1 .\" Copyright (c) 2009 Rick Macklem, University of Guelph
2 .\" All rights reserved.
4 .\" Redistribution and use in source and binary forms, with or without
5 .\" modification, are permitted provided that the following conditions
7 .\" 1. Redistributions of source code must retain the above copyright
8 .\" notice, this list of conditions and the following disclaimer.
9 .\" 2. Redistributions in binary form must reproduce the above copyright
10 .\" notice, this list of conditions and the following disclaimer in the
11 .\" documentation and/or other materials provided with the distribution.
13 .\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
14 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
16 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
17 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
18 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
19 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
20 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
21 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
22 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 .Nd NFS Version 4 Protocol
34 The NFS client and server provides support for the
37 .%T "Network File System (NFS) Version 4 Protocol RFC 7530" ,
38 .%T "Network File System (NFS) Version 4 Minor Version 1 Protocol RFC 5661" ,
39 .%T "Network File System (NFS) Version 4 Minor Version 2 Protocol RFC 7862" ,
40 .%T "File System Extended Attributes in NFSv4 RFC 8276" and
41 .%T "Parallel NFS (pNFS) Flexible File Layout RFC 8435" .
42 The protocol is somewhat similar to NFS Version 3, but differs in significant
44 It uses a single compound RPC that concatenates operations to-gether.
45 Each of these operations are similar to the RPCs of NFS Version 3.
46 The operations in the compound are performed in order, until one of
47 them fails (returns an error) and then the RPC terminates at that point.
50 integrated locking support, which implies that the server is no longer
54 server remains in recovery mode for a grace period (always greater than the
55 lease duration the server uses) after a reboot.
56 During this grace period, clients may recover state but not perform other
57 open/lock state changing operations.
58 To provide for correct recovery semantics, a small file described by
60 is used by the server during the recovery phase.
61 If this file is missing or empty, there is a backup copy maintained by
64 If either file is missing, they will be created by the
66 If both the file and the backup copy are empty,
67 it will result in the server starting without providing a grace period
69 Note that recovery only occurs when the server
70 machine is rebooted, not when the
74 It provides several optional features not present in NFS Version 3:
76 .Bd -literal -offset indent -compact
78 - Referrals, which redirect subtrees to other servers
80 - Delegations, which allow a client to operate on a file locally
81 - pNFS, where I/O operations are separated from Metadata operations
83 - User namespace extended attributes
84 - lseek(SEEK_DATA/SEEK_HOLE)
85 - File copying done locally on the server for copy_file_range(2)
87 - posix_fadvise(POSIX_FADV_WILLNEED/POSIX_FADV_DONTNEED)
92 protocol does not use a separate mount protocol and assumes that the
93 server provides a single file system tree structure, rooted at the point
94 in the local file system tree specified by one or more
96 .Bd -literal -offset indent -compact
97 V4: <rootdir> [-sec=secflavors] [host(s) or net]
108 allows a limited subset of operations to be performed on non-exported subtrees
109 of the local file system, so that traversal of the tree to the exported
110 subtrees is possible.
111 As such, the ``<rootdir>'' can be in a non-exported file system.
112 The exception is ZFS, which checks exports and, as such, all ZFS file systems
113 below the ``<rootdir>'' must be exported.
115 the entire tree that is rooted at that point must be in local file systems
116 that are of types that can be NFS exported.
119 file system is rooted at ``<rootdir>'', setting this to anything other
120 than ``/'' will result in clients being required to use different mount
123 than for NFS Version 2 or 3.
124 Unlike NFS Version 2 and 3, Version 4 allows a client mount to span across
125 multiple server file systems, although not all clients are capable of doing
129 uses strings for users and groups instead of numbers.
130 On the wire, these strings can either have the numbers in the string or
133 .Bd -literal -offset indent -compact
137 where ``<dns.domain>'' is not the same as the DNS domain used
138 for host name lookups, but is usually set to the same string.
139 Most systems set this ``<dns.domain>''
140 to the domain name part of the machine's
143 However, this can normally be overridden by a command line
144 option or configuration file for the daemon used to do the name<->number
148 the mapping daemon is called
150 and has a command line option that overrides the domain component of the
152 For use of this form of string on
154 either client or server, this daemon must be running.
156 The form where the numbers are in the strings can only be used for AUTH_SYS.
157 To configure your systems this way, the
159 daemon does not need to be running on the server, but the following sysctls need to be
160 set to 1 on the server.
