maps, Overview, Top
@comment node-name, next, previous, up
@chapter Supported Platforms
@i{Amd} has been ported to a wide variety of machines and operating systems.
The table below lists those platforms supported by the current release.
@menu
* Supported Operating Systems::
* Supported Machine Architectures::
@end menu
@node Supported Operating Systems, Supported Machine Architectures, Supported Platforms, Supported Platforms
@comment node-name, next, previous, up
@section Supported Operating Systems
@cindex Operating system names
@cindex Operating systems supported by Amd
@cindex Supported operating systems
The following operating systems are currently supported by @i{Amd}.
@i{Amd}'s conventional name for each system is given.
@table @code
@item acis43
4.3 BSD for IBM RT. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item aix3
AIX 3.1. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item aux
System V for Mac-II. Contributed by Julian Onions @t{<jpo@@cs.nott.ac.uk>}
@item bsd44
4.4 BSD. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item concentrix
Concentrix 5.0. Contributed by Sjoerd Mullender @t{<sjoerd@@cwi.nl>}
@item convex
Convex OS 7.1. Contributed by Eitan Mizrotsky @t{<eitan@@shumuji.ac.il>}
@item dgux
Data General DG/UX. Contributed by Mark Davies @t{<mark@@comp.vuw.ac.nz>}
@item fpx4
Celerity FPX 4.1/2. Contributed by Stephen Pope @t{<scp@@grizzly.acl.lanl.gov>}
@item hcx
Harris HCX/UX. Contributed by Chris Metcalf @t{<metcalf@@masala.lcs.mit.edu>}
@item hlh42
HLH OTS 1.@i{x} (4.2 BSD). Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item hpux
HP-UX 6.@i{x} or 7.0. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item irix
SGI Irix. Contributed by Scott R. Presnell @t{<srp@@cgl.ucsf.edu>}
@item next
Mach for NeXT. Contributed by Bill Trost @t{<trost%reed@@cse.ogi.edu>}
@item pyrOSx
Pyramid OSx. Contributed by Stefan Petri @t{<petri@@tubsibr.UUCP>}
@item riscix
Acorn RISC iX. Contributed by Piete Brooks @t{<pb@@cam.cl.ac.uk>}
@item sos3
SunOS 3.4 & 3.5. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item sos4
SunOS 4.@i{x}. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@item u2_2
Ultrix 2.2. Contributed by Piete Brooks @t{<pb@@cam.cl.ac.uk>}
@item u3_0
Ultrix 3. Contributed by Piete Brooks @t{<pb@@cam.cl.ac.uk>}
@item u4_0
Ultrix 4.0. Contributed by Chris Lindblad @t{<cjl@@ai.mit.edu>}
@item umax43
Umax 4.3 BSD. Contributed by Sjoerd Mullender @t{<sjoerd@@cwi.nl>}
@item utek
Utek 4.0. Contributed by Bill Trost @t{<trost%reed@@cse.ogi.edu>}
@item xinu43
mt Xinu MORE/bsd. Contributed by Jan-Simon Pendry @t{<jsp@@doc.ic.ac.uk>}
@end table
@node Supported Machine Architectures, , Supported Operating Systems, Supported Platforms
@comment node-name, next, previous, up
@section Supported Machine Architectures
@cindex Supported machine architectures
@cindex Machine architecture names
@cindex Machine architectures supported by Amd
@table @code
@item alliant
Alliant FX/4
@item arm
Acorn ARM
@item aviion
Data General AViiON
@item encore
Encore
@item fps500
FPS Model 500
@item hp9000
HP 9000/300 family
@item hp9k8
HP 9000/800 family
@item ibm032
IBM RT
@item ibm6000
IBM RISC System/6000
@item iris4d
SGI Iris 4D
@item macII
Apple Mac II
@item mips
MIPS RISC
@item multimax
Encore Multimax
@item orion105
HLH Orion 1/05
@item sun3
Sun-3 family
@item sun4
Sun-4 family
@item tahoe
Tahoe family
@item vax
DEC Vax
@end table
@node Mount Maps, Amd Command Line Options, Supported Platforms, Top
@comment node-name, next, previous, up
@chapter Mount Maps
@cindex Mount maps
@cindex Automounter configuration maps
@cindex Mount information
@i{Amd} has no built-in knowledge of machines or filesystems.
External @dfn{mount-maps} are used to provide the required information.
Specifically, @i{Amd} needs to know when and under what conditions it
should mount filesystems.
The map entry corresponding to the requested name contains a list of
possible locations from which to resolve the request. Each location
specifies filesystem type, information required by that filesystem (for
example the block special device in the case of UFS), and some
information describing where to mount the filesystem (@pxref{fs Option}). A
location may also contain @dfn{selectors} (@pxref{Selectors}).@refill
@menu
* Map Types::
* Key Lookup::
* Location Format::
@end menu
@node Map Types, Key Lookup, Mount Maps, Mount Maps
@comment node-name, next, previous, up
@section Map Types
@cindex Mount map types
@cindex Map types
@cindex Configuration map types
@cindex Types of mount map
@cindex Types of configuration map
@cindex Determining the map type
A mount-map provides the run-time configuration information to @i{Amd}.
Maps can be implemented in many ways. Some of the forms supported by
@i{Amd} are regular files, NIS maps the @dfn{Hesiod} name server and
even the password file.
A mount-map @dfn{name} is a sequence of characters. When an automount
point is created a handle on the mount-map is obtained. For each map
type configured @i{Amd} attempts to reference the a map of the
appropriate type. If a map is found, @i{Amd} notes the type for future
use and deletes the reference, for example closing any open file
descriptors. The available maps are configure when @i{Amd} is built and
can be displayed by running the command @samp{amd -v}.
By default, @i{Amd} caches data in a mode dependent on the type of map.
This is the same as specifying @samp{cache:=mapdefault} and selects a
suitable default cache mode depending on the map type. The individual
defaults are described below. The @var{cache} option can be specified
on automount points to alter the caching behaviour (@pxref{Automount
Filesystem}).@refill
The following map types have been implemented, though some are not
available on all machines. Run the command @samp{amd -v} to obtain a
list of map types configured on your machine.
@menu
* File maps::
* NIS maps::
* Hesiod maps::
* Password maps::
* Union maps::
@end menu
@node File maps, NIS maps, Map Types, Map Types
@comment node-name, next, previous, up
@subsection File maps
@cindex File maps
@cindex Flat file maps
@cindex File map syntactic conventions
When @i{Amd} searches a file for a map entry it does a simple scan of
the file and supports both comments and continuation lines.
Continuation lines are indicated by a backslash character (@samp{\}) as
the last character of a line in the file. The backslash, newline character
@emph{and any leading white space on the following line} are discarded. A maximum
line length of 2047 characters is enforced after continuation lines are read
but before comments are stripped. Each line must end with
a newline character; that is newlines are terminators, not separators.
The following examples illustrate this:
@example
key valA valB; \
valC
@end example
specifies @emph{three} locations, and is identical to
@example
key valA valB; valC
@end example
However,
@example
key valA valB;\
valC
@end example
specifies only @emph{two} locations, and is identical to
@example
key valA valB;valC
@end example
After a complete line has been read from the file, including
continuations, @i{Amd} determines whether there is a comment on the
line. A comment begins with a hash (``@samp{#}'') character and
continues to the end of the line. There is no way to escape or change
the comment lead-in character.
Note that continuation lines and comment support @dfn{only} apply to
file maps.
When caching is enabled, file maps have a default cache mode of
@code{all} (@pxref{Automount Filesystem}).
@node NIS maps, Hesiod maps, File maps, Map Types
@comment node-name, next, previous, up
@subsection NIS maps
@cindex NIS (YP) maps
When using NIS (formerly YP), an @i{Amd} map is implemented directly
by the underlying NIS map. Comments and continuation lines are
@emph{not} supported in the automounter and must be stripped when
constructing the NIS server's database.
NIS maps do not support cache mode @code{all} and, when caching is
enabled, have a default cache mode of @code{inc} (@pxref{Automount Filesystem}).
The following rule illustrates what could be added to your NIS @file{Makefile},
in this case causing the @file{amd.home} map to be rebuilt:
@example
$(YPTSDIR)/amd.home.time: $(ETCDIR)/amd.home
-@@sed -e "s/#.*$$//" -e "/^$$/d" $(ETCDIR)/amd.home | \
awk '@{ \
for (i = 1; i <= NF; i++) \
if (i == NF) @{ \
if (substr($$i, length($$i), 1) == "\\") \
printf("%s", substr($$i, 1, length($$i) - 1)); \
else \
printf("%s\n", $$i); \
@} \
else \
printf("%s ", $$i); \
@}' | \
$(MAKEDBM) - $(YPDBDIR)/amd.home; \
touch $(YPTSDIR)/amd.home.time; \
echo "updated amd.home"; \
if [ ! $(NOPUSH) ]; then \
$(YPPUSH) amd.home; \
echo "pushed amd.home"; \
else \
: ; \
fi
@end example
Here @code{$(YPTSDIR)} contains the time stamp files, and @code{$(YPDBDIR)} contains
the dbm format NIS files.
@node Hesiod maps, Password maps, NIS maps, Map Types
@comment node-name, next, previous, up
@subsection Hesiod maps
@cindex Hesiod maps
When the map name begins with the string @samp{hesiod.} lookups are made
using the @dfn{Hesiod} name server. The string following the dot is
used as a name qualifier and is prepended with the key being located.
The entire string is then resolved in the @code{automount} context. For
example, if the key is @samp{jsp} and map name is
@samp{hesiod.homes} then @dfn{Hesiod} is asked to resolve
@samp{jsp.homes.automount}.
Hesiod maps do not support cache mode @samp{all} and, when caching is
enabled, have a default cache mode of @samp{inc} (@pxref{Automount Filesystem}).
The following is an example of a @dfn{Hesiod} map entry:
@example
jsp.homes.automount HS TXT "rfs:=/home/charm;rhost:=charm;sublink:=jsp"
njw.homes.automount HS TXT "rfs:=/home/dylan/dk2;rhost:=dylan;sublink:=njw"
@end example
@node Password maps, Union maps, Hesiod maps, Map Types
@comment node-name, next, previous, up
@subsection Password maps
@cindex Password file maps
@cindex /etc/passwd maps
@cindex User maps, automatic generation
@cindex Automatic generation of user maps
@cindex Using the password file as a map
The password map support is unlike the four previous map types. When
the map name is the string @file{/etc/passwd} @i{Amd} can lookup a user
name in the password file and re-arrange the home directory field to
produce a usable map entry.
@i{Amd} assumes the home directory has the format
`@t{/}@i{anydir}@t{/}@i{dom1}@t{/../}@i{domN}@t{/}@i{login}'.
@c @footnote{This interpretation is not necessarily exactly what you want.}
It breaks this string into a map entry where @code{$@{rfs@}} has the
value `@t{/}@i{anydir}@t{/}@i{domN}', @code{$@{rhost@}} has the value
`@i{domN}@t{.}@i{...}@t{.}@i{dom1}', and @code{$@{sublink@}} has the
value @samp{login}.@refill
Thus if the password file entry was
@example
/home/achilles/jsp
@end example
the map entry used by @i{Amd} would be
@example
rfs:=/home/achilles;rhost:=achilles;sublink:=jsp
@end example
Similarly, if the password file entry was
@example
/home/cc/sugar/mjh
@end example
the map entry used by @i{Amd} would be
@example
rfs:=/home/sugar;rhost:=sugar.cc;sublink:=jsp
@end example
@node Union maps, , Password maps, Map Types
@comment node-name, next, previous, up
@subsection Union maps
@cindex Union file maps
The union map support is provided specifically for use with the union
filesystem, @pxref{Union Filesystem}.
It is identified by the string @samp{union:} which is followed by a
colon separated list of directories. The directories are read in order,
and the names of all entries are recorded in the map cache. Later
directories take precedence over earlier ones. The union filesystem
type then uses the map cache to determine the union of the names in all
the directories.
@c subsection Gdbm
@node Key Lookup, Location Format, Map Types, Mount Maps
@comment node-name, next, previous, up
@section How keys are looked up
@cindex Key lookup
@cindex Map lookup
@cindex Looking up keys
@cindex How keys are looked up
@cindex Wildcards in maps
The key is located in the map whose type was determined when the
automount point was first created. In general the key is a pathname
component. In some circumstances this may be modified by variable
expansion (@pxref{Variable Expansion}) and prefixing. If the automount
point has a prefix, specified by the @var{pref} option, then that is
prepended to the search key before the map is searched.
