freebsd-skq/usr.sbin/mountd/exports.5

513 lines
15 KiB
Groff
Raw Normal View History

.\" Copyright (c) 1989, 1991, 1993
.\" The Regents of the University of California. All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\" 4. Neither the name of the University nor the names of its contributors
.\" may be used to endorse or promote products derived from this software
.\" without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" @(#)exports.5 8.3 (Berkeley) 3/29/95
1999-08-28 00:22:10 +00:00
.\" $FreeBSD$
.\"
.Dd December 23, 2012
.Dt EXPORTS 5
.Os
.Sh NAME
.Nm exports
.Nd define remote mount points for
.Tn NFS
mount requests
.Sh SYNOPSIS
.Nm
.Sh DESCRIPTION
The
.Nm
file specifies remote mount points for the
.Tn NFS
mount protocol per the
.Tn NFS
server specification; see
2000-12-18 15:16:24 +00:00
.%T "Network File System Protocol Specification" ,
RFC1094, Appendix A and
.%T "NFS: Network File System Version 3 Specification" ,
Appendix I.
.Pp
Each line in the file
(other than comment lines that begin with a #)
specifies the mount point(s) and export flags within one local server
file system or the NFSv4 tree root for one or more hosts.
A long line may be split over several lines by ending all but the
2001-07-04 13:57:34 +00:00
last line with a backslash
.Pq Ql \e .
A host may be specified only once for each local file or the NFSv4 tree root on the
server and there may be only one default entry for each server
file system that applies to all other hosts.
The latter exports the file system to the
.Dq world
and should
be used only when the file system contains public information.
.Pp
In a mount entry,
the first field(s) specify the directory path(s) within a server file system
that can be mounted on by the corresponding client(s).
There are three forms of this specification.
The first is to list all mount points as absolute
directory paths separated by whitespace.
This list of directory paths should be considered an
.Dq administrative control ,
since it is only enforced by the
.Xr mountd 8
daemon and not the kernel.
As such, it only applies to NFSv2 and NFSv3 mounts and only
with respect to the client's use of the mount protocol.
The second is to specify the pathname of the root of the file system
followed by the
.Fl alldirs
flag;
this form allows the host(s) to mount at any point within the file system,
including regular files if the
.Fl r
option is used on
.Xr mountd 8 .
Because NFSv4 does not use the mount protocol,
the
.Dq administrative controls
are not applied.
Thus, all the above export line(s) should be considered to have the
.Fl alldirs
flag, even if the line is specified without it.
The third form has the string ``V4:'' followed by a single absolute path
name, to specify the NFSv4 tree root.
This line does not export any file system, but simply marks where the root
of the server's directory tree is for NFSv4 clients.
The exported file systems for NFSv4 are specified via the other lines
in the
2012-10-07 19:53:04 +00:00
.Nm
file in the same way as for NFSv2 and NFSv3.
The pathnames must not have any symbolic links in them and should not have
any
.Dq Pa \&.
or
.Dq Pa ..
components.
Mount points for a file system may appear on multiple lines each with
different sets of hosts and export options.
.Pp
The second component of a line specifies how the file system is to be
exported to the host set.
The option flags specify whether the file system
is exported read-only or read-write and how the client UID is mapped to
user credentials on the server.
For the NFSv4 tree root, the only option that can be specified in this
section is
.Fl sec .
.Pp
Export options are specified as follows:
.Pp
.Sm off
2008-07-01 19:54:02 +00:00
.Fl maproot Li = Sy user
.Sm on
The credential of the specified user is used for remote access by root.
The credential includes all the groups to which the user is a member
on the local machine (see
2001-01-16 09:15:57 +00:00
.Xr id 1 ) .
The user may be specified by name or number.
.Pp
.Sm off
2008-07-01 19:54:02 +00:00
.Fl maproot Li = Sy user:group1:group2:...
.Sm on
The colon separated list is used to specify the precise credential
to be used for remote access by root.
The elements of the list may be either names or numbers.
Note that user: should be used to distinguish a credential containing
no groups from a complete credential for that user.
.Pp
.Sm off
2008-07-01 19:54:02 +00:00
.Fl mapall Li = Sy user
.Sm on
or
.Sm off
2008-07-01 19:54:02 +00:00
.Fl mapall Li = Sy user:group1:group2:...
.Sm on
specifies a mapping for all client UIDs (including root)
using the same semantics as
.Fl maproot .
