3 .TH exports 5 "31 December 2009"
5 exports \- NFS server export table
9 contains a table of local physical file systems on an NFS server
10 that are accessible to NFS clients.
11 The contents of the file are maintained by the server's system
14 Each file system in this table has a list of options and an
18 to give information to
21 The file format is similar to the SunOS
23 file. Each line contains an export point and a whitespace-separated list
24 of clients allowed to mount the file system at that point. Each listed
25 client may be immediately followed by a parenthesized, comma-separated
26 list of export options for that client. No whitespace is permitted
27 between a client and its option list.
29 Also, each line may have one or more specifications for default options
30 after the path name, in the form of a dash ("\-") followed by an option
31 list. The option list is used for all subsequent exports on that line
34 Blank lines are ignored. A pound sign ("#") introduces a comment to the
35 end of the line. Entries may be continued across newlines using a
36 backslash. If an export name contains spaces it should be quoted using
37 double quotes. You can also specify spaces or other unusual character in
38 the export name using a backslash followed by the character code as three
41 To apply changes to this file, run
43 or restart the NFS server.
45 .SS Machine Name Formats
46 NFS clients may be specified in a number of ways:
48 This is the most common format. You may specify a host either by an
49 abbreviated name recognized be the resolver, the fully qualified domain
50 name, or an IP address.
52 NIS netgroups may be given as
54 Only the host part of each
55 netgroup members is consider in checking for membership. Empty host
56 parts or those containing a single dash (\-) are ignored.
58 Machine names may contain the wildcard characters \fI*\fR and \fI?\fR.
59 This can be used to make the \fIexports\fR file more compact; for instance,
60 \fI*.cs.foo.edu\fR matches all hosts in the domain
61 \fIcs.foo.edu\fR. As these characters also match the dots in a domain
62 name, the given pattern will also match all hosts within any subdomain
65 You can also export directories to all hosts on an IP (sub-) network
66 simultaneously. This is done by specifying an IP address and netmask pair
69 where the netmask can be specified in dotted-decimal format, or as a
70 contiguous mask length.
71 For example, either `/255.255.252.0' or `/22' appended
72 to the network base IPv4 address results in identical subnetworks with 10 bits of
73 host. Wildcard characters generally do not work on IP addresses, though they
74 may work by accident when reverse DNS lookups fail.
77 '''This is a special ``hostname'' that identifies the given directory name
78 '''as the public root directory (see the section on WebNFS in
80 '''for a discussion of WebNFS and the public root handle). When using this
83 '''must be the only entry on this line, and must have no export options
84 '''associated with it. Note that this does
86 '''actually export the named directory; you still have to set the exports
87 '''options in a separate entry.
89 '''The public root path can also be specified by invoking
92 '''.B \-\-public\-root
93 '''option. Multiple specifications of a public root will be ignored.
95 .SS RPCSEC_GSS security
96 You may use the special strings "gss/krb5", "gss/krb5i", or "gss/krb5p"
97 to restrict access to clients using rpcsec_gss security. However, this
98 syntax is deprecated; on linux kernels since 2.6.23, you should instead
99 use the "sec=" export option:
102 The sec= option, followed by a colon-delimited list of security flavors,
103 restricts the export to clients using those flavors. Available security
104 flavors include sys (the default--no cryptographic security), krb5
105 (authentication only), krb5i (integrity protection), and krb5p (privacy
106 protection). For the purposes of security flavor negotiation, order
107 counts: preferred flavors should be listed first. The order of the sec=
108 option with respect to the other options does not matter, unless you
109 want some options to be enforced differently depending on flavor.
110 In that case you may include multiple sec= options, and following options
111 will be enforced only for access using flavors listed in the immediately
112 preceding sec= option. The only options that are permitted to vary in
113 this way are ro, rw, no_root_squash, root_squash, and all_squash.