If the map cache is a @samp{regexp} cache then the key is treated as an
egrep-style regular expression, otherwise a normal string comparison is
made.
If the key cannot be found then a @dfn{wildcard} match is attempted.
@i{Amd} repeatedly strips the basename from the key, appends @samp{/*} and
attempts a lookup. Finally, @i{Amd} attempts to locate the special key @samp{*}.
@group
For example, the following sequence would be checked if @file{home/dylan/dk2} was
being located:
@example
home/dylan/dk2
home/dylan/*
home/*
*
@end example
@end group
At any point when a wildcard is found, @i{Amd} proceeds as if an exact
match had been found and the value field is then used to resolve the
mount request, otherwise an error code is propagated back to the kernel.
(@pxref{Filesystem Types}).@refill
@node Location Format, , Key Lookup, Mount Maps
@comment node-name, next, previous, up
@section Location Format
@cindex Location format
@cindex Map entry format
@cindex How locations are parsed
The value field from the lookup provides the information required to
mount a filesystem. The information is parsed according to the syntax
shown below.
@display
@i{location-list}:
@i{location-selection}
@i{location-list} @i{white-space} @t{||} @i{white-space} @i{location-selection}
@i{location-selection}:
@i{location}
@i{location-selection} @i{white-space} @i{location}
@i{location}:
@i{location-info}
@t{-}@i{location-info}
@t{-}
@i{location-info}:
@i{sel-or-opt}
@i{location-info}@t{;}@i{sel-or-opt}
@t{;}
@i{sel-or-opt}:
@i{selection}
@i{opt-ass}
@i{selection}:
selector@t{==}@i{value}
selector@t{!=}@i{value}
@i{opt-ass}:
option@t{:=}@i{value}
@i{white-space}:
space
tab
@end display
Note that unquoted whitespace is not allowed in a location description.
White space is only allowed, and is mandatory, where shown with non-terminal
@samp{white-space}.
A @dfn{location-selection} is a list of possible volumes with which to
satisfy the request. @dfn{location-selection}s are separated by the
@samp{||} operator. The effect of this operator is to prevent use of
location-selections to its right if any of the location-selections on
its left were selected whether or not any of them were successfully
mounted (@pxref{Selectors}).@refill
The location-selection, and singleton @dfn{location-list},
@samp{type:=ufs;dev:=/dev/xd1g} would inform @i{Amd} to mount a UFS
filesystem from the block special device @file{/dev/xd1g}.
The @dfn{sel-or-opt} component is either the name of an option required
by a specific filesystem, or it is the name of a built-in, predefined
selector such as the architecture type. The value may be quoted with
double quotes @samp{"}, for example
@samp{type:="ufs";dev:="/dev/xd1g"}. These quotes are stripped when the
value is parsed and there is no way to get a double quote into a value
field. Double quotes are used to get white space into a value field,
which is needed for the program filesystem (@pxref{Program Filesystem}).@refill
@menu
* Map Defaults::
* Variable Expansion::
* Selectors::
* Map Options::
@end menu
@node Map Defaults, Variable Expansion, Location Format, Location Format
@comment node-name, next, previous, up
@subsection Map Defaults
@cindex Map defaults
@cindex How to set default map parameters
@cindex Setting default map parameters
A location beginning with a dash @samp{-} is used to specify default
values for subsequent locations. Any previously specified defaults in
the location-list are discarded. The default string can be empty in
which case no defaults apply.
The location @samp{-fs:=/mnt;opts:=ro} would set the local mount point
to @file{/mnt} and cause mounts to be read-only by default. Defaults
specified this way are appended to, and so override, any global map
defaults given with @samp{/defaults}).
@c
@c A @samp{/defaults} value @dfn{gdef} and a location list
@c \begin{quote}
@c $@samp{-}@dfn{def}_a $\verb*+ +$ @dfn{loc}_{a_1} $\verb*+ +$ @dfn{loc}_{a_2} $\verb*+ +$ @samp{-}@dfn{def}_b $\verb*+ +$ @dfn{loc}_{b_1} \ldots$
@c \end{quote}
@c is equivalent to
@c \begin{quote}
@c $@samp{-}@dfn{gdef}@samp{;}@dfn{def}_a $\verb*+ +$ @dfn{loc}_{a_1} $\verb*+ +$ @dfn{loc}_{a_2} $\verb*+ +$ @samp{-}@dfn{gdef}@samp{;}@dfn{def}_b $\verb*+ +$ @dfn{loc}_{b_1} \ldots$
@c \end{quote}
@c which is equivalent to
@c \begin{quote}
@c $@dfn{gdef}@samp{;}@dfn{def}_a@samp{;}@dfn{loc}_{a_1} $\verb*+ +$@dfn{gdef}@samp{;}@dfn{def}_a@samp{;}@dfn{loc}_{a_2} $\verb*+ +$@dfn{gdef}@samp{;}@dfn{def}_b@samp{;}@dfn{loc}_{b_1} \ldots$
@c \end{quote}
@node Variable Expansion, Selectors, Map Defaults, Location Format
@comment node-name, next, previous, up
@subsection Variable Expansion
@cindex Variable expansion
@cindex How variables are expanded
@cindex Pathname operators
@cindex Domain stripping
@cindex Domainname operators
@cindex Stripping the local domain name
@cindex Environment variables
@cindex How to access environment variables in maps
To allow generic location specifications @i{Amd} does variable expansion
on each location and also on some of the option strings. Any option or
selector appearing in the form @code{$@dfn{var}} is replaced by the
current value of that option or selector. For example, if the value of
@code{$@{key@}} was @samp{bin}, @code{$@{autodir@}} was @samp{/a} and
@code{$@{fs@}} was `@t{$@{autodir@}}@t{/local/}@t{$@{key@}}' then
after expansion @code{$@{fs@}} would have the value @samp{/a/local/bin}.
Any environment variable can be accessed in a similar way.@refill
Two pathname operators are available when expanding a variable. If the
variable name begins with @samp{/} then only the last component of
then pathname is substituted. For example, if @code{$@{path@}} was
@samp{/foo/bar} then @code{$@{/path@}} would be expanded to @samp{bar}.
Similarly, if the variable name ends with @samp{/} then all but the
last component of the pathname is substituted. In the previous example,
@code{$@{path/@}} would be expanded to @samp{/foo}.@refill
Two domain name operators are also provided. If the variable name
begins with @samp{.} then only the domain part of the name is
substituted. For example, if @code{$@{rhost@}} was
@samp{swan.doc.ic.ac.uk} then @code{$@{.rhost@}} would be expanded to
@samp{doc.ic.ac.uk}. Similarly, if the variable name ends with @samp{.}
then only the host component is substituted. In the previous example,
@code{$@{rhost.@}} would be expanded to @samp{swan}.@refill
Variable expansion is a two phase process. Before a location is parsed,
all references to selectors, @i{eg} @code{$@{path@}}, are expanded. The
location is then parsed, selections are evaluated and option assignments
recorded. If there were no selections or they all succeeded the
location is used and the values of the following options are expanded in
the order given: @var{sublink}, @var{rfs}, @var{fs}, @var{opts},
@var{remopts}, @var{mount} and @var{unmount}.
Note that expansion of option values is done after @dfn{all} assignments
have been completed and not in a purely left to right order as is done
by the shell. This generally has the desired effect but care must be
taken if one of the options references another, in which case the
ordering can become significant.
There are two special cases concerning variable expansion:
@enumerate
@item
before a map is consulted, any selectors in the name received
from the kernel are expanded. For example, if the request from the
kernel was for `@t{$@{arch@}}@t{.bin}' and the machine architecture
was @samp{vax}, the value given to @code{$@{key@}} would be
@samp{vax.bin}.@refill
@item
the value of @code{$@{rhost@}} is expanded and normalized before the
other options are expanded. The normalization process strips any local
sub-domain components. For example, if @code{$@{domain@}} was
@samp{Berkeley.EDU} and @code{$@{rhost@}} was initially
@samp{snow.Berkeley.EDU}, after the normalization it would simply be
@samp{snow}. Hostname normalization is currently done in a
@emph{case-dependent} manner.@refill
@end enumerate
@node Selectors, Map Options, Variable Expansion, Location Format
@comment node-name, next, previous, up
@subsection Selectors
@cindex Selectors
Selectors are used to control the use of a location. It is possible to
share a mount map between many machines in such a way that filesystem
location, architecture and operating system differences are hidden from
the users. A selector of the form @samp{arch==sun3;os==sos4} would only
apply on Sun-3s running SunOS 4.x.
Selectors are evaluated left to right. If a selector fails then that
location is ignored. Thus the selectors form a conjunction and the
locations form a disjunction. If all the locations are ignored or
otherwise fail then @i{Amd} uses the @dfn{error} filesystem
(@pxref{Error Filesystem}). This is equivalent to having a location
@samp{type:=error} at the end of each mount-map entry.@refill
The selectors currently implemented are:
@table @samp
@cindex arch, mount selector
@cindex Mount selector; arch
@cindex Selector; arch
@item arch
the machine architecture which was automatically determined at compile
time. The architecture type can be displayed by running the command
@samp{amd -v}. @xref{Supported Machine Architectures}.@refill
@item autodir
@cindex autodir, mount selector
@cindex Mount selector; autodir
@cindex Selector; autodir
the default directory under which to mount filesystems. This may be
changed by the ``-a'' command line option. See the @var{fs} option.
@item byte
@cindex byte, mount selector
@cindex Mount selector; byte
@cindex Selector; byte
the machine's byte ordering. This is either @samp{little}, indicating
little-endian, or @samp{big}, indicating big-endian. One possible use
is to share @samp{rwho} databases (@pxref{rwho servers}).
@item cluster
@cindex cluster, mount selector
@cindex Mount selector; cluster
@cindex Selector; cluster
is provided as a hook for the name of the local cluster. This can be
used to decide which servers to use for copies of replicated
filesystems. @code{$@{cluster@}} defaults to the value of
@code{$@{domain@}} unless a different value is set with the ``-C''
command line option.
@item domain
@cindex domain, mount selector
@cindex Mount selector; domain
@cindex Selector; domain
the local domain name as specified by the ``-d'' command line option.
See @samp{host}.
@item host
@cindex host, mount selector
@cindex Mount selector; host
@cindex Selector; host
the local hostname as determined by @b{gethostname}(2). If no domain
name was specified on the command line and the hostname contains a
period @samp{.} then the string before the period is used as the
host name, and the string after the period is assigned to
@code{$@{domain@}}. For example, if the hostname is
@samp{styx.doc.ic.ac.uk} then @code{host} would be @samp{styx} and
@code{domain} would be @samp{doc.ic.ac.uk}. @code{hostd} would be
@samp{styx.doc.ic.ac.uk}.@refill
@item hostd
@cindex hostd, mount selector
@cindex Mount selector; hostd
@cindex Selector; hostd
is @code{$@{host@}} and @code{$@{domain@}} concatenated with a
@samp{.} inserted between them if required. If @code{$@{domain@}}
is an empty string then @code{$@{host@}} and @code{$@{hostd@}} will be
identical.
@item karch
@cindex karch, mount selector
@cindex Mount selector; karch
@cindex Selector; karch
is provided as a hook for the kernel architecture. This is used on
SunOS 4, for example, to distinguish between different @samp{/usr/kvm}
volumes. @code{$@{karch@}} defaults to the value of @code{$@{arch@}}
unless a different value is set with the ``-k'' command line option.
@item os
@cindex os, mount selector
@cindex Mount selector; os
@cindex Selector; os
the operating system. Like the machine architecture, this is
automatically determined at compile time. The operating system name can
be displayed by running the command @samp{amd -v}. @xref{Supported
Operating Systems}.@refill
@end table
The following selectors are also provided. Unlike the other selectors,
they vary for each lookup. Note that when the name from the kernel is
expanded prior to a map lookup, these selectors are all defined as empty
strings.
