1EXPORTS(5) Linux File Formats Manual EXPORTS(5)
2
6 exports - NFS file systems being exported (for Kernel based NFS)
7
9 /etc/exports
10
12 The file /etc/exports serves as the access control list for file sys‐
13 tems which may be exported to NFS clients. It is used by exportfs(8)
14 to give information to mountd(8) and to the kernel based NFS file
15 server daemon nfsd(8).
16 The file format is similar to the SunOS exports file. Each line con‐
18 tains an export point and a whitespace-separated list of clients
19 allowed to mount the file system at that point. Each listed client may
20 be immediately followed by a parenthesized, comma-separated list of
21 export options for that client. No whitespace is permitted between a
22 client and its option list.
23 Also, each line may have one or more specifications for default options
25 after the path name, in the form of a dash ("-") followed by an option
26 list. The option list is used for all subsequent exports on that line
27 only.
28 Blank lines are ignored. A pound sign ("#") introduces a comment to
30 the end of the line. Entries may be continued across newlines using a
31 backslash. If an export name contains spaces it should be quoted using
32 double quotes. You can also specify spaces or other unusual character
33 in the export name using a backslash followed by the character code as
34 three octal digits.
35 Machine Name Formats
37 NFS clients may be specified in a number of ways:
38 single host
40 This is the most common format. You may specify a host either by
41 an abbreviated name recognized be the resolver, the fully quali‐
42 fied domain name, or an IP address.
43 netgroups
45 NIS netgroups may be given as @group. Only the host part of
46 each netgroup members is consider in checking for membership.
47 Empty host parts or those containing a single dash (-) are
48 ignored.
49 wildcards
51 Machine names may contain the wildcard characters * and ?. This
52 can be used to make the exports file more compact; for instance,
53 *.cs.foo.edu matches all hosts in the domain cs.foo.edu. As
54 these characters also match the dots in a domain name, the given
55 pattern will also match all hosts within any subdomain of
56 cs.foo.edu.
57 IP networks
59 You can also export directories to all hosts on an IP (sub-)
60 network simultaneously. This is done by specifying an IP address
61 and netmask pair as address/netmask where the netmask can be
62 specified in dotted-decimal format, or as a contiguous mask
63 length (for example, either `/255.255.252.0' or `/22' appended
64 to the network base address result in identical subnetworks with
65 10 bits of host). Wildcard characters generally do not work on
66 IP addresses, though they may work by accident when reverse DNS
67 lookups fail.
68 RPCSEC_GSS security
70 To restrict access to an export using rpcsec_gss security, use the spe‐
71 cial string "gss/krb5" as the client. It is not possible to simultane‐
72 ously require rpcsec_gss and to make requirements on the IP address of
73 the client.
74 General Options
76 exportfs understands the following export options:
77 secure This option requires that requests originate on an Internet port
79 less than IPPORT_RESERVED (1024). This option is on by default.
80 To turn it off, specify insecure.
81 rw Allow both read and write requests on this NFS volume. The
83 default is to disallow any request which changes the filesystem.
84 This can also be made explicit by using the ro option.
85 async This option allows the NFS server to violate the NFS protocol
87 and reply to requests before any changes made by that request
88 have been committed to stable storage (e.g. disc drive).
89 Using this option usually improves performance, but at the cost
91 that an unclean server restart (i.e. a crash) can cause data to
92 be lost or corrupted.
93 sync Reply to requests only after the changes have been committed to
96 stable storage (see async above).
97 In releases of nfs-utils up to and including 1.0.0, this option
99 was the default. In all subsequence releases, sync is the
100 default, and async must be explicitly requested if needed. To
101 help make system administrators aware of this change, 'exportfs'
102 will issue a warning if neither sync nor async is specified.
103 no_wdelay
105 This option has no effect if async is also set. The NFS server
106 will normally delay committing a write request to disc slightly
107 if it suspects that another related write request may be in
108 progress or may arrive soon. This allows multiple write
109 requests to be committed to disc with the one operation which
110 can improve performance. If an NFS server received mainly small
111 unrelated requests, this behaviour could actually reduce perfor‐
112 mance, so no_wdelay is available to turn it off. The default
113 can be explicitly requested with the wdelay option.
114 nohide This option is based on the option of the same name provided in
116 IRIX NFS. Normally, if a server exports two filesystems one of
117 which is mounted on the other, then the client will have to
118 mount both filesystems explicitly to get access to them. If it
119 just mounts the parent, it will see an empty directory at the
120 place where the other filesystem is mounted. That filesystem is
121 "hidden".
122 Setting the nohide option on a filesystem causes it not to be
124 hidden, and an appropriately authorised client will be able to
125 move from the parent to that filesystem without noticing the
126 change.