162 .Bd -literal -offset indent -compact
163 vfs.nfs.enable_uidtostring
164 vfs.nfsd.enable_stringtouid
167 On the client, the sysctl
169 .Bd -literal -offset indent -compact
170 vfs.nfs.enable_uidtostring
173 must be set to 1 and the
175 daemon does not need to be running.
177 If these strings are not configured correctly, ``ls -l'' will typically
178 report a lot of ``nobody'' and ``nogroup'' ownerships.
180 Although uid/gid numbers are no longer used in the
182 protocol except optionally in the above strings, they will still be in the RPC authentication fields when
183 using AUTH_SYS (sec=sys), which is the default.
184 As such, in this case both the user/group name and number spaces must
185 be consistent between the client and server.
189 with RPCSEC_GSS (sec=krb5, krb5i, krb5p), only names and KerberosV tickets
192 To set up the NFS server that supports
194 you will need to set the variables in
198 .Bd -literal -offset indent -compact
199 nfs_server_enable="YES"
200 nfsv4_server_enable="YES"
205 .Bd -literal -offset indent -compact
206 nfsuserd_enable="YES"
209 if the server is using the ``<user>@<domain>'' form of user/group strings or
210 is using the ``-manage-gids'' option for
213 You will also need to add at least one ``V4:'' line to the
219 If the file systems you are exporting are only being accessed via
221 there are a couple of
223 variables that you can change, which might improve performance.
225 .It Cm vfs.nfsd.issue_delegations
226 when set non-zero, allows the server to issue Open Delegations to
228 These delegations permit the client to manipulate the file
229 locally on the client.
230 Unfortunately, at this time, client use of
231 delegations is limited, so performance gains may not be observed.
232 This can only be enabled when the file systems being exported to
234 clients are not being accessed locally on the server and, if being
235 accessed via NFS Version 2 or 3 clients, these clients cannot be
237 .It Cm vfs.nfsd.enable_locallocks
238 can be set to 0 to disable acquisition of local byte range locks.
239 Disabling local locking can only be done if neither local accesses
240 to the exported file systems nor the NLM is operating on them.
243 Note that Samba server access would be considered ``local access'' for the above
246 To build a kernel with the NFS server that supports
250 .Bd -literal -offset indent -compact
254 must be specified in the kernel's
260 mount, specify the ``nfsv4'' option on the
263 This will force use of the client that supports
270 must be running if name<->uid/gid mapping is being used, as above.
273 mount uses the host uuid to identify the client uniquely to the server,
274 you cannot safely do an
278 .Bd -literal -offset indent -compact
287 server that is being mounted on supports delegations, you can start the
289 daemon to handle client side callbacks.
292 .Bd -literal -offset indent -compact
293 nfsuserd_enable="YES" <-- If name<->uid/gid mapping is being used.
300 Without a functioning callback path, a server will never issue Delegations
303 For NFSv4.0, by default, the callback address will be set to the IP address acquired via
305 in the kernel and port# 7745.
306 To override the default port#, a command line option for
310 To get callbacks to work when behind a NAT gateway, a port for the callback
311 service will need to be set up on the NAT gateway and then the address
312 of the NAT gateway (host IP plus port#) will need to be set by assigning the
314 variable vfs.nfs.callback_addr to a string of the form:
318 where the first 4 Ns are the host IP address and the last two are the
319 port# in network byte order (all decimal #s in the range 0-255).
321 For NFSv4.1 and NFSv4.2, the callback path (called a backchannel) uses the same TCP connection as the mount,
322 so none of the above applies and should work through gateways without
325 To build a kernel with the client that supports
327 linked into it, the option
329 .Bd -literal -offset indent -compact
333 must be specified in the kernel's
337 Options can be specified for the
341 daemons at boot time via the ``nfsuserd_flags'' and ``nfscbd_flags''
345 NFSv4 mount(s) against exported volume(s) on the same host are not recommended,
346 since this can result in a hung NFS server.
347 It occurs when an nfsd thread tries to do an NFSv4
349 / Close RPC as part of acquiring a new vnode.
350 If all other nfsd threads are blocked waiting for lock(s) held by this nfsd
351 thread, then there isn't an nfsd thread to service the Close RPC.
353 .Bl -tag -width /var/db/nfs-stablerestart.bak -compact
354 .It Pa /var/db/nfs-stablerestart
355 NFS V4 stable restart file
356 .It Pa /var/db/nfs-stablerestart.bak
357 backup copy of the file
360 .Xr stablerestart 5 ,
368 At this time, there is no recall of delegations for local file system
370 As such, delegations should only be enabled for file systems
371 that are being used solely as NFS export volumes and are not being accessed
372 via local system calls nor services such as Samba.