2 .TH NFS 5 "2 November 2007"
4 nfs \- fstab format and options for the
12 NFS is an Internet Standard protocol
13 created by Sun Microsystems in 1984. NFS was developed
14 to allow file sharing between systems residing
15 on a local area network.
16 The Linux NFS client supports three versions
18 NFS version 2 [RFC1094],
19 NFS version 3 [RFC1813],
20 and NFS version 4 [RFC3530].
24 command attaches a file system to the system's
25 name space hierarchy at a given mount point.
30 should assemble a system's file name hierarchy
31 from various independent file systems
32 (including file systems exported by NFS servers).
35 file describes a single file system, its mount point,
36 and a set of default mount options for that mount point.
38 For NFS file system mounts, a line in the
40 file specifies the server name,
41 the path name of the exported server directory to mount,
42 the local directory that is the mount point,
43 the type of file system that is being mounted,
44 and a list of mount options that control
45 the way the filesystem is mounted and
46 how the NFS client behaves when accessing
47 files on this mount point.
48 The fifth and sixth fields on each line are not used
49 by NFS, thus conventionally each contain the digit zero. For example:
53 .TA 2.5i +0.75i +0.75i +1.0i
54 server:path /mountpoint fstype option,option,... 0 0
57 The server's hostname and export pathname
58 are separated by a colon, while
59 the mount options are separated by commas. The remaining fields
60 are separated by blanks or tabs.
61 The server's hostname can be an unqualified hostname,
62 a fully qualified domain name,
63 or a dotted quad IPv4 address.
66 field contains either "nfs" (for version 2 or version 3 NFS mounts)
67 or "nfs4" (for NFS version 4 mounts).
72 file system types share similar mount options,
73 which are described below.
77 for a description of generic mount options
78 available for all file systems. If you do not need to
79 specify any mount options, use the generic option
85 .SS "Valid options for either the nfs or nfs4 file system type"
86 These options are valid to use when mounting either
91 They imply the same behavior
92 and have the same default for both file system types.
95 Determines the recovery behavior of the NFS client
96 after an NFS request times out.
97 If neither option is specified (or if the
99 option is specified), NFS requests are retried indefinitely.
102 option is specified, then the NFS client fails an NFS request
105 retransmissions have been sent,
106 causing the NFS client to return an error
107 to the calling application.
110 A so-called "soft" timeout can cause
111 silent data corruption in certain cases. As such, use the
113 option only when client responsiveness
114 is more important than data integrity.
115 Using NFS over TCP or increasing the value of the
117 option may mitigate some of the risks of using the
122 The time (in tenths of a second) the NFS client waits for a
123 response before it retries an NFS request. If this
124 option is not specified, requests are retried every
125 60 seconds for NFS over TCP.
126 The NFS client does not perform any kind of timeout backoff
129 However, for NFS over UDP, the client uses an adaptive
130 algorithm to estimate an appropriate timeout value for frequently used
131 request types (such as READ and WRITE requests), but uses the
133 setting for infrequently used request types (such as FSINFO requests).
136 option is not specified,
137 infrequently used request types are retried after 1.1 seconds.
138 After each retransmission, the NFS client doubles the timeout for
140 up to a maximum timeout length of 60 seconds.
143 The number of times the NFS client retries a request before
144 it attempts further recovery action. If the
146 option is not specified, the NFS client tries each request
149 The NFS client generates a "server not responding" message
152 retries, then attempts further recovery (depending on whether the
154 mount option is in effect).
157 The maximum number of bytes in each network READ request
158 that the NFS client can receive when reading data from a file
160 The actual data payload size of each NFS READ request is equal to
163 setting. The largest read payload supported by the Linux NFS client
164 is 1,048,576 bytes (one megabyte).
168 value is a positive integral multiple of 1024.
171 values lower than 1024 are replaced with 4096; values larger than
172 1048576 are replaced with 1048576. If a specified value is within the supported
173 range but not a multiple of 1024, it is rounded down to the nearest
178 value is not specified, or if the specified
180 value is larger than the maximum that either client or server can support,
181 the client and server negotiate the largest
183 value that they can both support.
