1 .TH EXPORTS 5 "28 October 1999"
4 exports \- NFS file systems being exported (for Kernel based NFS)
10 serves as the access control list for file systems which may be
11 exported to NFS clients. It is used by
13 to give information to
15 and to the kernel based NFS file server daemon
18 The file format is similar to the SunOS
20 file, except that several additional options are permitted. Each line
21 contains an export point and a list of machine or netgroup names allowed
22 to mount the file system at that point. An optional parenthesized list
23 of export parameters may follow each machine name. Blank lines are
24 ignored, and a # introduces a comment to the end of the line. Entries may
25 be continued across newlines using a backslash.
27 .SS Machine Name Formats
28 NFS clients may be specified in a number of ways:
30 This is the most common format. You may specify a host either by an
31 abbreviated name recognized be the resolver, the fully qualified domain
32 name, or an IP address.
34 NIS netgroups may be given as
36 Only the host part of each
37 netgroup members is consider in checking for membership. Empty host
38 parts or those containing a single dash (\-) are ignored.
40 Machine names may contain the wildcard characters \fI*\fR and \fI?\fR.
41 This can be used to make the \fIexports\fR file more compact; for instance,
42 \fI*.cs.foo.edu\fR matches all hosts in the domain \fIcs.foo.edu\fR. However,
43 these wildcard characters do not match the dots in a domain name, so the
44 above pattern does not include hosts such as \fIa.b.cs.foo.edu\fR.
46 You can also export directories to all hosts on an IP (sub-) network
47 simultaneously. This is done by specifying an IP address and netmask pair
52 '''This is a special ``hostname'' that identifies the given directory name
53 '''as the public root directory (see the section on WebNFS in
55 '''for a discussion of WebNFS and the public root handle). When using this
58 '''must be the only entry on this line, and must have no export options
59 '''associated with it. Note that this does
61 '''actually export the named directory; you still have to set the exports
62 '''options in a separate entry.
64 '''The public root path can also be specified by invoking
67 '''.B \-\-public\-root
68 '''option. Multiple specifications of a public root will be ignored.
72 understands the following export options:
75 This option requires that requests originate on an internet port less
76 than IPPORT_RESERVED (1024). This option is on by default. To turn it
81 Allow both read and write requests on this NFS volume. The
82 default is to disallow any request which changes the filesystem.
83 This can also be made explicit by using
88 This option requests that all file writes be committed to disc before
89 the write request completes. This is required for complete safety of
90 data in the face of a server crash, but incurs a performance hit.
91 The default is to allow the server to write the data out whenever it
92 is ready. This can be explicitly requested with the
96 This option only has effect if
98 is also set. The NFS server will normally delay committing a write request
99 to disc slightly if it suspects that another related write request may be in
100 progress or may arrive soon. This allows multiple write requests to
101 be committed to disc with the one operation which can improve
102 performance. If an NFS server received mainly small unrelated
103 requests, this behaviour could actually reduce performance, so
105 is available to turn it off.
106 The default can be explicitly requested with the
110 This option is based on the option of the same name provided in IRIX
111 NFS. Normally, if a server exports two filesystems one of which is
112 mounted on the other, then the client will have to mount both
113 filesystems explicitly to get access to them. If it just mounts the
114 parent, it will see an empty directory at the place where the other
115 filesystem is mounted. That filesystem is "hidden".
119 option on a filesystem causes it not to be hidden, and an
120 appropriately authorised client will be able to move from the parent to
121 that filesystem without noticing the change.
123 However, some NFS clients do not cope well with this situation as, for
124 instance, it is then possible for two files in the one apparent
125 filesystem to have the same inode number.
129 option is currently only effective on
131 exports. It does not work reliably with netgroup, subnet, or wildcard
134 This option can be very useful in some situations, but it should be
135 used with due care, and only after confirming that the client system
136 copes with the situation effectively.
138 The option can be explicitly disabled with
142 This option disables subtree checking, which has mild security
143 implications, but can improve reliability is some circumstances.
