1
.TH EXPORTS 5 "4 March 2005" "Linux" "Linux File Formats Manual"
3
exports \- NFS file systems being exported (for Kernel based NFS)
9
serves as the access control list for file systems which may be
10
exported to NFS clients. It is used by
12
to give information to
14
and to the kernel based NFS file server daemon
17
The file format is similar to the SunOS
19
file. Each line contains an export point and a whitespace-separated list
20
of clients allowed to mount the file system at that point. Each listed
21
client may be immediately followed by a parenthesized, comma-separated
22
list of export options for that client. No whitespace is permitted
23
between a client and its option list.
25
Also, each line may have one or more specifications for default options
26
after the path name, in the form of a dash ("\-") followed by an option
27
list. The option list is used for all subsequent exports on that line
30
Blank lines are ignored. A pound sign ("#") introduces a comment to the
31
end of the line. Entries may be continued across newlines using a
32
backslash. If an export name contains spaces it should be quoted using
33
double quotes. You can also specify spaces or other unusual character in
34
the export name using a backslash followed by the character code as three
37
To apply changes to this file, run exportfs \-ra or restart the NFS server.
39
.SS Machine Name Formats
40
NFS clients may be specified in a number of ways:
42
This is the most common format. You may specify a host either by an
43
abbreviated name recognized be the resolver, the fully qualified domain
44
name, or an IP address.
46
NIS netgroups may be given as
48
Only the host part of each
49
netgroup members is consider in checking for membership. Empty host
50
parts or those containing a single dash (\-) are ignored.
52
Machine names may contain the wildcard characters \fI*\fR and \fI?\fR.
53
This can be used to make the \fIexports\fR file more compact; for instance,
54
\fI*.cs.foo.edu\fR matches all hosts in the domain
55
\fIcs.foo.edu\fR. As these characters also match the dots in a domain
56
name, the given pattern will also match all hosts within any subdomain
59
You can also export directories to all hosts on an IP (sub-) network
60
simultaneously. This is done by specifying an IP address and netmask pair
63
where the netmask can be specified in dotted-decimal format, or as a
64
contiguous mask length (for example, either `/255.255.252.0' or `/22' appended
65
to the network base address result in identical subnetworks with 10 bits of
66
host). Wildcard characters generally do not work on IP addresses, though they
67
may work by accident when reverse DNS lookups fail.
70
'''This is a special ``hostname'' that identifies the given directory name
71
'''as the public root directory (see the section on WebNFS in
73
'''for a discussion of WebNFS and the public root handle). When using this
76
'''must be the only entry on this line, and must have no export options
77
'''associated with it. Note that this does
79
'''actually export the named directory; you still have to set the exports
80
'''options in a separate entry.
82
'''The public root path can also be specified by invoking
85
'''.B \-\-public\-root
86
'''option. Multiple specifications of a public root will be ignored.
88
.SS RPCSEC_GSS security
89
You may use the special strings "gss/krb5", "gss/krb5i", or "gss/krb5p"
90
to restrict access to clients using rpcsec_gss security. However, this
91
syntax is deprecated; on linux kernels since 2.6.23, you should instead
92
use the "sec=" export option:
95
The sec= option, followed by a colon-delimited list of security flavors,
96
restricts the export to clients using those flavors. Available security
97
flavors include sys (the default--no cryptographic security), krb5
98
(authentication only), krb5i (integrity protection), and krb5p (privacy
99
protection). For the purposes of security flavor negotiation, order
100
counts: preferred flavors should be listed first. The order of the sec=
101
option with respect to the other options does not matter, unless you
102
want some options to be enforced differently depending on flavor.
103
In that case you may include multiple sec= options, and following options
104
will be enforced only for access using flavors listed in the immediately
105
preceding sec= option. The only options that are permitted to vary in
106
this way are ro, rw, no_root_squash, root_squash, and all_squash.