.Pp
The option
.Fl r
is a synonym for
.Fl maproot
in an effort to be backward compatible with older export file formats.
.Pp
In the absence of
.Fl maproot
and
.Fl mapall
options, remote accesses by root will result in using a credential of -2:-2.
All other users will be mapped to their remote credential.
If a
.Fl maproot
option is given,
remote access by root will be mapped to that credential instead of -2:-2.
If a
.Fl mapall
option is given,
all users (including root) will be mapped to that credential in
place of their own.
.Pp
Implement support for RPCSEC_GSS authentication to both the NFS client and server. This replaces the RPC implementation of the NFS client and server with the newer RPC implementation originally developed (actually ported from the userland sunrpc code) to support the NFS Lock Manager. I have tested this code extensively and I believe it is stable and that performance is at least equal to the legacy RPC implementation. The NFS code currently contains support for both the new RPC implementation and the older legacy implementation inherited from the original NFS codebase. The default is to use the new implementation - add the NFS_LEGACYRPC option to fall back to the old code. When I merge this support back to RELENG_7, I will probably change this so that users have to 'opt in' to get the new code. To use RPCSEC_GSS on either client or server, you must build a kernel which includes the KGSSAPI option and the crypto device. On the userland side, you must build at least a new libc, mountd, mount_nfs and gssd. You must install new versions of /etc/rc.d/gssd and /etc/rc.d/nfsd and add 'gssd_enable=YES' to /etc/rc.conf. As long as gssd is running, you should be able to mount an NFS filesystem from a server that requires RPCSEC_GSS authentication. The mount itself can happen without any kerberos credentials but all access to the filesystem will be denied unless the accessing user has a valid ticket file in the standard place (/tmp/krb5cc_<uid>). There is currently no support for situations where the ticket file is in a different place, such as when the user logged in via SSH and has delegated credentials from that login. This restriction is also present in Solaris and Linux. In theory, we could improve this in future, possibly using Brooks Davis' implementation of variant symlinks. Supporting RPCSEC_GSS on a server is nearly as simple. You must create service creds for the server in the form 'nfs/<fqdn>@<REALM>' and install them in /etc/krb5.keytab. The standard heimdal utility ktutil makes this fairly easy. After the service creds have been created, you can add a '-sec=krb5' option to /etc/exports and restart both mountd and nfsd. The only other difference an administrator should notice is that nfsd doesn't fork to create service threads any more. In normal operation, there will be two nfsd processes, one in userland waiting for TCP connections and one in the kernel handling requests. The latter process will create as many kthreads as required - these should be visible via 'top -H'. The code has some support for varying the number of service threads according to load but initially at least, nfsd uses a fixed number of threads according to the value supplied to its '-n' option. Sponsored by: Isilon Systems MFC after: 1 month
2008-11-03 10:38:00 +00:00
.Sm off
.Fl sec Li = Sy flavor1:flavor2...
.Sm on
specifies a colon separated list of acceptable security flavors to be
used for remote access.
Supported security flavors are sys, krb5, krb5i and krb5p.
If multiple flavors are listed, they should be ordered with the most
preferred flavor first.
If this option is not present,
the default security flavor list of just sys is used.
.Pp
The
.Fl ro
option specifies that the file system should be exported read-only
(default read/write).
The option
.Fl o
is a synonym for
.Fl ro
in an effort to be backward compatible with older export file formats.
.Pp
.Tn WebNFS
exports strictly according to the spec (RFC 2054 and RFC 2055) can
be done with the
.Fl public
flag.
However, this flag in itself allows r/w access to all files in
the file system, not requiring reserved ports and not remapping UIDs.
It
is only provided to conform to the spec, and should normally not be used.
For a
.Tn WebNFS
export,
use the
.Fl webnfs
flag, which implies
.Fl public ,
.Sm off
.Fl mapall No = Sy nobody
.Sm on
and
.Fl ro .
2004-03-27 14:23:25 +00:00
Note that only one file system can be
.Tn WebNFS
exported on a server.
.Pp
A
.Sm off
.Fl index No = Pa file
.Sm on
option can be used to specify a file whose handle will be returned if
a directory is looked up using the public filehandle
.Pq Tn WebNFS .
This is to mimic the behavior of URLs.
If no
.Fl index
option is specified, a directory filehandle will be returned as usual.
The
.Fl index
option only makes sense in combination with the
.Fl public
or
.Fl webnfs
flags.