@table @samp
@item key
@cindex key, mount selector
@cindex Mount selector; key
@cindex Selector; key
the name being resolved. For example, if @file{/home} is an automount
point, then accessing @file{/home/foo} would set @code{$@{key@}} to the
string @samp{foo}. The key is prefixed by the @var{pref} option set in
the parent mount point. The default prefix is an empty string. If the
prefix was @file{blah/} then @code{$@{key@}} would be set to
@file{blah/foo}.@refill
@item map
@cindex map, mount selector
@cindex Mount selector; map
@cindex Selector; map
the name of the mount map being used.
@item path
@cindex path, mount selector
@cindex Mount selector; path
@cindex Selector; path
the full pathname of the name being resolved. For example
@file{/home/foo} in the example above.
@item wire
@cindex wire, mount selector
@cindex Mount selector; wire
@cindex Selector; wire
the name of the network to which the primary network interface is
attached. If a symbolic name cannot be found in the networks or hosts
database then dotted IP address format is used. This value is also
output by the ``-v'' option.
@end table
Selectors can be negated by using @samp{!=} instead of @samp{==}. For
example to select a location on all non-Vax machines the selector
@samp{arch!=vax} would be used.
@node Map Options, , Selectors, Location Format
@comment node-name, next, previous, up
@subsection Map Options
@cindex Map options
@cindex Setting map options
Options are parsed concurrently with selectors. The difference is that
when an option is seen the string following the @samp{:=} is
recorded for later use. As a minimum the @var{type} option must be
specified. Each filesystem type has other options which must also be
specified. @xref{Filesystem Types}, for details on the filesystem
specific options.@refill
Superfluous option specifications are ignored and are not reported
as errors.
The following options apply to more than one filesystem type.
@menu
* delay Option::
* fs Option::
* opts Option::
* remopts Option::
* sublink Option::
* type Option::
@end menu
@node delay Option, fs Option, Map Options, Map Options
@comment node-name, next, previous, up
@subsubsection delay Option
@cindex Setting a delay on a mount location
@cindex Delaying mounts from specific locations
@cindex Primary server
@cindex Secondary server
@cindex delay, mount option
@cindex Mount option; delay
The delay, in seconds, before an attempt will be made to mount from the current location.
Auxilliary data, such as network address, file handles and so on are computed
regardless of this value.
A delay can be used to implement the notion of primary and secondary file servers.
The secondary servers would have a delay of a few seconds,
thus giving the primary servers a chance to respond first.
@node fs Option, opts Option, delay Option, Map Options
@comment node-name, next, previous, up
@subsubsection fs Option
@cindex Setting the local mount point
@cindex Overriding the default mount point
@cindex fs, mount option
@cindex Mount option; fs
The local mount point. The semantics of this option vary between
filesystems.
For NFS and UFS filesystems the value of @code{$@{fs@}} is used as the
local mount point. For other filesystem types it has other meanings
which are described in the section describing the respective filesystem
type. It is important that this string uniquely identifies the
filesystem being mounted. To satisfy this requirement, it should
contain the name of the host on which the filesystem is resident and the
pathname of the filesystem on the local or remote host.
The reason for requiring the hostname is clear if replicated filesystems
are considered. If a fileserver goes down and a replacement filesystem
is mounted then the @dfn{local} mount point @dfn{must} be different from
that of the filesystem which is hung. Some encoding of the filesystem
name is required if more than one filesystem is to be mounted from any
given host.
If the hostname is first in the path then all mounts from a particular
host will be gathered below a single directory. If that server goes
down then the hung mount points are less likely to be accidentally
referenced, for example when @b{getwd}(3) traverses the namespace to
find the pathname of the current directory.
The @samp{fs} option defaults to
@code{$@{autodir@}/$@{rhost@}$@{rfs@}}. In addition,
@samp{rhost} defaults to the local host name (@code{$@{host@}}) and
@samp{rfs} defaults to the value of @code{$@{path@}}, which is the full
path of the requested file; @samp{/home/foo} in the example above
(@pxref{Selectors}). @code{$@{autodir@}} defaults to @samp{/a} but may
be changed with the ``-a'' command line option. Sun's automounter
defaults to @samp{/tmp_mnt}. Note that there is no @samp{/} between
the @code{$@{rhost@}} and @code{$@{rfs@}} since @code{$@{rfs@}} begins
with a @samp{/}.@refill
@node opts Option, remopts Option, fs Option, Map Options
@comment node-name, next, previous, up
@subsubsection opts Option
@cindex Setting system mount options
@cindex Passing parameters to the mount system call
@cindex mount system call
@cindex mount system call flags
@cindex The mount system call
@cindex opts, mount option
@cindex Mount option; opts
The options to pass to the mount system call. A leading @samp{-} is
silently ignored. The mount options supported generally correspond to
those used by @b{mount}(8) and are listed below. Some additional
pseudo-options are interpreted by @i{Amd} and are also listed.
Unless specifically overridden, each of the system default mount options
applies. Any options not recognised are ignored. If no options list is
supplied the string @samp{rw,defaults} is used and all the system
default mount options apply. Options which are not applicable for a
particular operating system are silently ignored. For example, only 4.4
BSD is known to implement the @code{compress} and @code{spongy} options.
@table @code
@item compress
Use NFS compression protocol.
@item grpid
Use BSD directory group-id semantics.
@item intr
Allow keyboard interrupts on hard mounts.
@item noconn
Don't make a connection on datagram transports.
@item nocto
No close-to-open consistency.
@item nodevs
Don't allow local special devices on this filesystem.
@item nolooklease
When nqnfs is selected, disable the automatic acquiring of leases on
lookups.
@item nosuid
Don't allow set-uid or set-gid executables on this filesystem.
@item nqnfs
Enable leasing extensions to the NFS protocol know as Not Quite NFS.
Only supported by the 4.4BSD NFS implementation.
@item quota
Enable quota checking on this mount.
@item resvport
Use a reserved port number (one less than 1024). Some NFS servers
require this.
@item retrans=@i{n}
The number of NFS retransmits made before a user error is generated by a
@samp{soft} mounted filesystem, and before a @samp{hard} mounted
filesystem reports @samp{NFS server @dfn{yoyo} not responding still
trying}.
@item ro
Mount this filesystem readonly.
@item rsize=@var{n}
The NFS read packet size. You may need to set this if you are using
NFS/UDP through a gateway.
@item soft
Give up after @dfn{retrans} retransmissions.
@item spongy
Like @samp{soft} for status requests, and @samp{hard} for data transfers.
@item tcp
Use TCP/IP instead of UDP/IP, ignored if the NFS implementation does not
support TCP/IP mounts.
@item timeo=@var{n}
The NFS timeout, in tenth-seconds, before a request is retransmitted.
@item wsize=@var{n}
The NFS write packet size. You may need to set this if you are using
NFS/UDP through a gateway.
@end table
The following options are implemented by @i{Amd}, rather than being
passed to the kernel.
@table @code
@item nounmount
Configures the mount so that its time-to-live will
never expire. This is also the default for some filesystem types.
@c
@c Implementation broken:
@item ping=@var{n}
The interval, in seconds, between keep-alive pings. When four
consecutive pings have failed the mount point is marked as hung. This
interval defaults to 30 seconds. If the ping interval is less than zero,
no pings are sent and the host is assumed to be always
up. By default, pings are not sent for an NFS/TCP mount.
@item retry=@var{n}
The number of times to retry the mount system call.
@item utimeout=@var{n}
The interval, in seconds, by which the mount's
time-to-live is extended after an unmount attempt
has failed. In fact the interval is extended before the unmount is
attempted to avoid thrashing. The default value is 120 seconds (two
minutes) or as set by the ``-w'' command line option.
@end table
@node remopts Option, sublink Option, opts Option, Map Options
@comment node-name, next, previous, up
@subsubsection remopts Option
@cindex Setting system mount options for non-local networks
@cindex remopts, mount option
@cindex Mount option; remopts
This option has the same use as @code{$@{opts@}} but applies only when
the remote host is on a non-local network. For example, when using NFS
across a gateway it is often necessary to use smaller values for the
data read and write sizes. This can simply be done by specifying the
small values in @var{remopts}. When a non-local host is accessed, the
smaller sizes will automatically be used.
@i{Amd} determines whether a host is local by examining the network
interface configuration at startup. Any interface changes made after
@i{Amd} has been started will not be noticed. The likely effect will
be that a host may incorrectly be declared non-local.
Unless otherwise set, the value of @code{$@{rem@}} is the same as the
value of @code{$@{opts@}}.
@node sublink Option, type Option, remopts Option, Map Options
@comment node-name, next, previous, up
@subsubsection sublink Option
@cindex Setting the sublink option
@cindex sublink, mount option
@cindex Mount option; sublink
The subdirectory within the mounted filesystem to which the reference
should point. This can be used to prevent duplicate mounts in cases
where multiple directories in the same mounted filesystem are used.
@node type Option, , sublink Option, Map Options
@comment node-name, next, previous, up
@subsubsection type Option
@cindex Setting the filesystem type option
@cindex type, mount option
@cindex Mount option; type
The filesystem type to be used. @xref{Filesystem Types}, for a full
description of each type.@refill
@node Amd Command Line Options, Filesystem Types, Mount Maps, Top
@comment node-name, next, previous, up
@chapter @i{Amd} Command Line Options
@cindex Command line options, Amd
@cindex Amd command line options
@cindex Overriding defaults on the command line
Many of @i{Amd}'s parameters can be set from the command line. The
command line is also used to specify automount points and maps.
The general format of a command line is
@example
amd [@i{options}] @{ @i{directory} @i{map-name} [-@i{map-options}] @} ...
@end example
For each directory and map-name given, @i{Amd} establishes an
automount point. The @dfn{map-options} may be any sequence of options
or selectors---@pxref{Location Format}. The @dfn{map-options}
apply only to @i{Amd}'s mount point.
@samp{type:=toplvl;cache:=mapdefault;fs:=$@{map@}} is the default value for the
map options. Default options for a map are read from a special entry in
the map whose key is the string @samp{/defaults}. When default options
are given they are prepended to any options specified in the mount-map
locations as explained in. @xref{Map Defaults}, for more details.
The @dfn{options} are any combination of those listed below.
Once the command line has been parsed, the automount points are mounted.
The mount points are created if they do not already exist, in which case they
will be removed when @i{Amd} exits.
Finally, @i{Amd} disassociates itself from its controlling terminal and
forks into the background.
Note: Even if @i{Amd} has been built with @samp{-DDEBUG} it will still
background itself and disassociate itself from the controlling terminal.
To use a debugger it is necessary to specify @samp{-D nodaemon} on the
command line.
@menu
* -a Option:: Automount directory.
* -c Option:: Cache timeout interval.
* -d Option:: Domain name.
* -k Option:: Kernel architecture.
* -l Option:: Log file.
* -n Option:: Hostname normalisation.
* -p Option:: Output process id.
* -r Option:: Restart existing mounts.
* -t Option:: Kernel RPC timeout.
* -v Option:: Version information.
* -w Option:: Wait interval after failed unmount.
* -x Option:: Log options.
* -y Option:: NIS domain.
* -C-Option:: Cluster name.
* -D-Option:: Debug flags.
@end menu
@node -a Option, -c Option, Amd Command Line Options, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-a} @var{directory}
@cindex Automount directory
@cindex Setting the default mount directory
Specifies the default mount directory. This option changes the variable
@code{$@{autodir@}} which otherwise defaults to @file{/a}. For example,
some sites prefer @file{/amd}.
@example
amd -a /amd ...
@end example
@node -c Option, -d Option, -a Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-c} @var{cache-interval}
@cindex Cache interval
@cindex Interval before a filesystem times out
@cindex Setting the interval before a filesystem times out
@cindex Changing the interval before a filesystem times out
Selects the period, in seconds, for which a name is cached by @i{Amd}.
If no reference is made to the volume in this period, @i{Amd} discards
the volume name to filesystem mapping.
Once the last reference to a filesystem has been removed, @i{Amd}
attempts to unmount the filesystem. If the unmount fails the interval
is extended by a further period as specified by the @samp{-w} command
line option or by the @samp{utimeout} mount option.
The default @dfn{cache-interval} is 300 seconds (five minutes).
@node -d Option, -k Option, -c Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-d} @var{domain}
@cindex Domain name
@cindex Setting the local domain name
@cindex Overriding the local domain name
Specifies the host's domain. This sets the internal variable
@code{$@{domain@}} and affects the @code{$@{hostd@}} variable.
If this option is not specified and the hostname already contains the
local domain then that is used, otherwise the default value of
@code{$@{domain@}} is @samp{unknown.domain}.
