.\"@(#)exports.5"
.\"
.TH exports 5 "31 December 2009"
.SH NAME
exports \- NFS server export table
.SH DESCRIPTION
The file
.I /etc/exports
contains a table of local physical file systems on an NFS server
that are accessible to NFS clients.
The contents of the file are maintained by the server's system
administrator.
.PP
Each file system in this table has a list of options and an
access control list.
The table is used by
.BR exportfs (8)
to give information to
.BR mountd (8).
.PP
The file format is similar to the SunOS
.I exports
file. Each line contains an export point and a whitespace-separated list
of clients allowed to mount the file system at that point. Each listed
client may be immediately followed by a parenthesized, comma-separated
list of export options for that client. No whitespace is permitted
between a client and its option list.
.PP
Also, each line may have one or more specifications for default options
after the path name, in the form of a dash ("\-") followed by an option
list. The option list is used for all subsequent exports on that line
only.
.PP
Blank lines are ignored.  A pound sign ("#") introduces a comment to the
end of the line. Entries may be continued across newlines using a
backslash. If an export name contains spaces it should be quoted using
double quotes. You can also specify spaces or other unusual character in
the export name using a backslash followed by the character code as three
octal digits.
.PP
To apply changes to this file, run
.BR "exportfs \-ra"
or restart the NFS server.
.PP
.SS Machine Name Formats
NFS clients may be specified in a number of ways:
.IP "single host
You may specify a host either by an
abbreviated name recognized be the resolver, the fully qualified domain
name, an IPv4 address, or an IPv6 address. IPv6 addresses must not be
inside square brackets in /etc/exports lest they be confused with
character-class wildcard matches.
.IP "IP networks
You can also export directories to all hosts on an IP (sub-) network
simultaneously. This is done by specifying an IP address and netmask pair
as
.IR address/netmask
where the netmask can be specified in dotted-decimal format, or as a
contiguous mask length.
For example, either `/255.255.252.0' or `/22' appended
to the network base IPv4 address results in identical subnetworks with 10 bits
of host. IPv6 addresses must use a contiguous mask length and must not be inside square brackets to avoid confusion with character-class wildcards. Wildcard characters generally do not work on IP addresses, though they
may work by accident when reverse DNS lookups fail.
.IP "wildcards
Machine names may contain the wildcard characters \fI*\fR and \fI?\fR, or may contain character class lists within [square brackets].
This can be used to make the \fIexports\fR file more compact; for instance,
\fI*.cs.foo.edu\fR matches all hosts in the domain
\fIcs.foo.edu\fR.  As these characters also match the dots in a domain
name, the given pattern will also match all hosts within any subdomain
of \fIcs.foo.edu\fR.
.IP "netgroups
NIS netgroups may be given as
.IR @group .
Only the host part of each
netgroup members is consider in checking for membership.  Empty host
parts or those containing a single dash (\-) are ignored.
.IP "anonymous
This is specified by a single
.I *
character (not to be confused with the
.I wildcard
entry above) and will match all clients.
.\".TP
.\".B =public
.\"This is a special ``hostname'' that identifies the given directory name
.\"as the public root directory (see the section on WebNFS in
.\".BR nfsd (8)
.\"for a discussion of WebNFS and the public root handle). When using this
.\"convention,
.\".B =public
.\"must be the only entry on this line, and must have no export options
.\"associated with it. Note that this does
.\".I not
.\"actually export the named directory; you still have to set the exports
.\"options in a separate entry.
.\".PP
.\"The public root path can also be specified by invoking
.\".I nfsd
.\"with the
.\".B \-\-public\-root
.\"option. Multiple specifications of a public root will be ignored.
.PP
If a client matches more than one of the specifications above, then
the first match from the above list order takes precedence - regardless of
the order they appear on the export line. However, if a client matches
more than one of the same type of specification (e.g. two netgroups),
then the first match from the order they appear on the export line takes
precedence.
.SS RPCSEC_GSS security
You may use the special strings "gss/krb5", "gss/krb5i", or "gss/krb5p"
to restrict access to clients using rpcsec_gss security.  However, this
syntax is deprecated; on linux kernels since 2.6.23, you should instead
use the "sec=" export option:
.TP
.IR sec=
The sec= option, followed by a colon-delimited list of security flavors,
restricts the export to clients using those flavors.  Available security
flavors include sys (the default--no cryptographic security), krb5
(authentication only), krb5i (integrity protection), and krb5p (privacy
protection).  For the purposes of security flavor negotiation, order
counts: preferred flavors should be listed first.  The order of the sec=
option with respect to the other options does not matter, unless you
want some options to be enforced differently depending on flavor.
In that case you may include multiple sec= options, and following options
will be enforced only for access using flavors listed in the immediately
preceding sec= option.  The only options that are permitted to vary in
this way are ro, rw, no_root_squash, root_squash, and all_squash.
.PP
.SS General Options
.BR exportfs
understands the following export options:
.TP
.IR secure
This option requires that requests not using gss originate on an
Internet port less than IPPORT_RESERVED (1024). This option is on by default.
To turn it off, specify
.IR insecure .
(NOTE: older kernels (before upstream kernel version 4.17) enforced this
requirement on gss requests as well.)
.TP
.IR rw
Allow both read and write requests on this NFS volume. The
default is to disallow any request which changes the filesystem.
This can also be made explicit by using
the
.IR ro " option.
.TP
.IR async
This option allows the NFS server to violate the NFS protocol and
reply to requests before any changes made by that request have been
committed to stable storage (e.g. disc drive).