117 understands the following export options:
120 This option requires that requests originate on an Internet port less
121 than IPPORT_RESERVED (1024). This option is on by default. To turn it
126 Allow both read and write requests on this NFS volume. The
127 default is to disallow any request which changes the filesystem.
128 This can also be made explicit by using
133 This option allows the NFS server to violate the NFS protocol and
134 reply to requests before any changes made by that request have been
135 committed to stable storage (e.g. disc drive).
137 Using this option usually improves performance, but at the cost that
138 an unclean server restart (i.e. a crash) can cause data to be lost or
143 Reply to requests only after the changes have been committed to stable
148 In releases of nfs-utils up to and including 1.0.0, this option was the
149 default. In all releases after 1.0.0,
153 must be explicitly requested if needed.
154 To help make system administrators aware of this change,
156 will issue a warning if neither
163 This option has no effect if
165 is also set. The NFS server will normally delay committing a write request
166 to disc slightly if it suspects that another related write request may be in
167 progress or may arrive soon. This allows multiple write requests to
168 be committed to disc with the one operation which can improve
169 performance. If an NFS server received mainly small unrelated
170 requests, this behaviour could actually reduce performance, so
172 is available to turn it off.
173 The default can be explicitly requested with the
177 This option is based on the option of the same name provided in IRIX
178 NFS. Normally, if a server exports two filesystems one of which is
179 mounted on the other, then the client will have to mount both
180 filesystems explicitly to get access to them. If it just mounts the
181 parent, it will see an empty directory at the place where the other
182 filesystem is mounted. That filesystem is "hidden".
186 option on a filesystem causes it not to be hidden, and an
187 appropriately authorised client will be able to move from the parent to
188 that filesystem without noticing the change.
190 However, some NFS clients do not cope well with this situation as, for
191 instance, it is then possible for two files in the one apparent
192 filesystem to have the same inode number.
196 option is currently only effective on
198 exports. It does not work reliably with netgroup, subnet, or wildcard
201 This option can be very useful in some situations, but it should be
202 used with due care, and only after confirming that the client system
203 copes with the situation effectively.
205 The option can be explicitly disabled with
209 This option is similar to
211 but it makes it possible for clients to move from the filesystem marked
212 with crossmnt to exported filesystems mounted on it. Thus when a child
213 filesystem "B" is mounted on a parent "A", setting crossmnt on "A" has
214 the same effect as setting "nohide" on B.
217 This option disables subtree checking, which has mild security
218 implications, but can improve reliability in some circumstances.
220 If a subdirectory of a filesystem is exported, but the whole
221 filesystem isn't then whenever a NFS request arrives, the server must
222 check not only that the accessed file is in the appropriate filesystem
223 (which is easy) but also that it is in the exported tree (which is
224 harder). This check is called the
227 In order to perform this check, the server must include some
228 information about the location of the file in the "filehandle" that is
229 given to the client. This can cause problems with accessing files that
230 are renamed while a client has them open (though in many simple cases
233 subtree checking is also used to make sure that files inside
234 directories to which only root has access can only be accessed if the
235 filesystem is exported with
237 (see below), even if the file itself allows more general access.
239 As a general guide, a home directory filesystem, which is normally
240 exported at the root and may see lots of file renames, should be
241 exported with subtree checking disabled. A filesystem which is mostly
242 readonly, and at least doesn't see many file renames (e.g. /usr or
243 /var) and for which subdirectories may be exported, should probably be
244 exported with subtree checks enabled.
246 The default of having subtree checks enabled, can be explicitly
250 From release 1.1.0 of nfs-utils onwards, the default will be
252 as subtree_checking tends to cause more problems than it is worth.
253 If you genuinely require subtree checking, you should explicitly put
256 file. If you put neither option,
258 will warn you that the change is pending.