187 mount option as specified on the
189 command line appears in the
191 file. However, the effective
193 value negotiated by the client and server is reported in the
198 The maximum number of bytes per network WRITE request
199 that the NFS client can send when writing data to a file
200 on an NFS server. The actual data payload size of each
201 NFS WRITE request is equal to
204 setting. The largest write payload supported by the Linux NFS client
205 is 1,048,576 bytes (one megabyte).
211 value is a positive integral multiple of 1024.
214 values lower than 1024 are replaced with 4096; values larger than
215 1048576 are replaced with 1048576. If a specified value is within the supported
216 range but not a multiple of 1024, it is rounded down to the nearest
221 value is not specified, or if the specified
223 value is larger than the maximum that either client or server can support,
224 the client and server negotiate the largest
226 value that they can both support.
230 mount option as specified on the
232 command line appears in the
234 file. However, the effective
236 value negotiated by the client and server is reported in the
241 Selects whether the client may cache file attributes. If neither
242 option is specified (or if
244 is specified), the client caches file
247 To improve performance, NFS clients cache file
248 attributes. Every few seconds, an NFS client checks the server's version of each
249 file's attributes for updates. Changes that occur on the server in
250 those small intervals remain undetected until the client checks the
253 option prevents clients from caching file
254 attributes so that applications can more quickly detect file changes
257 In addition to preventing the client from caching file attributes,
260 option forces application writes to become synchronous so
261 that local changes to a file become visible on the server
262 immediately. That way, other clients can quickly detect recent
263 writes when they check the file's attributes.
267 option provides greater cache coherence among NFS clients
268 accessing the same files,
269 but it extracts a significant performance penalty.
270 As such, judicious use of file locking is encouraged instead.
271 The DATA AND METADATA COHERENCE section contains a detailed discussion
275 The minimum time (in seconds) that the NFS client caches
276 attributes of a regular file before it requests
277 fresh attribute information from a server.
278 If this option is not specified, the NFS client uses
282 The maximum time (in seconds) that the NFS client caches
283 attributes of a regular file before it requests
284 fresh attribute information from a server.
285 If this option is not specified, the NFS client uses
289 The minimum time (in seconds) that the NFS client caches
290 attributes of a directory before it requests
291 fresh attribute information from a server.
292 If this option is not specified, the NFS client uses
296 The maximum time (in seconds) that the NFS client caches
297 attributes of a directory before it requests
298 fresh attribute information from a server.
299 If this option is not specified, the NFS client uses
312 If this option is not specified, the NFS client uses
313 the defaults for each of these options listed above.
318 command behaves if an attempt to mount an export fails.
323 to exit with an error status if any part of the mount request
324 times out or fails outright.
325 This is called a "foreground" mount,
326 and is the default behavior if neither the
330 mount option is specified.
334 option is specified, a timeout or failure causes the
336 command to fork a child which continues to attempt
338 The parent immediately returns with a zero exit code.
339 This is known as a "background" mount.
341 If the local mount point directory is missing, the
343 command acts as if the mount request timed out.
344 This permits nested NFS mounts specified in
346 to proceed in any order during system initialization,
347 even if some NFS servers are not yet available.
348 Alternatively these issues can be addressed
349 using an automounter (refer to
354 The number of minutes that the
356 command retries an NFS mount operation
357 in the foreground or background before giving up.
358 If this option is not specified, the default value for foreground mounts
359 is 2 minutes, and the default value for background mounts is 10000 minutes (80 minutes shy of one week).
362 The RPCGSS security flavor to use for accessing files on this mount point.
365 option is not specified, or if
367 is specified, the NFS client uses the AUTH_SYS security flavor
368 for all NFS requests on this mount point.
369 Valid security flavors are
382 Refer to the SECURITY CONSIDERATIONS section for details.
384 .BR sharecache " / " nosharecache
385 Determines how the client's data cache and attribute cache are shared
386 when mounting the same export more than once concurrently. Using the
387 same cache reduces memory requirements on the client and presents
388 identical file contents to applications when the same remote file is
389 accessed via different mount points.