110
understands the following export options:
113
This option requires that requests originate on an Internet port less
114
than IPPORT_RESERVED (1024). This option is on by default. To turn it
119
Allow both read and write requests on this NFS volume. The
120
default is to disallow any request which changes the filesystem.
121
This can also be made explicit by using
126
This option allows the NFS server to violate the NFS protocol and
127
reply to requests before any changes made by that request have been
128
committed to stable storage (e.g. disc drive).
130
Using this option usually improves performance, but at the cost that
131
an unclean server restart (i.e. a crash) can cause data to be lost or
136
Reply to requests only after the changes have been committed to stable
141
In releases of nfs-utils up to and including 1.0.0, this option was the
142
default. In all releases after 1.0.0,
146
must be explicitly requested if needed.
147
To help make system administrators aware of this change, 'exportfs'
148
will issue a warning if neither
155
This option has no effect if
157
is also set. The NFS server will normally delay committing a write request
158
to disc slightly if it suspects that another related write request may be in
159
progress or may arrive soon. This allows multiple write requests to
160
be committed to disc with the one operation which can improve
161
performance. If an NFS server received mainly small unrelated
162
requests, this behaviour could actually reduce performance, so
164
is available to turn it off.
165
The default can be explicitly requested with the
169
This option is based on the option of the same name provided in IRIX
170
NFS. Normally, if a server exports two filesystems one of which is
171
mounted on the other, then the client will have to mount both
172
filesystems explicitly to get access to them. If it just mounts the
173
parent, it will see an empty directory at the place where the other
174
filesystem is mounted. That filesystem is "hidden".
178
option on a filesystem causes it not to be hidden, and an
179
appropriately authorised client will be able to move from the parent to
180
that filesystem without noticing the change.
182
However, some NFS clients do not cope well with this situation as, for
183
instance, it is then possible for two files in the one apparent
184
filesystem to have the same inode number.
188
option is currently only effective on
190
exports. It does not work reliably with netgroup, subnet, or wildcard
193
This option can be very useful in some situations, but it should be
194
used with due care, and only after confirming that the client system
195
copes with the situation effectively.
197
The option can be explicitly disabled with
201
This option is similar to
203
but it makes it possible for clients to move from the filesystem marked
204
with crossmnt to exported filesystems mounted on it. Thus when a child
205
filesystem "B" is mounted on a parent "A", setting crossmnt on "A" has
206
the same effect as setting "nohide" on B.
209
This option disables subtree checking, which has mild security
210
implications, but can improve reliability in some circumstances.
212
If a subdirectory of a filesystem is exported, but the whole
213
filesystem isn't then whenever a NFS request arrives, the server must
214
check not only that the accessed file is in the appropriate filesystem
215
(which is easy) but also that it is in the exported tree (which is
216
harder). This check is called the
219
In order to perform this check, the server must include some
220
information about the location of the file in the "filehandle" that is
221
given to the client. This can cause problems with accessing files that
222
are renamed while a client has them open (though in many simple cases
225
subtree checking is also used to make sure that files inside
226
directories to which only root has access can only be accessed if the
227
filesystem is exported with
229
(see below), even if the file itself allows more general access.
231
As a general guide, a home directory filesystem, which is normally
232
exported at the root and may see lots of file renames, should be
233
exported with subtree checking disabled. A filesystem which is mostly
234
readonly, and at least doesn't see many file renames (e.g. /usr or
235
/var) and for which subdirectories may be exported, should probably be
236
exported with subtree checks enabled.
238
The default of having subtree checks enabled, can be explicitly
242
From release 1.1.0 of nfs-utils onwards, the default will be
244
as subtree_checking tends to cause more problems than it is worth.
245
If you genuinely require subtree checking, you should explicitly put
248
file. If you put neither option,
250
will warn you that the change is pending.