.Pp
Specifying the
.Fl quiet
option will inhibit some of the syslog diagnostics for bad lines in
.Pa /etc/exports .
This can be useful to avoid annoying error messages for known possible
problems (see
.Sx EXAMPLES
below).
.Pp
The third component of a line specifies the host set to which the line applies.
The set may be specified in three ways.
The first way is to list the host name(s) separated by white space.
(Standard Internet
.Dq dot
addresses may be used in place of names.)
The second way is to specify a
.Dq netgroup
as defined in the
.Pa netgroup
file (see
2001-01-16 09:15:57 +00:00
.Xr netgroup 5 ) .
The third way is to specify an Internet subnetwork using a network and
network mask that is defined as the set of all hosts with addresses within
the subnetwork.
This latter approach requires less overhead within the
kernel and is recommended for cases where the export line refers to a
large number of clients within an administrative subnet.
.Pp
The first two cases are specified by simply listing the name(s) separated
by whitespace.
All names are checked to see if they are
.Dq netgroup
names
first and are assumed to be hostnames otherwise.
Using the full domain specification for a hostname can normally
circumvent the problem of a host that has the same name as a netgroup.
The third case is specified by the flag
.Sm off
2008-07-01 19:54:02 +00:00
.Fl network Li = Sy netname Op Li / Ar prefixlength
.Sm on
and optionally
.Sm off
.Fl mask No = Sy netmask .
.Sm on
The netmask may be specified either by attaching a
.Ar prefixlength
to the
.Fl network
option, or by using a separate
.Fl mask
option.
If the mask is not specified, it will default to the mask for that network
class (A, B or C; see
2001-01-16 09:15:57 +00:00
.Xr inet 4 ) .
See the
.Sx EXAMPLES
section below.
.Pp
Scoped IPv6 address must carry scope identifier as documented in
.Xr inet6 4 .
For example,
.Dq Li fe80::%re2/10
is used to specify
.Li fe80::/10
on
.Li re2
interface.
.Pp
For the third form which specifies the NFSv4 tree root, the directory path
specifies the location within the server's file system tree which is the
root of the NFSv4 tree.
All entries of this form must specify the same directory path.
For file systems other than ZFS,
this location can be any directory and does not
need to be within an exported file system. If it is not in an exported
file system, a very limited set of operations are permitted, so that an
NFSv4 client can traverse the tree to an exported file system.
Although parts of the NFSv4 tree can be non-exported, the entire NFSv4 tree
must consist of local file systems capable of being exported via NFS.
All ZFS file systems in the subtree below the NFSv4 tree root must be
exported.
NFSv4 does not use the mount protocol and does permit clients to cross server
mount point boundaries, although not all clients are capable of crossing the
mount points.
.Pp
The
.Fl sec
option on these line(s) specifies what security flavors may be used for
NFSv4 operations that do not use file handles. Since these operations
(SetClientID, SetClientIDConfirm, Renew, DelegPurge and ReleaseLockOnwer)
allocate/modify state in the server, it is possible to restrict some clients to
the use of the krb5[ip] security flavors, via this option.
See the
.Sx EXAMPLES
section below.
This third form is meaningless for NFSv2 and NFSv3 and is ignored for them.
.Pp
The
.Xr mountd 8
utility can be made to re-read the
.Nm
file by sending it a hangup signal as follows:
.Bd -literal -offset indent
/etc/rc.d/mountd reload
.Ed
.Pp
After sending the
.Dv SIGHUP ,
check the
.Xr syslogd 8
output to see whether
.Xr mountd 8
logged any parsing errors in the
.Nm
file.