For example, if the local domain was @samp{doc.ic.ac.uk}, @i{Amd} could
be started as follows:
@example
amd -d doc.ic.ac.uk ...
@end example
@node -k Option, -l Option, -d Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-k} @var{kernel-architecture}
@cindex Setting the Kernel architecture
Specifies the kernel architecture of the system. This is usually the
output of @samp{arch -k} and its only effect is to set the variable
@code{$@{karch@}}. If this option is not given, @code{$@{karch@}} has
the same value as @code{$@{arch@}}.
This would be used as follows:
@example
amd -k `arch -k` ...
@end example
@node -l Option, -n Option, -k Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-l} @var{log-option}
@cindex Log filename
@cindex Setting the log file
@cindex Using syslog to log errors
@cindex syslog
Selects the form of logging to be made. Two special @dfn{log-options}
are recognised.
@enumerate
@item
If @dfn{log-option} is the string @samp{syslog}, @i{Amd} will use the
@b{syslog}(3) mechanism.@refill
@item
If @dfn{log-option} is the string @samp{/dev/stderr}, @i{Amd} will use
standard error, which is also the default target for log messages. To
implement this, @i{Amd} simulates the effect of the @samp{/dev/fd}
driver.
@end enumerate
Any other string is taken as a filename to use for logging. Log
messages are appended to the file if it already exists, otherwise a new
file is created. The file is opened once and then held open, rather
than being re-opened for each message.
If the @samp{syslog} option is specified but the system does not support
syslog or if the named file cannot be opened or created, @i{Amd} will
use standard error. Error messages generated before @i{Amd} has
finished parsing the command line are printed on standard error.
Using @samp{syslog} is usually best, in which case @i{Amd} would be
started as follows:
@example
amd -l syslog ...
@end example
@node -n Option, -p Option, -l Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-n}
@cindex Hostname normalisation
@cindex Aliased hostnames
@cindex Resolving aliased hostnames
@cindex Normalising hostnames
Normalises the remote hostname before using it. Normalisation is done
by replacing the value of @code{$@{rhost@}} with the primary name
returned by a hostname lookup.
This option should be used if several names are used to refer to a
single host in a mount map.
@node -p Option, -r Option, -n Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-p}
@cindex Process id
@cindex Displaying the process id
@cindex process id of Amd daemon
@cindex pid file, creating with -p option
@cindex Creating a pid file
Causes @i{Amd}'s process id to be printed on standard output.
This can be redirected to a suitable file for use with kill:
@example
amd -p > /var/run/amd.pid ...
@end example
This option only has an affect if @i{Amd} is running in daemon mode.
If @i{Amd} is started with the @code{-D nodaemon} debug flag, this
option is ignored.
@node -r Option, -t Option, -p Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-r}
@cindex Restarting existing mounts
@cindex Picking up existing mounts
Tells @i{Amd} to restart existing mounts (@pxref{Inheritance Filesystem}).
@c @dfn{This option will be made the default in the next release.}
@node -t Option, -v Option, -r Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-t} @var{timeout.retransmit}
@cindex Setting Amd's RPC parameters
Specifies the RPC @dfn{timeout} and @dfn{retransmit} intervals used by
the kernel to communicate to @i{Amd}. These are used to set the
@samp{timeo} and @samp{retrans} mount options.
@i{Amd} relies on the kernel RPC retransmit mechanism to trigger mount
retries. The value of this parameter changes the retry interval. Too
long an interval gives poor interactive response, too short an interval
causes excessive retries.
@node -v Option, -w Option, -t Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-v}
@cindex Version information
@cindex Discovering version information
@cindex How to discover your version of Amd
Print version information on standard error and then exit. The output
is of the form:
@example
amd 5.2.1.11 of 91/03/17 18:04:05 5.3Alpha11 #0: Sun Mar 17 18:07:28 GMT 1991
Built by pendry@@vangogh.Berkeley.EDU for a hp300 running bsd44 (big-endian).
Map support for: root, passwd, union, file, error.
FS: ufs, nfs, nfsx, host, link, program, union, auto, direct, toplvl, error.
Primary network is 128.32.130.0.
@end example
The information includes the version number, release date and name of
the release. The architecture (@pxref{Supported Machine Architectures}),
operating system (@pxref{Supported Operating Systems})
and byte ordering are also printed as they appear in the @code{$@{os@}},
@code{$@{arch@}} and @code{$@{byte@}} variables.@refill
@node -w Option, -x Option, -v Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-w} @var{wait-timeout}
@cindex Setting the interval between unmount attempts
@cindex unmount attempt backoff interval
Selects the interval in seconds between unmount attempts after the
initial time-to-live has expired.
This defaults to 120 seconds (two minutes).
@node -x Option, -y Option, -w Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-x} @var{opts}
@cindex Log message selection
@cindex Selecting specific log messages
@cindex How to select log messages
@cindex syslog priorities
Specifies the type and verbosity of log messages. @dfn{opts} is
a comma separated list selected from the following options:
@table @code
@item fatal
Fatal errors
@item error
Non-fatal errors
@item user
Non-fatal user errors
@item warn
Recoverable errors
@item warning
Alias for @code{warn}
@item info
Information messages
@item map
Mount map usage
@item stats
Additional statistics
@item all
All of the above
@end table
Initially a set of default logging flags is enabled. This is as if
@samp{-x all,nomap,nostats} had been selected. The command line is
parsed and logging is controlled by the ``-x'' option. The very first
set of logging flags is saved and can not be subsequently disabled using
@i{Amq}. This default set of options is useful for general production
use.@refill
The @samp{info} messages include details of what is mounted and
unmounted and when filesystems have timed out. If you want to have the
default set of messages without the @samp{info} messages then you simply
need @samp{-x noinfo}. The messages given by @samp{user} relate to
errors in the mount maps, so these are useful when new maps are
installed. The following table lists the syslog priorites used for each
of the message types.@refill
@table @code
@item fatal
LOG_CRIT
@item error
LOG_ERR
@item user
LOG_WARNING
@item warning
LOG_WARNING
@item info
LOG_INFO
@item debug
LOG_DEBUG
@item map
LOG_DEBUG
@item stats
LOG_INFO
@end table
The options can be prefixed by the string @samp{no} to indicate
that this option should be turned off. For example, to obtain all
but @samp{info} messages the option @samp{-x all,noinfo} would be used.
If @i{Amd} was built with debugging enabled the @code{debug} option is
automatically enabled regardless of the command line options.
@node -y Option, -C-Option, -x Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-y} @var{NIS-domain}
@cindex NIS (YP) domain name
@cindex Overriding the NIS (YP) domain name
@cindex Setting the NIS (YP) domain name
@cindex YP domain name
Selects an alternate NIS domain. This is useful for debugging and
cross-domain shared mounting. If this flag is specified, @i{Amd}
immediately attempts to bind to a server for this domain.
@c @i{Amd} refers to NIS maps when it starts, unless the ``-m'' option
@c is specified, and whenever required in a mount map.
@node -C-Option, -D-Option, -y Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-C} @var{cluster-name}
@cindex Cluster names
@cindex Setting the cluster name
Specifies the name of the cluster of which the local machine is a member.
The only effect is to set the variable @code{$@{cluster@}}.
The @dfn{cluster-name} is will usually obtained by running another command which uses
a database to map the local hostname into a cluster name.
@code{$@{cluster@}} can then be used as a selector to restrict mounting of
replicated data.
If this option is not given, @code{$@{cluster@}} has the same value as @code{$@{domain@}}.
This would be used as follows:
@example
amd -C `clustername` ...
@end example
@node -D-Option, , -C-Option, Amd Command Line Options
@comment node-name, next, previous, up
@section @code{-D} @var{opts}
@cindex Debug options
@cindex Setting debug flags
Controls the verbosity and coverage of the debugging trace; @dfn{opts}
is a comma separated list of debugging options. The ``-D'' option is
only available if @i{Amd} was compiled with @samp{-DDEBUG}. The memory
debugging facilities are only available if @i{Amd} was compiled with
@samp{-DDEBUG_MEM} (in addition to @samp{-DDEBUG}).
The most common options to use are @samp{-D trace} and @samp{-D test}
(which turns on all the useful debug options). See the program source
for a more detailed explanation of the available options.
@node Filesystem Types, Run-time Administration, Amd Command Line Options, Top
@comment node-name, next, previous, up
@chapter Filesystem Types
@cindex Filesystem types
@cindex Mount types
@cindex Types of filesystem
To mount a volume, @i{Amd} must be told the type of filesystem to be
used. Each filesystem type typically requires additional information
such as the fileserver name for NFS.
From the point of view of @i{Amd}, a @dfn{filesystem} is anything that
can resolve an incoming name lookup. An important feature is support
for multiple filesystem types. Some of these filesystems are
implemented in the local kernel and some on remote fileservers, whilst
the others are implemented internally by @i{Amd}.@refill
The two common filesystem types are UFS and NFS. Four other user
accessible filesystems (@samp{link}, @samp{program}, @samp{auto} and
@samp{direct}) are also implemented internally by @i{Amd} and these are
described below. There are two additional filesystem types internal to
@i{Amd} which are not directly accessible to the user (@samp{inherit}
and @samp{error}). Their use is described since they may still have an
effect visible to the user.@refill
@menu
* Network Filesystem:: A single NFS filesystem.
* Network Host Filesystem:: NFS mount a host's entire export tree.
* Network Filesystem Group:: An atomic group of NFS filesystems.
* Unix Filesystem:: Native disk filesystem.
* Program Filesystem:: Generic Program mounts.
* Symbolic Link Filesystem:: Local link referencing existing filesystem.
* Automount Filesystem::
* Direct Automount Filesystem::
* Union Filesystem::
* Error Filesystem::
* Top-level Filesystem::
* Root Filesystem::
* Inheritance Filesystem::
@end menu
@node Network Filesystem, Network Host Filesystem, Filesystem Types, Filesystem Types
@comment node-name, next, previous, up
@section Network Filesystem (@samp{type:=nfs})
@cindex NFS
@cindex Mounting an NFS filesystem
@cindex How to mount and NFS filesystem
@cindex nfs, filesystem type
@cindex Filesystem type; nfs
The @dfn{nfs} filesystem type provides access to Sun's NFS.
@noindent
The following options must be specified:
@table @code
@cindex rhost, mount option
@cindex Mount option; rhost
@item rhost
the remote fileserver. This must be an entry in the hosts database. IP
addresses are not accepted. The default value is taken
from the local host name (@code{$@{host@}}) if no other value is
specified.
@cindex rfs, mount option
@cindex Mount option; rfs
@item rfs
the remote filesystem.
If no value is specified for this option, an internal default of
@code{$@{path@}} is used.
@end table
NFS mounts require a two stage process. First, the @dfn{file handle} of
the remote file system must be obtained from the server. Then a mount
system call must be done on the local system. @i{Amd} keeps a cache
of file handles for remote file systems. The cache entries have a
lifetime of a few minutes.
If a required file handle is not in the cache, @i{Amd} sends a request
to the remote server to obtain it. @i{Amd} @dfn{does not} wait for
a response; it notes that one of the locations needs retrying, but
continues with any remaining locations. When the file handle becomes
available, and assuming none of the other locations was successfully
mounted, @i{Amd} will retry the mount. This mechanism allows several
NFS filesystems to be mounted in parallel.
@c @footnote{The mechanism
@c is general, however NFS is the only filesystem
@c for which the required hooks have been written.}
The first one which responds with a valid file handle will be used.
@noindent
An NFS entry might be:
@example
jsp host!=charm;type:=nfs;rhost:=charm;rfs:=/home/charm;sublink:=jsp
@end example
The mount system call and any unmount attempts are always done
in a new task to avoid the possibilty of blocking @i{Amd}.
@node Network Host Filesystem, Network Filesystem Group, Network Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Network Host Filesystem (@samp{type:=host})
@cindex Network host filesystem
@cindex Mounting entire export trees
@cindex How to mount all NFS exported filesystems
@cindex host, filesystem type
@cindex Filesystem type; host
@c NOTE: the current implementation of the @dfn{host} filesystem type
@c sometimes fails to maintain a consistent view of the remote mount tree.
@c This happens when the mount times out and only some of the remote mounts
@c are successfully unmounted. To prevent this from occurring, use the
@c @samp{nounmount} mount option.