256
This option (the two names are synonymous) tells the NFS server not to require authentication of
257
locking requests (i.e. requests which use the NLM protocol). Normally
258
the NFS server will require a lock request to hold a credential for a
259
user who has read access to the file. With this flag no access checks
262
Early NFS client implementations did not send credentials with lock
263
requests, and many current NFS clients still exist which are based on
264
the old implementations. Use this flag if you find that you can only
265
lock files which are world readable.
267
The default behaviour of requiring authentication for NLM requests can
268
be explicitly requested with either of the synonymous
274
On some specially patched kernels, and when exporting filesystems that
275
support ACLs, this option tells nfsd not to reveal ACLs to clients, so
276
they will see only a subset of actual permissions on the given file
277
system. This option is safe for filesystems used by NFSv2 clients and
278
old NFSv3 clients that perform access decisions locally. Current
279
NFSv3 clients use the ACCESS RPC to perform all access decisions on
280
the server. Note that the
282
option only has effect on kernels specially patched to support it, and
283
when exporting filesystems with ACL support. The default is to export
284
with ACL support (i.e. by default,
290
'''This makes everything below the directory inaccessible for the named
291
'''client. This is useful when you want to export a directory hierarchy to
292
'''a client, but exclude certain subdirectories. The client's view of a
293
'''directory flagged with noaccess is very limited; it is allowed to read
294
'''its attributes, and lookup `.' and `..'. These are also the only entries
295
'''returned by a readdir.
298
'''Convert absolute symbolic links (where the link contents start with a
299
'''slash) into relative links by prepending the necessary number of ../'s
300
'''to get from the directory containing the link to the root on the
301
'''server. This has subtle, perhaps questionable, semantics when the file
302
'''hierarchy is not mounted at its root.
305
'''Leave all symbolic link as they are. This is the default operation.
311
This option makes it possible to only export a directory if it has
312
successfully been mounted.
313
If no path is given (e.g.
314
.IR mountpoint " or " mp )
315
then the export point must also be a mount point. If it isn't then
316
the export point is not exported. This allows you to be sure that the
317
directory underneath a mountpoint will never be exported by accident
318
if, for example, the filesystem failed to mount due to a disc error.
320
If a path is given (e.g.
321
.IR mountpoint= "/path or " mp= /path)
322
then the nominated path must be a mountpoint for the exportpoint to be
326
.IR fsid= num|root|uuid
327
NFS needs to be able to identify each filesystem that it exports.
328
Normally it will use a UUID for the filesystem (if the filesystem has
329
such a thing) or the device number of the device holding the
330
filesystem (if the filesystem is stored on the device).
332
As not all filesystems are stored on devices, and not all filesystems
333
have UUIDs, it is sometimes necessary to explicitly tell NFS how to
334
identify a filesystem. This is done with the
338
For NFSv4, there is a distinguished filesystem which is the root of
339
all exported filesystem. This is specified with
343
both of which mean exactly the same thing.
345
Other filesystems can be identified with a small integer, or a UUID
346
which should contain 32 hex digits and arbitrary punctuation.
348
Linux kernels version 2.6.20 and earlier do not understand the UUID
349
setting so a small integer must be used if an fsid option needs to be
350
set for such kernels. Setting both a small number and a UUID is
351
supported so the same configuration can be made to work on old and new
355
.IR refer= path@host[+host][:path@host[+host]]
356
A client referencing the export point will be directed to choose from
357
the given list an alternative location for the filesystem.
358
(Note that the server must have a mountpoint here, though a different
359
filesystem is not required; so, for example,
360
.IR "mount --bind" " /path /path"
363
.IR replicas= path@host[+host][:path@host[+host]]
364
If the client asks for alternative locations for the export point, it
365
will be given this list of alternatives. (Note that actual replication
366
of the filesystem must be handled elsewhere.)
369
.IR refer= path@host[+host][:path@host[+host]]
370
A client referencing the export point will be directed to choose from
371
the given list an alternative location for the filesystem.