264 This option (the two names are synonymous) tells the NFS server not to require authentication of
265 locking requests (i.e. requests which use the NLM protocol). Normally
266 the NFS server will require a lock request to hold a credential for a
267 user who has read access to the file. With this flag no access checks
270 Early NFS client implementations did not send credentials with lock
271 requests, and many current NFS clients still exist which are based on
272 the old implementations. Use this flag if you find that you can only
273 lock files which are world readable.
275 The default behaviour of requiring authentication for NLM requests can
276 be explicitly requested with either of the synonymous
282 On some specially patched kernels, and when exporting filesystems that
283 support ACLs, this option tells
285 not to reveal ACLs to clients, so
286 they will see only a subset of actual permissions on the given file
287 system. This option is safe for filesystems used by NFSv2 clients and
288 old NFSv3 clients that perform access decisions locally. Current
289 NFSv3 clients use the ACCESS RPC to perform all access decisions on
290 the server. Note that the
292 option only has effect on kernels specially patched to support it, and
293 when exporting filesystems with ACL support. The default is to export
294 with ACL support (i.e. by default,
300 '''This makes everything below the directory inaccessible for the named
301 '''client. This is useful when you want to export a directory hierarchy to
302 '''a client, but exclude certain subdirectories. The client's view of a
303 '''directory flagged with noaccess is very limited; it is allowed to read
304 '''its attributes, and lookup `.' and `..'. These are also the only entries
305 '''returned by a readdir.
308 '''Convert absolute symbolic links (where the link contents start with a
309 '''slash) into relative links by prepending the necessary number of ../'s
310 '''to get from the directory containing the link to the root on the
311 '''server. This has subtle, perhaps questionable, semantics when the file
312 '''hierarchy is not mounted at its root.
315 '''Leave all symbolic link as they are. This is the default operation.
321 This option makes it possible to only export a directory if it has
322 successfully been mounted.
323 If no path is given (e.g.
324 .IR mountpoint " or " mp )
325 then the export point must also be a mount point. If it isn't then
326 the export point is not exported. This allows you to be sure that the
327 directory underneath a mountpoint will never be exported by accident
328 if, for example, the filesystem failed to mount due to a disc error.
330 If a path is given (e.g.
331 .IR mountpoint= "/path or " mp= /path)
332 then the nominated path must be a mountpoint for the exportpoint to be
336 .IR fsid= num|root|uuid
337 NFS needs to be able to identify each filesystem that it exports.
338 Normally it will use a UUID for the filesystem (if the filesystem has
339 such a thing) or the device number of the device holding the
340 filesystem (if the filesystem is stored on the device).
342 As not all filesystems are stored on devices, and not all filesystems
343 have UUIDs, it is sometimes necessary to explicitly tell NFS how to
344 identify a filesystem. This is done with the
348 For NFSv4, there is a distinguished filesystem which is the root of
349 all exported filesystem. This is specified with
353 both of which mean exactly the same thing.
355 Other filesystems can be identified with a small integer, or a UUID
356 which should contain 32 hex digits and arbitrary punctuation.
358 Linux kernels version 2.6.20 and earlier do not understand the UUID
359 setting so a small integer must be used if an fsid option needs to be
360 set for such kernels. Setting both a small number and a UUID is
361 supported so the same configuration can be made to work on old and new
365 .IR refer= path@host[+host][:path@host[+host]]
366 A client referencing the export point will be directed to choose from
367 the given list an alternative location for the filesystem.
368 (Note that the server must have a mountpoint here, though a different
369 filesystem is not required; so, for example,
370 .IR "mount --bind" " /path /path"
373 .IR replicas= path@host[+host][:path@host[+host]]
374 If the client asks for alternative locations for the export point, it
375 will be given this list of alternatives. (Note that actual replication
376 of the filesystem must be handled elsewhere.)
379 .IR refer= path@host[+host][:path@host[+host]]
380 A client referencing the export point will be directed to choose from
381 the given list an alternative location for the filesystem.