Using this option usually improves performance, but at the cost that
an unclean server restart (i.e. a crash) can cause data to be lost or
corrupted.

.TP
.IR sync
Reply to requests only after the changes have been committed to stable
storage (see
.IR async
above).

In releases of nfs-utils up to and including 1.0.0, the
.I async 
option was the
default.  In all releases after 1.0.0,
.I sync
is the default, and
.I async
must be explicitly requested if needed.
.TP
.IR no_wdelay
This option has no effect if
.I async
is also set.  The NFS server will normally delay committing a write request
to disc slightly if it suspects that another related write request may be in
progress or may arrive soon.  This allows multiple write requests to
be committed to disc with the one operation which can improve
performance.  If an NFS server received mainly small unrelated
requests, this behaviour could actually reduce performance, so
.IR no_wdelay
is available to turn it off.
The default can be explicitly requested with the
.IR wdelay " option.
.TP
.IR nohide
This option is based on the option of the same name provided in IRIX
NFS.  Normally, if a server exports two filesystems one of which is
mounted on the other, then the client will have to mount both
filesystems explicitly to get access to them.  If it just mounts the
parent, it will see an empty directory at the place where the other
filesystem is mounted.  That filesystem is "hidden".

Setting the
.I nohide
option on a filesystem causes it not to be hidden, and an
appropriately authorised client will be able to move from the parent to
that filesystem without noticing the change.

However, some NFS clients do not cope well with this situation as, for
instance, it is then possible for two files in the one apparent
filesystem to have the same inode number.

The
.I nohide
option is currently only effective on
.I "single host
exports.  It does not work reliably with netgroup, subnet, or wildcard
exports.

This option can be very useful in some situations, but it should be
used with due care, and only after confirming that the client system
copes with the situation effectively.

The option can be explicitly disabled for NFSv2 and NFSv3 with
.IR hide .

This option is not relevant when NFSv4 is use.  NFSv4 never hides
subordinate filesystems.  Any filesystem that is exported will be
visible where expected when using NFSv4.
.TP
.I crossmnt
This option is similar to
.I nohide
but it makes it possible for clients to access all filesystems mounted
on a filesystem marked with
.IR crossmnt .
Thus when a child filesystem "B" is mounted on a parent "A", setting
crossmnt on "A" has a similar effect to setting "nohide" on B.