145 If a subdirectory of a filesystem is exported, but the whole
146 filesystem isn't then whenever a NFS request arrives, the server must
147 check not only that the accessed file is in the appropriate filesystem
148 (which is easy) but also that it is in the exported tree (which is
149 harder). This check is called the
152 In order to perform this check, the server must include some
153 information about the location of the file in the "filehandle" that is
154 given to the client. This can cause problems with accessing files that
155 are renamed while a client has them open (though in many simple cases
158 subtree checking is also used to make sure that files inside
159 directories to which only root has access can only be accessed if the
160 filesystem is exported with
162 (see below), even the file itself allows more general access.
164 As a general guide, a home directory filesystem, which is normally
165 exported at the root and may see lots of file renames, should be
166 exported with subtree checking disabled. A filesystem which is mostly
167 readonly, and at least doesn't see many file renames (e.g. /usr or
168 /var) and for which subdirectories may be exported, should probably be
169 exported with subtree checks enabled.
171 The default of having subtree checks enabled, can be explicitly
179 This option (the two names are synonymous) tells the NFS server not to require authentication of
180 locking requests (i.e. requests which use the NLM protocol). Normally
181 the NFS server will require a lock request to hold a credential for a
182 user who has read access to the file. With this flag no access checks
185 Early NFS client implementations did not send credentials with lock
186 requests, and many current NFS clients still exist which are based on
187 the old implementations. Use this flag if you find that you can only
188 lock files which are world readable.
190 The default behaviour of requiring authentication for NLM requests can
191 be explicitly requested with either of the synonymous
198 '''This makes everything below the directory inaccessible for the named
199 '''client. This is useful when you want to export a directory hierarchy to
200 '''a client, but exclude certain subdirectories. The client's view of a
201 '''directory flagged with noaccess is very limited; it is allowed to read
202 '''its attributes, and lookup `.' and `..'. These are also the only entries
203 '''returned by a readdir.
206 '''Convert absolute symbolic links (where the link contents start with a
207 '''slash) into relative links by prepending the necessary number of ../'s
208 '''to get from the directory containing the link to the root on the
209 '''server. This has subtle, perhaps questionable, semantics when the file
210 '''hierarchy is not mounted at its root.
213 '''Leave all symbolic link as they are. This is the default operation.
217 bases its access control to files on the server machine on the uid and
218 gid provided in each NFS RPC request. The normal behavior a user would
219 expect is that she can access her files on the server just as she would
220 on a normal file system. This requires that the same uids and gids are
221 used on the client and the server machine. This is not always true, nor
222 is it always desirable.
224 Very often, it is not desirable that the root user on a client machine
225 is also treated as root when accessing files on the NFS server. To this
226 end, uid 0 is normally mapped to a different id: the so-called
229 uid. This mode of operation (called `root squashing') is the default,
230 and can be turned off with
235 '''tries to obtain the anonymous uid and gid by looking up user
237 '''in the password file at startup time. If it isn't found, a uid and gid
239 chooses a uid and gid
240 of -2 (i.e. 65534) for squashed access. These values can also be overridden by
242 .IR anonuid " and " anongid
245 '''In addition to this,
247 '''lets you specify arbitrary uids and gids that should be mapped to user
249 Finally, you can map all user requests to the
250 anonymous uid by specifying the
251 .IR all_squash " option.
253 '''For the benefit of installations where uids differ between different
256 '''provides several mechanism to dynamically map server uids to client
257 '''uids and vice versa: static mapping files, NIS-based mapping, and
262 '''mapping is enabled with the
264 '''option, and uses the UGID RPC protocol. For this to work, you have to run
267 '''mapping daemon on the client host. It is the least secure of the three methods,
268 '''because by running
270 '''everybody can query the client host for a list of valid user names. You
271 '''can protect yourself by restricting access to
273 '''to valid hosts only. This can be done by entering the list of valid
278 '''file. The service name is
280 '''For a description of the file's syntax, please read
281 '''.IR hosts_access (5).
283 '''Static mapping is enabled by using the
285 '''option, which takes a file name as an argument that describes the mapping.