127 However, some NFS clients do not cope well with this situation
129 as, for instance, it is then possible for two files in the one
130 apparent filesystem to have the same inode number.
131 The nohide option is currently only effective on single host
133 exports. It does not work reliably with netgroup, subnet, or
134 wildcard exports.
135 This option can be very useful in some situations, but it should
137 be used with due care, and only after confirming that the client
138 system copes with the situation effectively.
139 The option can be explicitly disabled with hide.
141 crossmnt
143 This option is similar to nohide but it makes it possible for
144 clients to move from the filesystem marked with crossmnt to
145 exported filesystems mounted on it. Thus when a child filesys‐
146 tem "B" is mounted on a parent "A", setting crossmnt on "A" has
147 the same effect as setting "nohide" on B.
148 no_subtree_check
150 This option disables subtree checking, which has mild security
151 implications, but can improve reliability in some circumstances.
152 If a subdirectory of a filesystem is exported, but the whole
154 filesystem isn't then whenever a NFS request arrives, the server
155 must check not only that the accessed file is in the appropriate
156 filesystem (which is easy) but also that it is in the exported
157 tree (which is harder). This check is called the subtree_check.
158 In order to perform this check, the server must include some
160 information about the location of the file in the "filehandle"
161 that is given to the client. This can cause problems with
162 accessing files that are renamed while a client has them open
163 (though in many simple cases it will still work).
164 subtree checking is also used to make sure that files inside
166 directories to which only root has access can only be accessed
167 if the filesystem is exported with no_root_squash (see below),
168 even if the file itself allows more general access.
169 As a general guide, a home directory filesystem, which is nor‐
171 mally exported at the root and may see lots of file renames,
172 should be exported with subtree checking disabled. A filesystem
173 which is mostly readonly, and at least doesn't see many file
174 renames (e.g. /usr or /var) and for which subdirectories may be
175 exported, should probably be exported with subtree checks
176 enabled.
177 The default of having subtree checks enabled, can be explicitly
179 requested with subtree_check.
180 From release 1.1.0 of nfs-utils onwards, the default will be
182 no_subtree_check as subtree_checking tends to cause more prob‐
183 lems than it is worth. If you genuinely require subtree check‐
184 ing, you should explicitly put that option in the exports file.
185 If you put neither option, exportfs will warn you that the
186 change is pending.
187 insecure_locks
190 no_auth_nlm
192 This option (the two names are synonymous) tells the NFS server
193 not to require authentication of locking requests (i.e. requests
194 which use the NLM protocol). Normally the NFS server will
195 require a lock request to hold a credential for a user who has
196 read access to the file. With this flag no access checks will
197 be performed.
198 Early NFS client implementations did not send credentials with
200 lock requests, and many current NFS clients still exist which
201 are based on the old implementations. Use this flag if you find
202 that you can only lock files which are world readable.
203 The default behaviour of requiring authentication for NLM
205 requests can be explicitly requested with either of the synony‐
206 mous auth_nlm, or secure_locks.
207 no_acl On some specially patched kernels, and when exporting filesys‐
209 tems that support ACLs, this option tells nfsd not to reveal
210 ACLs to clients, so they will see only a subset of actual per‐
211 missions on the given file system. This option is safe for
212 filesystems used by NFSv2 clients and old NFSv3 clients that
213 perform access decisions locally. Current NFSv3 clients use the
214 ACCESS RPC to perform all access decisions on the server. Note
215 that the no_acl option only has effect on kernels specially
216 patched to support it, and when exporting filesystems with ACL
217 support. The default is to export with ACL support (i.e. by
218 default, no_acl is off).
219 mountpoint=path
223 mp This option makes it possible to only export a directory if it
225 has successfully been mounted. If no path is given (e.g.
226 mountpoint or mp) then the export point must also be a mount
227 point. If it isn't then the export point is not exported. This
228 allows you to be sure that the directory underneath a mountpoint
229 will never be exported by accident if, for example, the filesys‐
230 tem failed to mount due to a disc error.
231 If a path is given (e.g. mountpoint=/path or mp=/path) then the
233 nominated path must be a mountpoint for the exportpoint to be
234 exported.
235 fsid=num|root|uuid
238 NFS needs to be able to identify each filesystem that it
239 exports. Normally it will use a UUID for the filesystem (if the
240 filesystem has such a thing) or the device number of the device
241 holding the filesystem (if the filesystem is stored on the
242 device).
243 As not all filesystems are stored on devices, and not all
245 filesystems have UUIDs, it is sometimes necessary to explicitly
246 tell NFS how to identify a filesystem. This is done with the
247 fsid= option.