.Sh FILES
.Bl -tag -width /etc/exports -compact
.It Pa /etc/exports
the default remote mount-point file
.El
.Sh EXAMPLES
.Bd -literal -offset indent
/usr /usr/local -maproot=0:10 friends
/usr -maproot=daemon grumpy.cis.uoguelph.ca 131.104.48.16
/usr -ro -mapall=nobody
/u -maproot=bin: -network 131.104.48 -mask 255.255.255.0
/a -network 192.168.0/24
/a -network 3ffe:1ce1:1:fe80::/64
/u2 -maproot=root friends
/u2 -alldirs -network cis-net -mask cis-mask
/cdrom -alldirs,quiet,ro -network 192.168.33.0 -mask 255.255.255.0
Implement support for RPCSEC_GSS authentication to both the NFS client and server. This replaces the RPC implementation of the NFS client and server with the newer RPC implementation originally developed (actually ported from the userland sunrpc code) to support the NFS Lock Manager. I have tested this code extensively and I believe it is stable and that performance is at least equal to the legacy RPC implementation. The NFS code currently contains support for both the new RPC implementation and the older legacy implementation inherited from the original NFS codebase. The default is to use the new implementation - add the NFS_LEGACYRPC option to fall back to the old code. When I merge this support back to RELENG_7, I will probably change this so that users have to 'opt in' to get the new code. To use RPCSEC_GSS on either client or server, you must build a kernel which includes the KGSSAPI option and the crypto device. On the userland side, you must build at least a new libc, mountd, mount_nfs and gssd. You must install new versions of /etc/rc.d/gssd and /etc/rc.d/nfsd and add 'gssd_enable=YES' to /etc/rc.conf. As long as gssd is running, you should be able to mount an NFS filesystem from a server that requires RPCSEC_GSS authentication. The mount itself can happen without any kerberos credentials but all access to the filesystem will be denied unless the accessing user has a valid ticket file in the standard place (/tmp/krb5cc_<uid>). There is currently no support for situations where the ticket file is in a different place, such as when the user logged in via SSH and has delegated credentials from that login. This restriction is also present in Solaris and Linux. In theory, we could improve this in future, possibly using Brooks Davis' implementation of variant symlinks. Supporting RPCSEC_GSS on a server is nearly as simple. You must create service creds for the server in the form 'nfs/<fqdn>@<REALM>' and install them in /etc/krb5.keytab. The standard heimdal utility ktutil makes this fairly easy. After the service creds have been created, you can add a '-sec=krb5' option to /etc/exports and restart both mountd and nfsd. The only other difference an administrator should notice is that nfsd doesn't fork to create service threads any more. In normal operation, there will be two nfsd processes, one in userland waiting for TCP connections and one in the kernel handling requests. The latter process will create as many kthreads as required - these should be visible via 'top -H'. The code has some support for varying the number of service threads according to load but initially at least, nfsd uses a fixed number of threads according to the value supplied to its '-n' option. Sponsored by: Isilon Systems MFC after: 1 month
2008-11-03 10:38:00 +00:00
/private -sec=krb5i
/secret -sec=krb5p
V4: / -sec=krb5:krb5i:krb5p -network 131.104.48 -mask 255.255.255.0
V4: / -sec=sys:krb5:krb5i:krb5p grumpy.cis.uoguelph.ca
.Ed
.Pp
Given that
2008-07-01 19:54:02 +00:00
.Pa /usr , /u , /a
and
.Pa /u2
are
local file system mount points, the above example specifies the following:
.Pp
The file system rooted at
.Pa /usr
is exported to hosts
.Em friends
where friends is specified in the netgroup file
with users mapped to their remote credentials and
root mapped to UID 0 and group 10.
It is exported read-write and the hosts in
.Dq friends
can mount either
.Pa /usr
or
.Pa /usr/local .
It is exported to
.Em 131.104.48.16
and
.Em grumpy.cis.uoguelph.ca
with users mapped to their remote credentials and
root mapped to the user and groups associated with
.Dq daemon ;
it is exported to the rest of the world as read-only with
all users mapped to the user and groups associated with
.Dq nobody .
.Pp
The file system rooted at
.Pa /u
is exported to all hosts on the subnetwork
.Em 131.104.48
with root mapped to the UID for
.Dq bin
and with no group access.
.Pp
The file system rooted at
.Pa /u2
is exported to the hosts in
.Dq friends
with root mapped to UID and groups
associated with
.Dq root ;
it is exported to all hosts on network
.Dq cis-net
allowing mounts at any
directory within /u2.
.Pp
The file system rooted at
.Pa /a
is exported to the network 192.168.0.0, with a netmask of 255.255.255.0.
However, the netmask length in the entry for
.Pa /a
is not specified through a
.Fl mask
option, but through the
.Li / Ns Ar prefix
notation.
.Pp
The file system rooted at
.Pa /a
is also exported to the IPv6 network
.Li 3ffe:1ce1:1:fe80::
address, using the upper 64 bits as the prefix.
Note that, unlike with IPv4 network addresses, the specified network
address must be complete, and not just contain the upper bits.
With IPv6 addresses, the
.Fl mask
option must not be used.