372
(Note that the server must have a mountpoint here, though a different
373
filesystem is not required; so, for example,
374
.IR "mount --bind" " /path /path"
377
.IR replicas= path@host[+host][:path@host[+host]]
378
If the client asks for alternative locations for the export point, it
379
will be given this list of alternatives. (Note that actual replication
380
of the filesystem must be handled elsewhere.)
385
bases its access control to files on the server machine on the uid and
386
gid provided in each NFS RPC request. The normal behavior a user would
387
expect is that she can access her files on the server just as she would
388
on a normal file system. This requires that the same uids and gids are
389
used on the client and the server machine. This is not always true, nor
390
is it always desirable.
392
Very often, it is not desirable that the root user on a client machine
393
is also treated as root when accessing files on the NFS server. To this
394
end, uid 0 is normally mapped to a different id: the so-called
397
uid. This mode of operation (called `root squashing') is the default,
398
and can be turned off with
403
'''tries to obtain the anonymous uid and gid by looking up user
405
'''in the password file at startup time. If it isn't found, a uid and gid
407
chooses a uid and gid
408
of 65534 for squashed access. These values can also be overridden by
410
.IR anonuid " and " anongid
413
'''In addition to this,
415
'''lets you specify arbitrary uids and gids that should be mapped to user
417
Finally, you can map all user requests to the
418
anonymous uid by specifying the
419
.IR all_squash " option.
421
Here's the complete list of mapping options:
424
Map requests from uid/gid 0 to the anonymous uid/gid. Note that this does
425
not apply to any other uids or gids that might be equally sensitive, such as
432
Turn off root squashing. This option is mainly useful for diskless clients.
435
Map all uids and gids to the anonymous user. Useful for NFS-exported
436
public FTP directories, news spool directories, etc. The opposite option
439
which is the default setting.
441
.IR anonuid " and " anongid
442
These options explicitly set the uid and gid of the anonymous account.
443
This option is primarily useful for PC/NFS clients, where you might want
444
all requests appear to be from one user. As an example, consider the
447
in the example section below, which maps all requests to uid 150 (which
448
is supposedly that of user joe).
454
# sample /etc/exports file
455
/ master(rw) trusty(rw,no_root_squash)
456
/projects proj*.local.domain(rw)
457
/usr *.local.domain(ro) @trusted(rw)
458
/home/joe pc001(rw,all_squash,anonuid=150,anongid=100)
459
/pub *(ro,insecure,all_squash)
460
/srv/www \-sync,rw server @trusted @external(ro)
461
'''/pub/private (noaccess)
464
The first line exports the entire filesystem to machines master and trusty.
465
In addition to write access, all uid squashing is turned off for host
466
trusty. The second and third entry show examples for wildcard hostnames
467
and netgroups (this is the entry `@trusted'). The fourth line shows the
468
entry for the PC/NFS client discussed above. Line 5 exports the
469
public FTP directory to every host in the world, executing all requests
470
under the nobody account. The
472
option in this entry also allows clients with NFS implementations that
473
don't use a reserved port for NFS.
474
The sixth line exports a directory read-write to the machine 'server'
475
as well as the `@trusted' netgroup, and read-only to netgroup `@external',
476
all three mounts with the `sync' option enabled.
477
''' The last line denies all NFS clients
478
'''access to the private directory.
480
'''Unlike other NFS server implementations, this
482
'''allows you to export both a directory and a subdirectory thereof to
483
'''the same host, for instance
484
'''.IR /usr " and " /usr/X11R6 .
485
'''In this case, the mount options of the most specific entry apply. For
486
'''instance, when a user on the client host accesses a file in
488
'''the mount options given in the
490
'''entry apply. This is also true when the latter is a wildcard or netgroup
501
'''An error parsing the file is reported using syslogd(8) as level NOTICE from
502
'''a DAEMON whenever nfsd(8) or mountd(8) is started up. Any unknown
503
'''host is reported at that time, but often not all hosts are not yet known
504
'''to named(8) at boot time, thus as hosts are found they are reported
505
'''with the same syslogd(8) parameters.