The @dfn{host} filesystem allows access to the entire export tree of an
NFS server. The implementation is layered above the @samp{nfs}
implementation so keep-alives work in the same way. The only option
which needs to specified is @samp{rhost} which is the name of the
fileserver to mount.
The @samp{host} filesystem type works by querying the mount daemon on
the given fileserver to obtain its export list. @i{Amd} then obtains
filehandles for each of the exported filesystems. Any errors at this
stage cause that particular filesystem to be ignored. Finally each
filesystem is mounted. Again, errors are logged but ignored. One
common reason for mounts to fail is that the mount point does not exist.
Although @i{Amd} attempts to automatically create the mount point, it
may be on a remote filesystem to which @i{Amd} does not have write
permission.
When an attempt to unmount a @samp{host} filesystem mount fails, @i{Amd}
remounts any filesystems which had successfully been unmounted. To do
this @i{Amd} queries the mount daemon again and obtains a fresh copy of
the export list. @i{Amd} then tries to mount any exported filesystems
which are not currently mounted.
Sun's automounter provides a special @samp{-hosts} map. To achieve the
same effect with @i{Amd} requires two steps. First a mount map must
be created as follows:
@example
/defaults type:=host;fs:=$@{autodir@}/$@{rhost@}/root;rhost:=$@{key@}
* opts:=rw,nosuid,grpid
@end example
@noindent
and then start @i{Amd} with the following command
@example
amd /n net.map
@end example
@noindent
where @samp{net.map} is the name of map described above. Note that the
value of @code{$@{fs@}} is overridden in the map. This is done to avoid
a clash between the mount tree and any other filesystem already mounted
from the same fileserver.
If different mount options are needed for different hosts then
additional entries can be added to the map, for example
@example
host2 opts:=ro,nosuid,soft
@end example
@noindent
would soft mount @samp{host2} read-only.
@node Network Filesystem Group, Unix Filesystem, Network Host Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Network Filesystem Group (@samp{type:=nfsx})
@cindex Network filesystem group
@cindex Atomic NFS mounts
@cindex Mounting an atomic group of NFS filesystems
@cindex How to mount an atomic group of NFS filesystems
@cindex nfsx, filesystem type
@cindex Filesystem type; nfsx
The @dfn{nfsx} filesystem allows a group of filesystems to be mounted
from a single NFS server. The implementation is layered above the
@samp{nfs} implementation so keep-alives work in the same way.
The options are the same as for the @samp{nfs} filesystem with one
difference.
@noindent
The following options must be specified:
@table @code
@item rhost
the remote fileserver. This must be an entry in the hosts database. IP
addresses are not accepted. The default value is taken from the local
host name (@code{$@{host@}}) if no other value is specified.
@item rfs
as a list of filesystems to mount. The list is in the form of a comma
separated strings.
@end table
@noindent
For example:
@example
pub type:=nfsx;rhost:=gould;\
rfs:=/public,/,graphics,usenet;fs:=$@{autodir@}/$@{rhost@}/root
@end example
The first string defines the root of the tree, and is applied as a
prefix to the remaining members of the list which define the individual
filesystems. The first string is @emph{not} used as a filesystem name.
A parallel operation is used to determine the local mount points to
ensure a consistent layout of a tree of mounts.
Here, the @emph{three} filesystems, @samp{/public},
@samp{/public/graphics} and @samp{/public/usenet}, would be mounted.@refill
A local mount point, @code{$@{fs@}}, @emph{must} be specified. The
default local mount point will not work correctly in the general case.
A suggestion is to use @samp{fs:=$@{autodir@}/$@{rhost@}/root}.@refill
@node Unix Filesystem, Program Filesystem, Network Filesystem Group, Filesystem Types
@comment node-name, next, previous, up
@section Unix Filesystem (@samp{type:=ufs})
@cindex Unix filesystem
@cindex UFS
@cindex Mounting a UFS filesystem
@cindex Mounting a local disk
@cindex How to mount a UFS filesystems
@cindex How to mount a local disk
@cindex Disk filesystems
@cindex ufs, filesystem type
@cindex Filesystem type; ufs
The @dfn{ufs} filesystem type provides access to the system's
standard disk filesystem---usually a derivative of the Berkeley Fast Filesystem.
@noindent
The following option must be specified:
@table @code
@cindex dev, mount option
@cindex Mount option; dev
@item dev
the block special device to be mounted.
@end table
A UFS entry might be:
@example
jsp host==charm;type:=ufs;dev:=/dev/xd0g;sublink:=jsp
@end example
@node Program Filesystem, Symbolic Link Filesystem, Unix Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Program Filesystem (@samp{type:=program})
@cindex Program filesystem
@cindex Mount a filesystem under program control
@cindex program, filesystem type
@cindex Filesystem type; program
The @dfn{program} filesystem type allows a program to be run whenever a
mount or unmount is required. This allows easy addition of support for
other filesystem types, such as MIT's Remote Virtual Disk (RVD)
which has a programmatic interface via the commands
@samp{rvdmount} and @samp{rvdunmount}.
@noindent
The following options must be specified:
@table @code
@cindex mount, mount option
@cindex Mount option; mount
@item mount
the program which will perform the mount.
@cindex unmount, mount option
@cindex Mount option; unmount
@item unmount
the program which will perform the unmount.
@end table
The exit code from these two programs is interpreted as a Unix error
code. As usual, exit code zero indicates success. To execute the
program @i{Amd} splits the string on whitespace to create an array of
substrings. Single quotes @samp{'} can be used to quote whitespace
if that is required in an argument. There is no way to escape or change
the quote character.
To run the program @samp{rvdmount} with a host name and filesystem as
arguments would be specified by @samp{mount:="/etc/rvdmount rvdmount
fserver $@{path@}"}.
The first element in the array is taken as the pathname of the program
to execute. The other members of the array form the argument vector to
be passed to the program, @dfn{including argument zero}. This means
that the split string must have at least two elements. The program is
directly executed by @i{Amd}, not via a shell. This means that scripts
must begin with a @code{#!} interpreter specification.
If a filesystem type is to be heavily used, it may be worthwhile adding
a new filesystem type into @i{Amd}, but for most uses the program
filesystem should suffice.
When the program is run, standard input and standard error are inherited
from the current values used by @i{Amd}. Standard output is a
duplicate of standard error. The value specified with the ``-l''
command line option has no effect on standard error.
@node Symbolic Link Filesystem, Symbolic Link Filesystem II, Program Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Symbolic Link Filesystem (@samp{type:=link})
@cindex Symbolic link filesystem
@cindex Referencing part of the local name space
@cindex Mounting part of the local name space
@cindex How to reference part of the local name space
@cindex link, filesystem type
@cindex symlink, link filesystem type
@cindex Filesystem type; link
Each filesystem type creates a symbolic link to point from the volume
name to the physical mount point. The @samp{link} filesystem does the
same without any other side effects. This allows any part of the
machines name space to be accessed via @i{Amd}.
One common use for the symlink filesystem is @file{/homes} which can be
made to contain an entry for each user which points to their
(auto-mounted) home directory. Although this may seem rather expensive,
it provides a great deal of administrative flexibility.
@noindent
The following option must be defined:
@table @code
@item fs
The value of @var{fs} option specifies the destination of the link, as
modified by the @var{sublink} option. If @var{sublink} is non-null, it
is appended to @code{$@{fs@}}@code{/} and the resulting string is used
as the target.
@end table
The @samp{link} filesystem can be though of as identical to the
@samp{ufs} filesystem but without actually mounting anything.
An example entry might be:
@example
jsp host==charm;type:=link;fs:=/home/charm;sublink:=jsp
@end example
which would return a symbolic link pointing to @file{/home/charm/jsp}.
@node Symbolic Link Filesystem II, Automount Filesystem, Symbolic Link Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Symbolic Link Filesystem II (@samp{type:=linkx})
@cindex Symbolic link filesystem II
@cindex Referencing an existing part of the local name space
@cindex Mounting an existing part of the local name space
@cindex How to reference an existing part of the local name space
@cindex linkx, filesystem type
@cindex symlink, linkx filesystem type
@cindex Filesystem type; linkx
The @samp{linkx} filesystem type is identical to @samp{link} with the
exception that the target of the link must exist. Existence is checked
with the @samp{lstat} system call.
The @samp{linkx} filesystem type is particularly useful for wildcard map
entries. In this case, a list of possible targets can be give and
@i{Amd} will choose the first one which exists on the local machine.
@node Automount Filesystem, Direct Automount Filesystem, Symbolic Link Filesystem II, Filesystem Types
@comment node-name, next, previous, up
@section Automount Filesystem (@samp{type:=auto})
@cindex Automount filesystem
@cindex Map cache types
@cindex Setting map cache parameters
@cindex How to set map cache parameters
@cindex How to start an indirect automount point
@cindex auto, filesystem type
@cindex Filesystem type; auto
@cindex SIGHUP signal
@cindex Map cache synchronising
@cindex Synchronising the map cache
@cindex Map cache options
@cindex Regular expressions in maps
The @dfn{auto} filesystem type creates a new automount point below an
existing automount point. Top-level automount points appear as system
mount points. An automount mount point can also appear as a
sub-directory of an existing automount point. This allows some
additional structure to be added, for example to mimic the mount tree of
another machine.
The following options may be specified:
@table @code
@cindex cache, mount option
@cindex Mount option; cache
@item cache
specifies whether the data in this mount-map should be
cached. The default value is @samp{none}, in which case
no caching is done in order to conserve memory.
However, better performance and reliability can be obtained by caching
some or all of a mount-map.
If the cache option specifies @samp{all},
the entire map is enumerated when the mount point is created.
If the cache option specifies @samp{inc}, caching is done incrementally
as and when data is required.
Some map types do not support cache mode @samp{all}, in which case @samp{inc}
is used whenever @samp{all} is requested.
Caching can be entirely disabled by using cache mode @samp{none}.
If the cache option specifies @samp{regexp} then the entire map will be
enumerated and each key will be treated as an egrep-style regular
expression. The order in which a cached map is searched does not
correspond to the ordering in the source map so the regular expressions
should be mutually exclusive to avoid confusion.
Each mount map type has a default cache type, usually @samp{inc}, which
can be selected by specifying @samp{mapdefault}.
The cache mode for a mount map can only be selected on the command line.
Starting @i{Amd} with the command:
@example
amd /homes hesiod.homes -cache:=inc
@end example
will cause @samp{/homes} to be automounted using the @dfn{Hesiod} name
server with local incremental caching of all successfully resolved names.
All cached data is forgotten whenever @i{Amd} receives a @samp{SIGHUP}
signal and, if cache @samp{all} mode was selected, the cache will be
reloaded. This can be used to inform @i{Amd} that a map has been
updated. In addition, whenever a cache lookup fails and @i{Amd} needs
to examine a map, the map's modify time is examined. If the cache is
out of date with respect to the map then it is flushed as if a
@samp{SIGHUP} had been received.
An additional option (@samp{sync}) may be specified to force @i{Amd} to
check the map's modify time whenever a cached entry is being used. For
example, an incremental, synchronised cache would be created by the
following command:
@example
amd /homes hesiod.homes -cache:=inc,sync
@end example
@item fs
specifies the name of the mount map to use for the new mount point.
Arguably this should have been specified with the @code{$@{rfs@}} option but
we are now stuck with it due to historical accident.
@c %If the string @samp{.} is used then the same map is used;
@c %in addition the lookup prefix is set to the name of the mount point followed
@c %by a slash @samp{/}.
@c %This is the same as specifying @samp{fs:=\$@{map@};pref:=\$@{key@}/}.
@c
@item pref
alters the name that is looked up in the mount map. If
@code{$@{pref@}}, the @dfn{prefix}, is non-null then it is prepended to
the name requested by the kernel @dfn{before} the map is searched.
@end table
The server @samp{dylan.doc.ic.ac.uk} has two user disks:
@samp{/dev/dsk/2s0} and @samp{/dev/dsk/5s0}. These are accessed as
@samp{/home/dylan/dk2} and @samp{/home/dylan/dk5} respectively. Since
@samp{/home} is already an automount point, this naming is achieved with
the following map entries:@refill
@example
dylan type:=auto;fs:=$@{map@};pref:=$@{key@}/
dylan/dk2 type:=ufs;dev:=/dev/dsk/2s0
dylan/dk5 type:=ufs;dev:=/dev/dsk/5s0
@end example
@node Direct Automount Filesystem, Union Filesystem, Automount Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Direct Automount Filesystem (@samp{type:=direct})
@cindex Direct automount filesystem
@cindex How to start a direct automount point
@cindex direct, filesystem type
@cindex Filesystem type; direct
The @dfn{direct} filesystem is almost identical to the automount
filesystem. Instead of appearing to be a directory of mount points, it
appears as a symbolic link to a mounted filesystem. The mount is done
at the time the link is accessed. @xref{Automount Filesystem}, for a
list of required options.