286 '''NIS-based mapping queries the client's NIS server to obtain a mapping from
287 '''user and group names on the server host to user and group names on the
290 Here's the complete list of mapping options:
293 Map requests from uid/gid 0 to the anonymous uid/gid. Note that this does
294 not apply to any other uids that might be equally sensitive, such as user
298 Turn off root squashing. This option is mainly useful for diskless clients.
300 '''.IR squash_uids " and " squash_gids
301 '''This option specifies a list of uids or gids that should be subject to
302 '''anonymous mapping. A valid list of ids looks like this:
304 '''.IR squash_uids=0-15,20,25-50
306 '''Usually, your squash lists will look a lot simpler.
309 Map all uids and gids to the anonymous user. Useful for NFS-exported
310 public FTP directories, news spool directories, etc. The opposite option
313 which is the default setting.
316 '''This option turns on dynamic uid/gid mapping. Each uid in an NFS request
317 '''will be translated to the equivalent server uid, and each uid in an
318 '''NFS reply will be mapped the other way round. This option requires that
320 '''runs on the client host. The default setting is
321 '''.IR map_identity ,
322 '''which leaves all uids untouched. The normal squash options apply regardless
323 '''of whether dynamic mapping is requested or not.
326 '''This option enables static mapping. It specifies the name of the file
327 '''that describes the uid/gid mapping, e.g.
329 '''.IR map_static=/etc/nfs/foobar.map
331 '''The file's format looks like this
335 '''# Mapping for client foobar:
337 '''uid 0-99 - # squash these
338 '''uid 100-500 1000 # map 100-500 to 1000-1400
339 '''gid 0-49 - # squash these
340 '''gid 50-100 700 # map 50-100 to 700-750
344 '''This option enables NIS-based uid/gid mapping. For instance, when
345 '''the server encounters the uid 123 on the server, it will obtain the
346 '''login name associated with it, and contact the NFS client's NIS server
347 '''to obtain the uid the client associates with the name.
349 '''In order to do this, the NFS server must know the client's NIS domain.
350 '''This is specified as an argument to the
354 '''.I map_nis=foo.com
356 '''Note that it may not be sufficient to simply specify the NIS domain
357 '''here; you may have to take additional actions before
359 '''is actually able to contact the server. If your distribution uses
360 '''the NYS library, you can specify one or more NIS servers for the
361 '''client's domain in
362 '''.IR /etc/yp.conf .
363 '''If you are using a different NIS library, you may have to obtain a
366 '''daemon that can be configured via
369 .IR anonuid " and " anongid
370 These options explicitly set the uid and gid of the anonymous account.
371 This option is primarily useful for PC/NFS clients, where you might want
372 all requests appear to be from one user. As an example, consider the
375 in the example section below, which maps all requests to uid 150 (which
376 is supposedly that of user joe).
382 # sample /etc/exports file
383 / master(rw) trusty(rw,no_root_squash)
384 /projects proj*.local.domain(rw)
385 /usr *.local.domain(ro) @trusted(rw)
386 /home/joe pc001(rw,all_squash,anonuid=150,anongid=100)
387 /pub (ro,insecure,all_squash)
388 '''/pub/private (noaccess)
391 The first line exports the entire filesystem to machines master and trusty.
392 In addition to write access, all uid squashing is turned off for host
393 trusty. The second and third entry show examples for wildcard hostnames
394 and netgroups (this is the entry `@trusted'). The fourth line shows the
395 entry for the PC/NFS client discussed above. Line 5 exports the
396 public FTP directory to every host in the world, executing all requests
397 under the nobody account. The
399 option in this entry also allows clients with NFS implementations that
400 don't use a reserved port for NFS.
401 ''' The last line denies all NFS clients
402 '''access to the private directory.
404 '''Unlike other NFS server implementations, this
406 '''allows you to export both a directory and a subdirectory thereof to
407 '''the same host, for instance
408 '''.IR /usr " and " /usr/X11R6 .
409 '''In this case, the mount options of the most specific entry apply. For
410 '''instance, when a user on the client host accesses a file in
412 '''the mount options given in the
414 '''entry apply. This is also true when the latter is a wildcard or netgroup
419 '''An error parsing the file is reported using syslogd(8) as level NOTICE from
420 '''a DAEMON whenever nfsd(8) or mountd(8) is started up. Any unknown
421 '''host is reported at that time, but often not all hosts are not yet known
422 '''to named(8) at boot time, thus as hosts are found they are reported
423 '''with the same syslogd(8) parameters.