382 (Note that the server must have a mountpoint here, though a different
383 filesystem is not required; so, for example,
384 .IR "mount --bind" " /path /path"
387 .IR replicas= path@host[+host][:path@host[+host]]
388 If the client asks for alternative locations for the export point, it
389 will be given this list of alternatives. (Note that actual replication
390 of the filesystem must be handled elsewhere.)
395 bases its access control to files on the server machine on the uid and
396 gid provided in each NFS RPC request. The normal behavior a user would
397 expect is that she can access her files on the server just as she would
398 on a normal file system. This requires that the same uids and gids are
399 used on the client and the server machine. This is not always true, nor
400 is it always desirable.
402 Very often, it is not desirable that the root user on a client machine
403 is also treated as root when accessing files on the NFS server. To this
404 end, uid 0 is normally mapped to a different id: the so-called
407 uid. This mode of operation (called `root squashing') is the default,
408 and can be turned off with
413 '''tries to obtain the anonymous uid and gid by looking up user
415 '''in the password file at startup time. If it isn't found, a uid and gid
417 chooses a uid and gid
418 of 65534 for squashed access. These values can also be overridden by
420 .IR anonuid " and " anongid
423 '''In addition to this,
425 '''lets you specify arbitrary uids and gids that should be mapped to user
427 Finally, you can map all user requests to the
428 anonymous uid by specifying the
429 .IR all_squash " option.
431 Here's the complete list of mapping options:
434 Map requests from uid/gid 0 to the anonymous uid/gid. Note that this does
435 not apply to any other uids or gids that might be equally sensitive, such as
442 Turn off root squashing. This option is mainly useful for diskless clients.
445 Map all uids and gids to the anonymous user. Useful for NFS-exported
446 public FTP directories, news spool directories, etc. The opposite option
449 which is the default setting.
451 .IR anonuid " and " anongid
452 These options explicitly set the uid and gid of the anonymous account.
453 This option is primarily useful for PC/NFS clients, where you might want
454 all requests appear to be from one user. As an example, consider the
457 in the example section below, which maps all requests to uid 150 (which
458 is supposedly that of user joe).
464 # sample /etc/exports file
465 / master(rw) trusty(rw,no_root_squash)
466 /projects proj*.local.domain(rw)
467 /usr *.local.domain(ro) @trusted(rw)
468 /home/joe pc001(rw,all_squash,anonuid=150,anongid=100)
469 /pub *(ro,insecure,all_squash)
470 /srv/www \-sync,rw server @trusted @external(ro)
471 '''/pub/private (noaccess)
474 The first line exports the entire filesystem to machines master and trusty.
475 In addition to write access, all uid squashing is turned off for host
476 trusty. The second and third entry show examples for wildcard hostnames
477 and netgroups (this is the entry `@trusted'). The fourth line shows the
478 entry for the PC/NFS client discussed above. Line 5 exports the
479 public FTP directory to every host in the world, executing all requests
480 under the nobody account. The
482 option in this entry also allows clients with NFS implementations that
483 don't use a reserved port for NFS.
484 The sixth line exports a directory read-write to the machine 'server'
485 as well as the `@trusted' netgroup, and read-only to netgroup `@external',
486 all three mounts with the `sync' option enabled.
487 ''' The last line denies all NFS clients
488 '''access to the private directory.
490 '''Unlike other NFS server implementations, this
492 '''allows you to export both a directory and a subdirectory thereof to
493 '''the same host, for instance
494 '''.IR /usr " and " /usr/X11R6 .
495 '''In this case, the mount options of the most specific entry apply. For
496 '''instance, when a user on the client host accesses a file in
498 '''the mount options given in the
500 '''entry apply. This is also true when the latter is a wildcard or netgroup
511 '''An error parsing the file is reported using syslogd(8) as level NOTICE from
516 '''is started up. Any unknown
517 '''host is reported at that time, but often not all hosts are not yet known
520 '''at boot time, thus as hosts are found they are reported
projects / nfs-utils.git / blob