Direct automount points are created by specifying the @samp{direct}
filesystem type on the command line:
@example
amd ... /usr/man auto.direct -type:=direct
@end example
where @samp{auto.direct} would contain an entry such as:
@example
usr/man -type:=nfs;rfs:=/usr/man \
rhost:=man-server1 rhost:=man-server2
@end example
In this example, @samp{man-server1} and @samp{man-server2} are file
servers which export copies of the manual pages. Note that the key
which is looked up is the name of the automount point without the
leading @samp{/}.
@node Union Filesystem, Error Filesystem, Direct Automount Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Union Filesystem (@samp{type:=union})
@cindex Union filesystem
@cindex union, filesystem type
@cindex Filesystem type; union
The @dfn{union} filesystem type allows the contents of several
directories to be merged and made visible in a single directory. This
can be used to overcome one of the major limitations of the Unix mount
mechanism which only allows complete directories to be mounted.
For example, supposing @file{/tmp} and @file{/var/tmp} were to be merged
into a new directory called @file{/mtmp}, with files in @file{/var/tmp}
taking precedence. The following command could be used to achieve this
effect:
@example
amd ... /mtmp union:/tmp:/var/tmp -type:=union
@end example
Currently, the unioned directories must @emph{not} be automounted. That
would cause a deadlock. This seriously limits the current usefulness of
this filesystem type and the problem will be addressed in a future
release of @i{Amd}.
Files created in the union directory are actually created in the last
named directory. This is done by creating a wildcard entry which points
to the correct directory. The wildcard entry is visible if the union
directory is listed, so allowing you to see which directory has
priority.
The files visible in the union directory are computed at the time
@i{Amd} is started, and are not kept uptodate with respect to the
underlying directories. Similarly, if a link is removed, for example
with the @samp{rm} command, it will be lost forever.
@node Error Filesystem, Top-level Filesystem, Union Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Error Filesystem (@samp{type:=error})
@cindex Error filesystem
@cindex error, filesystem type
@cindex Filesystem type; error
The @dfn{error} filesystem type is used internally as a catch-all in
the case where none of the other filesystems was selected, or some other
error occurred.
Lookups and mounts always fail with ``No such file or directory''.
All other operations trivially succeed.
The error filesystem is not directly accessible.
@node Top-level Filesystem, Root Filesystem, Error Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Top-level Filesystem (@samp{type:=toplvl})
@cindex Top level filesystem
@cindex toplvl, filesystem type
@cindex Filesystem type; toplvl
The @dfn{toplvl} filesystems is derived from the @samp{auto} filesystem
and is used to mount the top-level automount nodes. Requests of this
type are automatically generated from the command line arguments and
can also be passed in by using the ``-M'' option of the @dfn{Amq} command.
@node Root Filesystem, Inheritance Filesystem, Top-level Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Root Filesystem
@cindex Root filesystem
@cindex root, filesystem type
@cindex Filesystem type; root
The @dfn{root} (@samp{type:=root}) filesystem type acts as an internal
placeholder onto which @i{Amd} can pin @samp{toplvl} mounts. Only one
node of this type need ever exist and one is created automatically
during startup. The effect of creating a second root node is undefined.
@node Inheritance Filesystem, , Root Filesystem, Filesystem Types
@comment node-name, next, previous, up
@section Inheritance Filesystem
@cindex Inheritance filesystem
@cindex Nodes generated on a restart
@cindex inherit, filesystem type
@cindex Filesystem type; inherit
The @dfn{inheritance} (@samp{type:=inherit}) filesystem is not directly
accessible. Instead, internal mount nodes of this type are
automatically generated when @i{Amd} is started with the ``-r'' option.
At this time the system mount table is scanned to locate any filesystems
which are already mounted. If any reference to these filesystems is
made through @i{Amd} then instead of attempting to mount it, @i{Amd}
simulates the mount and @dfn{inherits} the filesystem. This allows a
new version of @i{Amd} to be installed on a live system simply by
killing the old daemon with @code{SIGTERM} and starting the new one.@refill
This filesystem type is not generally visible externally, but it is
possible that the output from @samp{amq -m} may list @samp{inherit} as
the filesystem type. This happens when an inherit operation cannot
be completed for some reason, usually because a fileserver is down.
@node Run-time Administration, Examples, Filesystem Types, Top
@comment node-name, next, previous, up
@chapter Run-time Administration
@cindex Run-time administration
@cindex Amq command
@menu
* Starting Amd::
* Stopping Amd::
* Controlling Amd::
@end menu
@node Starting Amd, Stopping Amd, Run-time Administration, Run-time Administration
@comment node-name, next, previous, up
@section Starting @i{Amd}
@cindex Starting Amd
@cindex Additions to /etc/rc.local
@cindex /etc/rc.local additions
@cindex /etc/amd.start
@i{Amd} is best started from @samp{/etc/rc.local}:
@example
if [ -f /etc/amd.start ]; then
sh /etc/amd.start; (echo -n ' amd') >/dev/console
fi
@end example
@noindent
The shell script, @samp{amd.start}, contains:
@example
#!/bin/sh -
PATH=/etc:/bin:/usr/bin:/usr/ucb:$PATH export PATH
#
# Either name of logfile or "syslog"
#
LOGFILE=syslog
#LOGFILE=/var/log/amd
#
# Figure out whether domain name is in host name
# If the hostname is just the machine name then
# pass in the name of the local domain so that the
# hostnames in the map are domain stripped correctly.
#
case `hostname` in
*.*) dmn= ;;
*) dmn='-d doc.ic.ac.uk'
esac
#
# Zap earlier log file
#
case "$LOGFILE" in
*/*)
mv "$LOGFILE" "$LOGFILE"-
> "$LOGFILE"
;;
syslog)
: nothing
;;
esac
cd /usr/sbin
#
# -r restart
# -d dmn local domain
# -w wait wait between unmount attempts
# -l log logfile or "syslog"
#
eval ./amd -r $dmn -w 240 -l "$LOGFILE" \
/homes amd.homes -cache:=inc \
/home amd.home -cache:=inc \
/vol amd.vol -cache:=inc \
/n amd.net -cache:=inc
@end example
If the list of automount points and maps is contained in a file or NIS map
it is easily incorporated onto the command line:
@example
...
eval ./amd -r $dmn -w 240 -l "$LOGFILE" `ypcat -k auto.master`
@end example
@node Stopping Amd, Controlling Amd, Starting Amd, Run-time Administration
@comment node-name, next, previous, up
@section Stopping @i{Amd}
@cindex Stopping Amd
@cindex SIGTERM signal
@cindex SIGINT signal
@i{Amd} stops in response to two signals.
@table @samp
@item SIGTERM
causes the top-level automount points to be unmounted and then @i{Amd}
to exit. Any automounted filesystems are left mounted. They can be
recovered by restarting @i{Amd} with the ``-r'' command line option.@refill
@item SIGINT
causes @i{Amd} to attempt to unmount any filesystems which it has
automounted, in addition to the actions of @samp{SIGTERM}. This signal
is primarly used for debugging.@refill
@end table
Actions taken for other signals are undefined.
@node Controlling Amd, , Stopping Amd, Run-time Administration
@comment node-name, next, previous, up
@section Controlling @i{Amd}
@cindex Controlling Amd
@cindex Discovering what is going on at run-time
@cindex Listing currently mounted filesystems
It is sometimes desirable or necessary to exercise external control
over some of @i{Amd}'s internal state. To support this requirement,
@i{Amd} implements an RPC interface which is used by the @dfn{Amq} program.
A variety of information is available.
@i{Amq} generally applies an operation, specified by a single letter option,
to a list of mount points. The default operation is to obtain statistics
about each mount point. This is similar to the output shown above
but includes information about the number and type of accesses to each
mount point.
@menu
* Amq default:: Default command behaviour.
* Amq -f option:: Flushing the map cache.
* Amq -h option:: Controlling a non-local host.
* Amq -m option:: Obtaining mount statistics.
* Amq -M-option:: Mounting a volume.
* Amq -s option:: Obtaining global statistics.
* Amq -u option:: Forcing volumes to time out.
* Amq -v option:: Version information.
* Other Amq options:: Three other special options.
@end menu
@node Amq default, Amq -f option, Controlling Amd, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} default information
With no arguments, @dfn{Amq} obtains a brief list of all existing
mounts created by @i{Amd}. This is different from the list displayed by
@b{df}(1) since the latter only includes system mount points.
@noindent
The output from this option includes the following information:
@itemize @bullet
@item
the automount point,
@item
the filesystem type,
@item
the mount map or mount information,
@item
the internal, or system mount point.
@end itemize
@noindent
For example:
@example
/ root "root" sky:(pid75)
/homes toplvl /usr/local/etc/amd.homes /homes
/home toplvl /usr/local/etc/amd.home /home
/homes/jsp nfs charm:/home/charm /a/charm/home/charm/jsp
/homes/phjk nfs toytown:/home/toytown /a/toytown/home/toytown/ai/phjk
@end example
@noindent
If an argument is given then statistics for that volume name will
be output. For example:
@example
What Uid Getattr Lookup RdDir RdLnk Statfs Mounted@@
/homes 0 1196 512 22 0 30 90/09/14 12:32:55
/homes/jsp 0 0 0 0 1180 0 90/10/13 12:56:58
@end example
@table @code
@item What
the volume name.
@item Uid
ignored.
@item Getattr
the count of NFS @dfn{getattr} requests on this node. This should only be
non-zero for directory nodes.
@item Lookup
the count of NFS @dfn{lookup} requests on this node. This should only be
non-zero for directory nodes.
@item RdDir
the count of NFS @dfn{readdir} requests on this node. This should only
be non-zero for directory nodes.
@item RdLnk
the count of NFS @dfn{readlink} requests on this node. This should be
zero for directory nodes.
@item Statfs
the could of NFS @dfn{statfs} requests on this node. This should only
be non-zero for top-level automount points.
@item Mounted@@
the date and time the volume name was first referenced.
@end table
@node Amq -f option, Amq -h option, Amq default, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -f option
@cindex Flushing the map cache
@cindex Map cache, flushing
The ``-f'' option causes @i{Amd} to flush the internal mount map cache.
This is useful for Hesiod maps since @i{Amd} will not automatically
notice when they have been updated. The map cache can also be
synchronised with the map source by using the @samp{sync} option
(@pxref{Automount Filesystem}).@refill
@node Amq -h option, Amq -m option, Amq -f option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -h option
@cindex Querying an alternate host
By default the local host is used. In an HP-UX cluster the root server
is used since that is the only place in the cluster where @i{Amd} will
be running. To query @i{Amd} on another host the ``-h'' option should
be used.
@node Amq -m option, Amq -M-option, Amq -h option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -m option
The ``-m'' option displays similar information about mounted
filesystems, rather than automount points. The output includes the
following information:
@itemize @bullet
@item
the mount information,
@item
the mount point,
@item
the filesystem type,
@item
the number of references to this filesystem,
@item
the server hostname,
@item
the state of the file server,
@item
any error which has occured.
@end itemize
For example:
@example
"root" truth:(pid602) root 1 localhost is up
hesiod.home /home toplvl 1 localhost is up
hesiod.vol /vol toplvl 1 localhost is up
hesiod.homes /homes toplvl 1 localhost is up
amy:/home/amy /a/amy/home/amy nfs 5 amy is up
swan:/home/swan /a/swan/home/swan nfs 0 swan is up (Permission denied)
ex:/home/ex /a/ex/home/ex nfs 0 ex is down
@end example
When the reference count is zero the filesystem is not mounted but
the mount point and server information is still being maintained
by @i{Amd}.
@node Amq -M-option, Amq -s option, Amq -m option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -M option
The ``-M'' option passes a new map entry to @i{Amd} and waits for it to
be evaluated, possibly causing a mount. For example, the following
command would cause @samp{/home/toytown} on host @samp{toytown} to be
mounted locally on @samp{/mnt/toytown}.