With
.I nohide
the child filesystem needs to be explicitly exported.  With
.I crossmnt
it need not.  If a child of a
.I crossmnt
file is not explicitly exported, then it will be implicitly exported
with the same export options as the parent, except for
.IR fsid= .
This makes it impossible to
.B not
export a child of a
.I crossmnt
filesystem.  If some but not all subordinate filesystems of a parent
are to be exported, then they must be explicitly exported and the
parent should not have
.I crossmnt
set.

The
.I nocrossmnt
option can explictly disable
.I crossmnt
if it was previously set.  This is rarely useful.
.TP
.IR no_subtree_check
This option disables subtree checking, which has mild security
implications, but can improve reliability in some circumstances.

If a subdirectory of a filesystem is exported, but the whole
filesystem isn't then whenever a NFS request arrives, the server must
check not only that the accessed file is in the appropriate filesystem
(which is easy) but also that it is in the exported tree (which is
harder). This check is called the
.IR subtree_check .

In order to perform this check, the server must include some
information about the location of the file in the "filehandle" that is
given to the client.  This can cause problems with accessing files that
are renamed while a client has them open (though in many simple cases
it will still work).

subtree checking is also used to make sure that files inside
directories to which only root has access can only be accessed if the
filesystem is exported with
.I no_root_squash
(see below), even if the file itself allows more general access.

As a general guide, a home directory filesystem, which is normally
exported at the root and may see lots of file renames, should be
exported with subtree checking disabled.  A filesystem which is mostly
readonly, and at least doesn't see many file renames (e.g. /usr or
/var) and for which subdirectories may be exported, should probably be
exported with subtree checks enabled.

The default of having subtree checks enabled, can be explicitly
requested with
.IR subtree_check .

From release 1.1.0 of nfs-utils onwards, the default will be
.I no_subtree_check
as subtree_checking tends to cause more problems than it is worth.
If you genuinely require subtree checking, you should explicitly put
that option in the
.B exports
file.  If you put neither option,
.B exportfs
will warn you that the change is pending.

.TP
.IR insecure_locks
.TP
.IR no_auth_nlm
This option (the two names are synonymous) tells the NFS server not to require authentication of
locking requests (i.e. requests which use the NLM protocol).  Normally
the NFS server will require a lock request to hold a credential for a
user who has read access to the file.  With this flag no access checks
will be performed.

Early NFS client implementations did not send credentials with lock
requests, and many current NFS clients still exist which are based on
the old implementations.  Use this flag if you find that you can only
lock files which are world readable.

The default behaviour of requiring authentication for NLM requests can
be explicitly requested with either of the synonymous
.IR auth_nlm ,
or
.IR secure_locks .
.\".TP
.\".I noaccess
.\"This makes everything below the directory inaccessible for the named
.\"client.  This is useful when you want to export a directory hierarchy to
.\"a client, but exclude certain subdirectories. The client's view of a
.\"directory flagged with noaccess is very limited; it is allowed to read
.\"its attributes, and lookup `.' and `..'. These are also the only entries
.\"returned by a readdir.
.\".TP
.\".IR link_relative
.\"Convert absolute symbolic links (where the link contents start with a
.\"slash) into relative links by prepending the necessary number of ../'s
.\"to get from the directory containing the link to the root on the
.\"server.  This has subtle, perhaps questionable, semantics when the file
.\"hierarchy is not mounted at its root.
.\".TP
.\".IR link_absolute
.\"Leave all symbolic link as they are. This is the default operation.

.TP
.IR mountpoint= path
.TP
.I mp
This option makes it possible to only export a directory if it has
successfully been mounted.
If no path is given (e.g.
.IR mountpoint " or " mp )
then the export point must also be a mount point.  If it isn't then
the export point is not exported.  This allows you to be sure that the
directory underneath a mountpoint will never be exported by accident
if, for example, the filesystem failed to mount due to a disc error.

If a path is given (e.g.
.IR mountpoint= "/path or " mp= /path)
then the nominated path must be a mountpoint for the exportpoint to be
exported.

.TP
.IR fsid= num|root|uuid
NFS needs to be able to identify each filesystem that it exports.
Normally it will use a UUID for the filesystem (if the filesystem has
such a thing) or the device number of the device holding the
filesystem (if the filesystem is stored on the device).