248 For NFSv4, there is a distinguished filesystem which is the root
250 of all exported filesystem. This is specified with fsid=root or
251 fsid=0 both of which mean exactly the same thing.
252 Other filesystems can be identified with a small integer, or a
254 UUID which should contain 32 hex digits and arbitrary punctua‐
255 tion.
256 Linux kernels version 2.6.20 and earlier do not understand the
258 UUID setting so a small integer must be used if an fsid option
259 needs to be set for such kernels. Setting both a small number
260 and a UUID is supported so the same configuration can be made to
261 work on old and new kernels alike.
262 refer=path@host[+host][:path@host[+host]]
265 A client referencing the export point will be directed to choose
266 from the given list an alternative location for the filesystem.
267 (Note that the server must have a mountpoint here, though a dif‐
268 ferent filesystem is not required; so, for example, mount --bind
269 /path /path is sufficient.)
270 replicas=path@host[+host][:path@host[+host]]
272 If the client asks for alternative locations for the export
273 point, it will be given this list of alternatives. (Note that
274 actual replication of the filesystem must be handled elsewhere.)
275 refer=path@host[+host][:path@host[+host]]
278 A client referencing the export point will be directed to choose
279 from the given list an alternative location for the filesystem.
280 (Note that the server must have a mountpoint here, though a dif‐
281 ferent filesystem is not required; so, for example, mount --bind
282 /path /path is sufficient.)
283 replicas=path@host[+host][:path@host[+host]]
285 If the client asks for alternative locations for the export
286 point, it will be given this list of alternatives. (Note that
287 actual replication of the filesystem must be handled elsewhere.)
288 User ID Mapping
291 nfsd bases its access control to files on the server machine on the uid
292 and gid provided in each NFS RPC request. The normal behavior a user
293 would expect is that she can access her files on the server just as she
294 would on a normal file system. This requires that the same uids and
295 gids are used on the client and the server machine. This is not always
296 true, nor is it always desirable.
297 Very often, it is not desirable that the root user on a client machine
299 is also treated as root when accessing files on the NFS server. To this
300 end, uid 0 is normally mapped to a different id: the so-called anony‐
301 mous or nobody uid. This mode of operation (called `root squashing') is
302 the default, and can be turned off with no_root_squash.
303 By default, exportfs chooses a uid and gid of 65534 for squashed
305 access. These values can also be overridden by the anonuid and anongid
306 options. Finally, you can map all user requests to the anonymous uid
307 by specifying the all_squash option.
308 Here's the complete list of mapping options:
310 root_squash
312 Map requests from uid/gid 0 to the anonymous uid/gid. Note that
313 this does not apply to any other uids or gids that might be
314 equally sensitive, such as user bin or group staff.
315 no_root_squash
317 Turn off root squashing. This option is mainly useful for disk‐
318 less clients.
319 all_squash
321 Map all uids and gids to the anonymous user. Useful for NFS-
322 exported public FTP directories, news spool directories, etc.
323 The opposite option is no_all_squash, which is the default set‐
324 ting.
325 anonuid and anongid
327 These options explicitly set the uid and gid of the anonymous
328 account. This option is primarily useful for PC/NFS clients,
329 where you might want all requests appear to be from one user. As
330 an example, consider the export entry for /home/joe in the exam‐
331 ple section below, which maps all requests to uid 150 (which is
332 supposedly that of user joe).
333
335 # sample /etc/exports file
336 / master(rw) trusty(rw,no_root_squash)
337 /projects proj*.local.domain(rw)
338 /usr *.local.domain(ro) @trusted(rw)
339 /home/joe pc001(rw,all_squash,anonuid=150,anongid=100)
340 /pub (ro,insecure,all_squash)
341 /srv/www -sync,rw server @trusted @external(ro)
342 The first line exports the entire filesystem to machines master and
344 trusty. In addition to write access, all uid squashing is turned off
345 for host trusty. The second and third entry show examples for wildcard
346 hostnames and netgroups (this is the entry `@trusted'). The fourth line
347 shows the entry for the PC/NFS client discussed above. Line 5 exports
348 the public FTP directory to every host in the world, executing all
349 requests under the nobody account. The insecure option in this entry
350 also allows clients with NFS implementations that don't use a reserved
351 port for NFS. The sixth line exports a directory read-write to the
352 machine 'server' as well as the `@trusted' netgroup, and read-only to
353 netgroup `@external', all three mounts with the `sync' option enabled.
354
356 /etc/exports
357
359 exportfs(8), netgroup(5), mountd(8), nfsd(8), showmount(8).
360Linux 4 March 2005 EXPORTS(5)