@example
amq -M '/mnt/toytown type:=nfs;rfs:=/home/toytown;rhost:=toytown;fs:=$@{key@}'
@end example
@i{Amd} applies some simple security checks before allowing this
operation. The check tests whether the incoming request is from a
privileged UDP port on the local machine. ``Permission denied'' is
returned if the check fails.
A future release of @i{Amd} will include code to allow the @b{mount}(8)
command to mount automount points:
@example
mount -t amd /vol hesiod.vol
@end example
This will then allow @i{Amd} to be controlled from the standard system
filesystem mount list.
@node Amq -s option, Amq -u option, Amq -M-option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -s option
@cindex Global statistics
@cindex Statistics
The ``-s'' option displays global statistics. If any other options are specified
or any filesystems named then this option is ignored. For example:
@example
requests stale mount mount unmount
deferred fhandles ok failed failed
1054 1 487 290 7017
@end example
@table @samp
@item Deferred requests
are those for which an immediate reply could not be constructed. For
example, this would happen if a background mount was required.
@item Stale filehandles
counts the number of times the kernel passes a stale filehandle to @i{Amd}.
Large numbers indicate problems.
@item Mount ok
counts the number of automounts which were successful.
@item Mount failed
counts the number of automounts which failed.
@item Unmount failed
counts the number of times a filesystem could not be unmounted. Very
large numbers here indicate that the time between unmount attempts
should be increased.
@end table
@node Amq -u option, Amq -v option, Amq -s option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -u option
@cindex Forcing filesystem to time out
@cindex Unmounting a filesystem
The ``-u'' option causes the time-to-live interval of the named mount
points to be expired, thus causing an unmount attempt. This is the only
safe way to unmount an automounted filesystem. It is not possible to
unmount a filesystem which has been mounted with the @samp{nounmount}
flag.
@c The ``-H'' option informs @i{Amd} that the specified mount point has hung -
@c as if its keepalive timer had expired.
@node Amq -v option, Other Amq options, Amq -u option, Controlling Amd
@comment node-name, next, previous, up
@subsection @i{Amq} -v option
@cindex Version information at run-time
The ``-v'' option displays the version of @i{Amd} in a similar way to
@i{Amd}'s ``-v'' option.
@node Other Amq options, , Amq -v option, Controlling Amd
@comment node-name, next, previous, up
@subsection Other @i{Amq} options
Three other operations are implemented. These modify the state of
@i{Amd} as a whole, rather than any particular filesystem. The ``-l'',
``-x'' and ``-D'' options have exactly the same effect as @i{Amd}'s
corresponding command line options. The ``-l'' option is rejected by
@i{Amd} in the current version for obvious security reasons. When
@i{Amd} receives a ``-x''flag it limits the log options being modified
to those which were not enabled at startup. This prevents a user
turning @emph{off} any logging option which was specified at startup,
though any which have been turned off since then can still be turned
off. The ``-D'' option has a similar behaviour.
@node Examples, Internals, Run-time Administration, Top
@comment node-name, next, previous, up
@chapter Examples
@menu
* User Filesystems::
* Home Directories::
* Architecture Sharing::
* Wildcard Names::
* rwho servers::
* /vol::
@end menu
@node User Filesystems, Home Directories, Examples, Examples
@comment node-name, next, previous, up
@section User Filesystems
@cindex User filesystems
@cindex Mounting user filesystems
With more than one fileserver, the directories most frequently
cross-mounted are those containing user home directories. A common
convention used at Imperial College is to mount the user disks under
@t{/home/}@i{machine}.
Typically, the @samp{/etc/fstab} file contained a long list of entries
such as:
@example
@i{machine}:/home/@i{machine} /home/@i{machine} nfs ...
@end example
for each fileserver on the network.
There are numerous problems with this system. The mount list can become
quite large and some of the machines may be down when a system is
booted. When a new fileserver is installed, @samp{/etc/fstab} must be
updated on every machine, the mount directory created and the filesystem
mounted.
In many environments most people use the same few workstations, but
it is convenient to go to a colleague's machine and access your own
files. When a server goes down, it can cause a process on a client
machine to hang. By minimising the mounted filesystems to only include
those actively being used, there is less chance that a filesystem will
be mounted when a server goes down.
The following is a short extract from a map taken from a research fileserver
at Imperial College.
Note the entry for @samp{localhost} which is used for users such as
the operator (@samp{opr}) who have a home directory on most machine as
@samp{/home/localhost/opr}.
@example
/defaults opts:=rw,intr,grpid,nosuid
charm host!=$@{key@};type:=nfs;rhost:=$@{key@};rfs:=/home/$@{key@} \
host==$@{key@};type:=ufs;dev:=/dev/xd0g
#
...
#
localhost type:=link;fs:=$@{host@}
...
#
# dylan has two user disks so have a
# top directory in which to mount them.
#
dylan type:=auto;fs:=$@{map@};pref:=$@{key@}/
#
dylan/dk2 host!=dylan;type:=nfs;rhost:=dylan;rfs:=/home/$@{key@} \
host==dylan;type:=ufs;dev:=/dev/dsk/2s0
#
dylan/dk5 host!=dylan;type:=nfs;rhost:=dylan;rfs:=/home/$@{key@} \
host==dylan;type:=ufs;dev:=/dev/dsk/5s0
...
#
toytown host!=$@{key@};type:=nfs;rhost:=$@{key@};rfs:=/home/$@{key@} \
host==$@{key@};type:=ufs;dev:=/dev/xy1g
...
#
zebedee host!=$@{key@};type:=nfs;rhost:=$@{key@};rfs:=/home/$@{key@} \
host==$@{key@};type:=ufs;dev:=/dev/dsk/1s0
#
# Just for access...
#
gould type:=auto;fs:=$@{map@};pref:=$@{key@}/
gould/staff host!=gould;type:=nfs;rhost:=gould;rfs:=/home/$@{key@}
#
gummo host!=$@{key@};type:=nfs;rhost:=$@{key@};rfs:=/home/$@{key@}
...
@end example
This map is shared by most of the machines listed so on those
systems any of the user disks is accessible via a consistent name.
@i{Amd} is started with the following command
@example
amd /home amd.home
@end example
Note that when mounting a remote filesystem, the @dfn{automounted}
mount point is referenced, so that the filesystem will be mounted if
it is not yet (at the time the remote @samp{mountd} obtains the file handle).
@node Home Directories, Architecture Sharing, User Filesystems, Examples
@comment node-name, next, previous, up
@section Home Directories
@cindex Home directories
@cindex Example of mounting home directories
@cindex Mount home directories
One convention for home directories is to locate them in @samp{/homes}
so user @samp{jsp}'s home directory is @samp{/homes/jsp}. With more
than a single fileserver it is convenient to spread user files across
several machines. All that is required is a mount-map which converts
login names to an automounted directory.
Such a map might be started by the command:
@example
amd /homes amd.homes
@end example
where the map @samp{amd.homes} contained the entries:
@example
/defaults type:=link # All the entries are of type:=link
jsp fs:=/home/charm/jsp
njw fs:=/home/dylan/dk5/njw
...
phjk fs:=/home/toytown/ai/phjk
sjv fs:=/home/ganymede/sjv
@end example
Whenever a login name is accessed in @samp{/homes} a symbolic link
appears pointing to the real location of that user's home directory. In
this example, @samp{/homes/jsp} would appear to be a symbolic link
pointing to @samp{/home/charm/jsp}. Of course, @samp{/home} would also
be an automount point.
This system causes an extra level of symbolic links to be used.
Although that turns out to be relatively inexpensive, an alternative is
to directly mount the required filesystems in the @samp{/homes}
map. The required map is simple, but long, and its creation is best automated.
The entry for @samp{jsp} could be:
@example
jsp -sublink:=$@{key@};rfs:=/home/charm \
host==charm;type:=ufs;dev:=/dev/xd0g \
host!=charm;type:=nfs;rhost:=charm
@end example
This map can become quite big if it contains a large number of entries.
By combining two other features of @i{Amd} it can be greatly simplified.
First the UFS partitions should be mounted under the control of
@samp{/etc/fstab}, taking care that they are mounted in the same place
that @i{Amd} would have automounted them. In most cases this would be
something like @samp{/a/@dfn{host}/home/@dfn{host}} and
@samp{/etc/fstab} on host @samp{charm} would have a line:@refill
@example
/dev/xy0g /a/charm/home/charm 4.2 rw,nosuid,grpid 1 5
@end example
The map can then be changed to:
@example
/defaults type:=nfs;sublink:=$@{key@};opts:=rw,intr,nosuid,grpid
jsp rhost:=charm;rfs:=/home/charm
njw rhost:=dylan;rfs:=/home/dylan/dk5
...
phjk rhost:=toytown;rfs:=/home/toytown;sublink:=ai/$@{key@}
sjv rhost:=ganymede;rfs:=/home/ganymede
@end example
This map operates as usual on a remote machine (@i{ie} @code{$@{host@}}
not equal to @code{$@{rhost@}}). On the machine where the filesystem is
stored (@i{ie} @code{$@{host@}} equal to @code{$@{rhost@}}), @i{Amd}
will construct a local filesystem mount point which corresponds to the
name of the locally mounted UFS partition. If @i{Amd} is started with
the ``-r'' option then instead of attempting an NFS mount, @i{Amd} will
simply inherit the UFS mount (@pxref{Inheritance Filesystem}). If
``-r'' is not used then a loopback NFS mount will be made. This type of
mount is known to cause a deadlock on many systems.
@node Architecture Sharing, Wildcard Names, Home Directories, Examples
@comment node-name, next, previous, up
@section Architecture Sharing
@cindex Architecture sharing
@cindex Sharing a fileserver between architectures
@cindex Architecture dependent volumes
@c %At the moment some of the research machines have sets of software
@c %mounted in @samp{/vol}. This contains subdirectories for \TeX,
@c %system sources, local sources, prolog libraries and so on.
Often a filesystem will be shared by machines of different architectures.
Separate trees can be maintained for the executable images for each
architecture, but it may be more convenient to have a shared tree,
with distinct subdirectories.
A shared tree might have the following structure on the fileserver (called
@samp{fserver} in the example):
@example
local/tex
local/tex/fonts
local/tex/lib
local/tex/bin
local/tex/bin/sun3
local/tex/bin/sun4
local/tex/bin/hp9000
...
@end example
In this example, the subdirectories of @samp{local/tex/bin} should be
hidden when accessed via the automount point (conventionally @samp{/vol}).
A mount-map for @samp{/vol} to achieve this would look like:
@example
/defaults sublink:=$@{/key@};rhost:=fserver;type:=link
tex type:=auto;fs:=$@{map@};pref:=$@{key@}/
tex/fonts host!=fserver;type:=nfs;rfs:=/vol/tex \
host==fserver;fs:=/usr/local/tex
tex/lib host!=fserver;type:=nfs;rfs:=/vol/tex \
host==fserver;fs:=/usr/local/tex
tex/bin -sublink:=$@{/key@}/$@{arch@} host!=fserver;type:=nfs;rfs:=/vol/tex \
host:=fserver;fs:=/usr/local/tex
@end example
When @samp{/vol/tex/bin} is referenced, the current machine architecture
is automatically appended to the path by the @code{$@{sublink@}}
variable. This means that users can have @samp{/vol/tex/bin} in their
@samp{PATH} without concern for architecture dependencies.
@node Wildcard Names, rwho servers, Architecture Sharing, Examples
@comment node-name, next, previous, up
@section Wildcard names & Replicated Servers
By using the wildcard facility, @i{Amd} can @dfn{overlay} an existing
directory with additional entries.
The system files are usually mounted under @samp{/usr}. If instead
@i{Amd} is mounted on @samp{/usr}, additional
names can be overlayed to augment or replace names in the ``master'' @samp{/usr}.
A map to do this would have the form:
@example
local type:=auto;fs:=local-map
share type:=auto;fs:=share-map
* -type:=nfs;rfs:=/export/exec/$@{arch@};sublink:="$@{key@}" \
rhost:=fserv1 rhost:=fserv2 rhost:=fserv3
@end example
Note that the assignment to @code{$@{sublink@}} is surrounded by double
quotes to prevent the incoming key from causing the map to be
misinterpreted. This map has the effect of directing any access to
@samp{/usr/local} or @samp{/usr/share} to another automount point.