As not all filesystems are stored on devices, and not all filesystems
have UUIDs, it is sometimes necessary to explicitly tell NFS how to
identify a filesystem.  This is done with the
.I fsid=
option.

For NFSv4, there is a distinguished filesystem which is the root of
all exported filesystem.  This is specified with
.I fsid=root
or
.I fsid=0
both of which mean exactly the same thing.

Other filesystems can be identified with a small integer, or a UUID
which should contain 32 hex digits and arbitrary punctuation.

Linux kernels version 2.6.20 and earlier do not understand the UUID
setting so a small integer must be used if an fsid option needs to be
set for such kernels.  Setting both a small number and a UUID is
supported so the same configuration can be made to work on old and new
kernels alike.

.TP
.IR nordirplus
This option will disable READDIRPLUS request handling.  When set,
READDIRPLUS requests from NFS clients return NFS3ERR_NOTSUPP, and
clients fall back on READDIR.  This option affects only NFSv3 clients.
.TP
.IR refer= path@host[+host][:path@host[+host]]
A client referencing the export point will be directed to choose from
the given list an alternative location for the filesystem.
(Note that the server must have a mountpoint here, though a different
filesystem is not required; so, for example,
.IR "mount --bind" " /path /path"
is sufficient.)
.TP
.IR replicas= path@host[+host][:path@host[+host]]
If the client asks for alternative locations for the export point, it
will be given this list of alternatives. (Note that actual replication
of the filesystem must be handled elsewhere.)

.TP
.IR pnfs
This option enables the use of the pNFS extension if the protocol level
is NFSv4.1 or higher, and the filesystem supports pNFS exports.  With
pNFS clients can bypass the server and perform I/O directly to storage
devices. The default can be explicitly requested with the
.I no_pnfs
option.

.TP
.IR security_label
With this option set, clients using NFSv4.2 or higher will be able to
set and retrieve security labels (such as those used by SELinux).  This
will only work if all clients use a consistent security policy.  Note
that early kernels did not support this export option, and instead
enabled security labels by default.

.SS User ID Mapping
.PP
.B nfsd
bases its access control to files on the server machine on the uid and
gid provided in each NFS RPC request. The normal behavior a user would
expect is that she can access her files on the server just as she would
on a normal file system. This requires that the same uids and gids are
used on the client and the server machine. This is not always true, nor
is it always desirable.
.PP
Very often, it is not desirable that the root user on a client machine
is also treated as root when accessing files on the NFS server. To this
end, uid 0 is normally mapped to a different id: the so-called
anonymous or
.I nobody
uid. This mode of operation (called `root squashing') is the default,
and can be turned off with
.IR no_root_squash .
.PP
By default,
.\".B nfsd
.\"tries to obtain the anonymous uid and gid by looking up user
.\".I nobody
.\"in the password file at startup time. If it isn't found, a uid and gid
.B exportfs
chooses a uid and gid
of 65534 for squashed access. These values can also be overridden by
the
.IR anonuid " and " anongid
options.
.\".PP
.\"In addition to this,
.\".B nfsd
.\"lets you specify arbitrary uids and gids that should be mapped to user
.\"nobody as well.
Finally, you can map all user requests to the
anonymous uid by specifying the
.IR all_squash " option.
.PP
Here's the complete list of mapping options:
.TP
.IR root_squash
Map requests from uid/gid 0 to the anonymous uid/gid. Note that this does
not apply to any other uids or gids that might be equally sensitive, such as
user
.IR bin
or group
.IR staff .
.TP
.IR no_root_squash
Turn off root squashing. This option is mainly useful for diskless clients.
.TP
.IR all_squash
Map all uids and gids to the anonymous user. Useful for NFS-exported
public FTP directories, news spool directories, etc. The opposite option
is
.IR no_all_squash ,
which is the default setting.
.TP
.IR anonuid " and " anongid
These options explicitly set the uid and gid of the anonymous account.
This option is primarily useful for PC/NFS clients, where you might want
all requests appear to be from one user. As an example, consider the
export entry for
.B /home/joe
in the example section below, which maps all requests to uid 150 (which
is supposedly that of user joe).