.Pp
The file system rooted at
.Pa /cdrom
will be exported read-only to the entire network 192.168.33.0/24, including
all its subdirectories.
Since
.Pa /cdrom
is the conventional mountpoint for a CD-ROM device, this export will
fail if no CD-ROM medium is currently mounted there since that line
would then attempt to export a subdirectory of the root file system
with the
.Fl alldirs
option which is not allowed.
The
.Fl quiet
option will then suppress the error message for this condition that
would normally be syslogged.
As soon as an actual CD-ROM is going to be mounted,
.Xr mount 8
will notify
.Xr mountd 8
about this situation, and the
.Pa /cdrom
2002-12-27 12:15:40 +00:00
file system will be exported as intended.
Note that without using the
.Fl alldirs
option, the export would always succeed.
While there is no CD-ROM medium mounted under
.Pa /cdrom ,
it would export the (normally empty) directory
.Pa /cdrom
of the root file system instead.
Implement support for RPCSEC_GSS authentication to both the NFS client and server. This replaces the RPC implementation of the NFS client and server with the newer RPC implementation originally developed (actually ported from the userland sunrpc code) to support the NFS Lock Manager. I have tested this code extensively and I believe it is stable and that performance is at least equal to the legacy RPC implementation. The NFS code currently contains support for both the new RPC implementation and the older legacy implementation inherited from the original NFS codebase. The default is to use the new implementation - add the NFS_LEGACYRPC option to fall back to the old code. When I merge this support back to RELENG_7, I will probably change this so that users have to 'opt in' to get the new code. To use RPCSEC_GSS on either client or server, you must build a kernel which includes the KGSSAPI option and the crypto device. On the userland side, you must build at least a new libc, mountd, mount_nfs and gssd. You must install new versions of /etc/rc.d/gssd and /etc/rc.d/nfsd and add 'gssd_enable=YES' to /etc/rc.conf. As long as gssd is running, you should be able to mount an NFS filesystem from a server that requires RPCSEC_GSS authentication. The mount itself can happen without any kerberos credentials but all access to the filesystem will be denied unless the accessing user has a valid ticket file in the standard place (/tmp/krb5cc_<uid>). There is currently no support for situations where the ticket file is in a different place, such as when the user logged in via SSH and has delegated credentials from that login. This restriction is also present in Solaris and Linux. In theory, we could improve this in future, possibly using Brooks Davis' implementation of variant symlinks. Supporting RPCSEC_GSS on a server is nearly as simple. You must create service creds for the server in the form 'nfs/<fqdn>@<REALM>' and install them in /etc/krb5.keytab. The standard heimdal utility ktutil makes this fairly easy. After the service creds have been created, you can add a '-sec=krb5' option to /etc/exports and restart both mountd and nfsd. The only other difference an administrator should notice is that nfsd doesn't fork to create service threads any more. In normal operation, there will be two nfsd processes, one in userland waiting for TCP connections and one in the kernel handling requests. The latter process will create as many kthreads as required - these should be visible via 'top -H'. The code has some support for varying the number of service threads according to load but initially at least, nfsd uses a fixed number of threads according to the value supplied to its '-n' option. Sponsored by: Isilon Systems MFC after: 1 month
2008-11-03 10:38:00 +00:00
.Pp
The file system rooted at
.Pa /private
will be exported using Kerberos 5 authentication and will require
integrity protected messages for all accesses.
The file system rooted at
.Pa /secret
will also be exported using Kerberos 5 authentication and all messages
used to access it will be encrypted.
.Pp
For the experimental server, the NFSv4 tree is rooted at ``/'',
and any client within the 131.104.48 subnet is permitted to perform NFSv4 state
operations on the server, so long as valid Kerberos credentials are provided.
The machine grumpy.cis.uoguelph.ca is permitted to perform NFSv4 state
operations on the server using AUTH_SYS credentials, as well as Kerberos ones.
.Sh SEE ALSO
.Xr nfsv4 4 ,
.Xr netgroup 5 ,
.Xr mountd 8 ,
.Xr nfsd 8 ,
.Xr showmount 8
.Sh BUGS
The export options are tied to the local mount points in the kernel and
must be non-contradictory for any exported subdirectory of the local
server mount point.
It is recommended that all exported directories within the same server
file system be specified on adjacent lines going down the tree.
You cannot specify a hostname that is also the name of a netgroup.
Specifying the full domain specification for a hostname can normally
circumvent the problem.