In this example, it is assumed that the @samp{/usr} files are replicated
on three fileservers: @samp{fserv1}, @samp{fserv2} and @samp{fserv3}.
For any references other than to @samp{local} and @samp{share} one of
the servers is used and a symbolic link to
@t{$@{autodir@}/$@{rhost@}/export/exec/$@{arch@}/@i{whatever}} is
returned once an appropriate filesystem has been mounted.@refill
@node rwho servers, /vol, Wildcard Names, Examples
@comment node-name, next, previous, up
@section @samp{rwho} servers
@cindex rwho servers
@cindex Architecture specific mounts
@cindex Example of architecture specific mounts
The @samp{/usr/spool/rwho} directory is a good candidate for automounting.
For efficiency reasons it is best to capture the rwho data on a small
number of machines and then mount that information onto a large number
of clients. The data written into the rwho files is byte order dependent
so only servers with the correct byte ordering can be used by a client:
@example
/defaults type:=nfs
usr/spool/rwho -byte==little;rfs:=/usr/spool/rwho \
rhost:=vaxA rhost:=vaxB \
|| -rfs:=/usr/spool/rwho \
rhost:=sun4 rhost:=hp300
@end example
@node /vol, , rwho servers, Examples
@comment node-name, next, previous, up
@section @samp{/vol}
@cindex /vol
@cindex Catch-all mount point
@cindex Generic volume name
@samp{/vol} is used as a catch-all for volumes which do not have other
conventional names.
Below is part of the @samp{/vol} map for the domain @samp{doc.ic.ac.uk}.
The @samp{r+d} tree is used for new or experimental software that needs
to be available everywhere without installing it on all the fileservers.
Users wishing to try out the new software then simply include
@samp{/vol/r+d/@{bin,ucb@}} in their path.@refill
The main tree resides on one host @samp{gould.doc.ic.ac.uk}, which has
different @samp{bin}, @samp{etc}, @samp{lib} and @samp{ucb}
sub-directories for each machine architecture. For example,
@samp{/vol/r+d/bin} for a Sun-4 would be stored in the sub-directory
@samp{bin/sun4} of the filesystem @samp{/usr/r+d}. When it was accessed
a symbolic link pointing to @samp{/a/gould/usr/r+d/bin/sun4} would be
returned.@refill
@example
/defaults type:=nfs;opts:=rw,grpid,nosuid,intr,soft
wp -opts:=rw,grpid,nosuid;rhost:=charm \
host==charm;type:=link;fs:=/usr/local/wp \
host!=charm;type:=nfs;rfs:=/vol/wp
...
#
src -opts:=rw,grpid,nosuid;rhost:=charm \
host==charm;type:=link;fs:=/usr/src \
host!=charm;type:=nfs;rfs:=/vol/src
#
r+d type:=auto;fs:=$@{map@};pref:=r+d/
# per architecture bin,etc,lib&ucb...
r+d/bin rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}/$@{arch@}
r+d/etc rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}/$@{arch@}
r+d/include rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}
r+d/lib rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}/$@{arch@}
r+d/man rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}
r+d/src rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}
r+d/ucb rhost:=gould.doc.ic.ac.uk;rfs:=/usr/r+d;sublink:=$@{/key@}/$@{arch@}
# hades pictures
pictures -opts:=rw,grpid,nosuid;rhost:=thpfs \
host==thpfs;type:=link;fs:=/nbsd/pictures \
host!=thpfs;type:=nfs;rfs:=/nbsd;sublink:=pictures
# hades tools
hades -opts:=rw,grpid,nosuid;rhost:=thpfs \
host==thpfs;type:=link;fs:=/nbsd/hades \
host!=thpfs;type:=nfs;rfs:=/nbsd;sublink:=hades
# bsd tools for hp.
bsd -opts:=rw,grpid,nosuid;arch==hp9000;rhost:=thpfs \
host==thpfs;type:=link;fs:=/nbsd/bsd \
host!=thpfs;type:=nfs;rfs:=/nbsd;sublink:=bsd
@end example
@node Internals, Acknowledgements & Trademarks, Examples, Top
@comment node-name, next, previous, up
@chapter Internals
@menu
* Log Messages::
@end menu
@node Log Messages, , Internals, Internals
@comment node-name, next, previous, up
@section Log Messages
In the following sections a brief explanation is given of some of the
log messages made by @i{Amd}. Where the message is in @samp{typewriter}
font, it corresponds exactly to the message produced by @i{Amd}. Words
in @dfn{italic} are replaced by an appropriate string. Variables,
@code{$@{var@}}, indicate that the value of the appropriate variable is
output.
Log messages are either sent direct to a file,
or logged via the @b{syslog}(3) mechanism.
Messages are logged with facility @samp{LOG_DAEMON} when using @b{syslog}(3).
In either case, entries in the file are of the form:
@example
@i{date-string} @i{hostname} @t{amd[}@i{pid}@t{]} @i{message}
@end example
@menu
* Fatal errors::
* Info messages::
@end menu
@node Fatal errors, Info messages, Log Messages, Log Messages
@comment node-name, next, previous, up
@subsection Fatal errors
@i{Amd} attempts to deal with unusual events. Whenever it is not
possible to deal with such an error, @i{Amd} will log an appropriate
message and, if it cannot possibly continue, will either exit or abort.
These messages are selected by @samp{-x fatal} on the command line.
When @b{syslog}(3) is being used, they are logged with level
@samp{LOG_FATAL}. Even if @i{Amd} continues to operate it is likely to
remain in a precarious state and should be restarted at the earliest
opportunity.
@table @asis
@item @t{Attempting to inherit not-a-filesystem}
The prototype mount point created during a filesystem restart did not
contain a reference to the restarted filesystem. This erorr ``should
never happen''.
@item @t{Can't bind to domain "@i{NIS-domain}"}
A specific NIS domain was requested on the command line, but no server
for that domain is available on the local net.
@item @t{Can't determine IP address of this host (@i{hostname})}
When @i{Amd} starts it determines its own IP address. If this lookup
fails then @i{Amd} cannot continue. The hostname it looks up is that
obtained returned by @b{gethostname}(2) system call.
@item @t{Can't find root file handle for @i{automount point}}
@i{Amd} creates its own file handles for the automount points. When it
mounts itself as a server, it must pass these file handles to the local
kernel. If the filehandle is not obtainable the mount point is ignored.
This error ``should never happen''.
@item @t{Must be root to mount filesystems (euid = @i{euid})}
To prevent embarrassment, @i{Amd} makes sure it has appropriate system
privileges. This amounts to having an euid of 0. The check is made
after argument processing complete to give non-root users a chance to
access the ``-v'' option.
@item @t{No work to do - quitting}
No automount points were given on the command line and so there is no
work to do.
@item @t{Out of memory in realloc}
While attempting to realloc some memory, the memory space available to
@i{Amd} was exhausted. This is an unrecoverable error.
@item @t{Out of memory}
While attempting to malloc some memory, the memory space available to
@i{Amd} was exhausted. This is an unrecoverable error.
@item @t{cannot create rpc/udp service}
Either the NFS or AMQ endpoint could not be created.
@item @t{gethostname:} @i{description}
The @b{gethostname}(2) system call failed during startup.
@item @t{host name is not set}
The @b{gethostname}(2) system call returned a zero length host name.
This can happen if @i{Amd} is started in single user mode just after
booting the system.
@item @t{ifs_match called!}
An internal error occurred while restarting a pre-mounted filesystem.
This error ``should never happen''.
@item @t{mount_afs:} @i{description}
An error occured while @i{Amd} was mounting itself.
@item @t{run_rpc failed}
Somehow the main NFS server loop failed. This error ``should never
happen''.
@item @t{unable to free rpc arguments in amqprog_1}
The incoming arguments to the AMQ server could not be free'ed.
@item @t{unable to free rpc arguments in nfs_program_1}
The incoming arguments to the NFS server could not be free'ed.
@item @t{unable to register (AMQ_PROGRAM, AMQ_VERSION, udp)}
The AMQ server could not be registered with the local portmapper or the
internal RPC dispatcher.
@item @t{unable to register (NFS_PROGRAM, NFS_VERSION, 0)}
The NFS server could not be registered with the internal RPC dispatcher.
@end table
@node Info messages, , Fatal errors, Log Messages
@comment node-name, next, previous, up
@subsection Info messages
@i{Amd} generates information messages to record state changes. These
messages are selected by @samp{-x info} on the command line. When
@b{syslog}(3) is being used, they are logged with level @samp{LOG_INFO}.
The messages listed below can be generated and are in a format suitable
for simple statistical analysis. @dfn{mount-info} is the string
that is displayed by @dfn{Amq} in its mount information column and
placed in the system mount table.
@table @asis
@item @t{mount of "@t{$@{@i{path}@}}" on @t{$@{@i{fs}@}} timed out}
Attempts to mount a filesystem for the given automount point have failed
to complete within 30 seconds.
@item @t{"@t{$@{@i{path}@}}" forcibly timed out}
An automount point has been timed out by the @i{Amq} command.
@item @t{restarting @i{mount-info} on @t{$@{@i{fs}@}}}
A pre-mounted file system has been noted.
@item @t{"@t{$@{@i{path}@}}" has timed out}
No access to the automount point has been made within the timeout
period.
@item @t{file server @t{$@{@i{rhost}@}} is down - timeout of "@t{$@{@i{path}@}}" ignored}
An automount point has timed out, but the corresponding file server is
known to be down. This message is only produced once for each mount
point for which the server is down.
@item @t{Re-synchronizing cache for map @t{$@{@i{map}@}}}
The named map has been modified and the internal cache is being re-synchronized.
@item @t{Filehandle denied for "@t{$@{@i{rhost}@}}:@t{$@{@i{rfs}@}}"}
The mount daemon refused to return a file handle for the requested filesystem.
@item @t{Filehandle error for "$@{@i{rhost}@}:$@{@i{rfs}@}":} @i{description}
The mount daemon gave some other error for the requested filesystem.
@item @t{file server @t{$@{@i{rhost}@}} type nfs starts up}
A new NFS file server has been referenced and is known to be up.
@item @t{file server @t{$@{@i{rhost}@}} type nfs starts down}
A new NFS file server has been referenced and is known to be down.
@item @t{file server @t{$@{@i{rhost}@}} type nfs is up}
An NFS file server that was previously down is now up.
@item @t{file server @t{$@{@i{rhost}@}} type nfs is down}
An NFS file server that was previously up is now down.
@item @t{Finishing with status @i{exit-status}}
@i{Amd} is about to exit with the given exit status.
@item @t{@i{mount-info} mounted fstype @t{$@{@i{type}@}} on @t{$@{@i{fs}@}}}
A new file system has been mounted.
@item @t{@i{mount-info} restarted fstype @t{$@{@i{type}@}} on @t{$@{@i{fs}@}}}
@i{Amd} is using a pre-mounted filesystem to satisfy a mount request.
@item @t{@i{mount-info} unmounted fstype @t{$@{@i{type}@}} from @t{$@{@i{fs}@}}}
A file system has been unmounted.
@item @t{@i{mount-info} unmounted fstype @t{$@{@i{type}@}} from @t{$@{@i{fs}@}} link @t{$@{@i{fs}@}}/@t{$@{@i{sublink}@}}}
A file system of which only a sub-directory was in use has been unmounted.
@end table
@node Acknowledgements & Trademarks, Index, Internals, Top
@comment node-name, next, previous, up
@unnumbered Acknowledgements & Trademarks
Thanks to the Formal Methods Group at Imperial College for
suffering patiently while @i{Amd} was being developed on their machines.
Thanks to the many people who have helped with the development of
@i{Amd}, especially Piete Brooks at the Cambridge University Computing
Lab for many hours of testing, experimentation and discussion.
@itemize @bullet
@item
@b{DEC}, @b{VAX} and @b{Ultrix} are registered trademarks of Digital
Equipment Corporation.
@item
@b{AIX} and @b{IBM} are registered trademarks of International Business
Machines Corporation.
@item
@b{Sun}, @b{NFS} and @b{SunOS} are registered trademarks of Sun
Microsystems, Inc.
@item
@b{Unix} is a registered trademark of AT&T Unix Systems Laboratories
in the USA and other countries.
@end itemize
@node Index, Intro, Acknowledgements & Trademarks, Top
@comment node-name, next, previous, up
@unnumbered Index
@printindex cp
@contents
@bye