.SS Subdirectory Exports

Normally you should only export only the root of a filesystem.  The NFS
server will also allow you to export a subdirectory of a filesystem,
however, this has drawbacks:

First, it may be possible for a malicious user to access files on the
filesystem outside of the exported subdirectory, by guessing filehandles
for those other files.  The only way to prevent this is by using the
.IR no_subtree_check
option, which can cause other problems.

Second, export options may not be enforced in the way that you would
expect.  For example, the
.IR security_label
option will not work on subdirectory exports, and if nested subdirectory
exports change the
.IR security_label
or
.IR sec=
options, NFSv4 clients will normally see only the options on the parent
export.  Also, where security options differ, a malicious client may use
filehandle-guessing attacks to access the files from one subdirectory
using the options from another.


.SS Extra Export Tables
After reading 
.I /etc/exports 
.B exportfs
reads files in the
.I /etc/exports.d
directory as extra export tables.  Only files ending in
.I .exports
are considered.  Files beginning with a dot are ignored.
The format for extra export tables is the same as 
.I /etc/exports
.
.IP
.SH EXAMPLE
.PP
.nf
.ta +3i
# sample /etc/exports file
/               master(rw) trusty(rw,no_root_squash)
/projects       proj*.local.domain(rw)
/usr            *.local.domain(ro) @trusted(rw)
/home/joe       pc001(rw,all_squash,anonuid=150,anongid=100)
/pub            *(ro,insecure,all_squash)
/srv/www        \-sync,rw server @trusted @external(ro)
/foo            2001:db8:9:e54::/64(rw) 192.0.2.0/24(rw)
/build          buildhost[0-9].local.domain(rw)
.\"/pub/private    (noaccess)
.fi
.PP
The first line exports the entire filesystem to machines master and trusty.
In addition to write access, all uid squashing is turned off for host
trusty. The second and third entry show examples for wildcard hostnames
and netgroups (this is the entry `@trusted'). The fourth line shows the
entry for the PC/NFS client discussed above. Line 5 exports the
public FTP directory to every host in the world, executing all requests
under the nobody account. The
.I insecure
option in this entry also allows clients with NFS implementations that
don't use a reserved port for NFS.
The sixth line exports a directory read-write to the machine 'server'
as well as the `@trusted' netgroup, and read-only to netgroup `@external',
all three mounts with the `sync' option enabled. The seventh line exports
a directory to both an IPv6 and an IPv4 subnet. The eighth line demonstrates
a character class wildcard match.
.\" The last line denies all NFS clients
.\"access to the private directory.
.\".SH CAVEATS
.\"Unlike other NFS server implementations, this
.\".B nfsd
.\"allows you to export both a directory and a subdirectory thereof to
.\"the same host, for instance
.\".IR /usr " and " /usr/X11R6 .
.\"In this case, the mount options of the most specific entry apply. For
.\"instance, when a user on the client host accesses a file in
.\".IR /usr/X11R6 ,
.\"the mount options given in the
.\".I /usr/X11R6
.\"entry apply. This is also true when the latter is a wildcard or netgroup
.\"entry.
.SH FILES
/etc/exports
/etc/exports.d
.SH SEE ALSO
.BR exportfs (8),
.BR netgroup (5),
.BR mountd (8),
.BR nfsd (8),
.BR showmount (8).
.\".SH DIAGNOSTICS
.\"An error parsing the file is reported using syslogd(8) as level NOTICE from
.\"a DAEMON whenever
.\".BR nfsd (8)
.\"or
.\".BR mountd (8)
.\"is started up.  Any unknown
.\"host is reported at that time, but often not all hosts are not yet known
.\"to
.\".BR named (8)
.\"at boot time, thus as hosts are found they are reported
.\"with the same
.\".BR syslogd (8)
.\"parameters.