391 If neither option is specified, or if the
394 specified, then a single cache is used for all mount points that
395 access the same export. If the
398 then that mount point gets a unique cache. Note that when data and
399 attribute caches are shared, the mount options from the first mount
400 point take effect for subsequent concurrent mounts of the same export.
402 As of kernel 2.6.18, the behavior specified by
404 is legacy caching behavior. This
405 is considered a data risk since multiple cached copies
406 of the same file on the same client can become out of sync
407 following a local update of one of the copies.
408 .SS "Valid options for the nfs file system type"
409 Use these options, along with the options in the above subsection,
415 The transport the NFS client uses
416 to transmit requests to the NFS server for this mount point.
422 Each transport uses different default
426 settings; refer to the description of these two mount options for details.
428 In addition to controlling how the NFS client transmits requests to
429 the server, this mount option also controls how the
431 command communicates with the server's rpcbind and mountd services.
434 forces all traffic from the
436 command and the NFS client to use TCP.
439 forces all traffic types to use UDP.
443 mount option is not specified, the
445 command discovers which protocols the server supports
446 and chooses an appropriate transport for each service.
447 Refer to the TRANSPORT METHODS section for more details.
452 option is an alternative to specifying
454 It is included for compatibility with other operating systems.
459 option is an alternative to specifying
461 It is included for compatibility with other operating systems.
464 The numeric value of the server's NFS service port.
465 If the server's NFS service is not available on the specified port,
466 the mount request fails.
468 If this option is not specified, or if the specified port value is 0,
469 then the NFS client uses the NFS service port number
470 advertised by the server's rpcbind service.
471 The mount request fails if the server's rpcbind service is not available,
472 the server's NFS service is not registered with its rpcbind service,
473 or the server's NFS service is not available on the advertised port.
476 The numeric value of the server's mountd port.
477 If the server's mountd service is not available on the specified port,
478 the mount request fails.
480 If this option is not specified,
481 or if the specified port value is 0, then the
483 command uses the mountd service port number
484 advertised by the server's rpcbind service.
485 The mount request fails if the server's rpcbind service is not available,
486 the server's mountd service is not registered with its rpcbind service,
487 or the server's mountd service is not available on the advertised port.
489 This option can be used when mounting an NFS server
490 through a firewall that blocks the rpcbind protocol.
492 .BI mountproto= transport
493 The transport the NFS client uses
494 to transmit requests to the NFS server's mountd service when performing
495 this mount request, and when later unmounting this mount point.
502 This option can be used when mounting an NFS server
503 through a firewall that blocks a particular transport.
504 When used in combination with the
506 option, different transports for mountd requests and NFS requests
508 If the server's mountd service is not available via the specified
509 transport, the mount request fails.
510 Refer to the TRANSPORT METHODS section for more on how the
512 mount option interacts with the
517 The hostname of the host running mountd.
518 If this option is not specified, the
520 command assumes that the mountd service runs
521 on the same host as the NFS service.
524 The RPC version number used to contact the server's mountd.
525 If this option is not specified, the client uses a version number
526 appropriate to the requested NFS version.
527 This option is useful when multiple NFS services
528 are running on the same remote server host.
531 The maximum length of a pathname component on this mount.
532 If this option is not specified, the maximum length is negotiated
533 with the server. In most cases, this maximum length is 255 characters.
535 Some early versions of NFS did not support this negotiation.
536 Using this option ensures that
538 reports the proper maximum component length to applications
542 The NFS protocol version number used to contact the server's NFS service.
543 The Linux client supports version 2 and version 3 of the NFS protocol
544 when using the file system type
546 If the server does not support the requested version,
547 the mount request fails.
548 If this option is not specified, the client attempts to use version 3,
549 but negotiates the NFS version with the server if version 3 support
553 This option is an alternative to the
556 It is included for compatibility with other operating systems.
558 .BR lock " / " nolock
559 Selects whether to use the NLM sideband protocol to lock files on the server.
560 If neither option is specified (or if
562 is specified), NLM locking is used for this mount point.
565 option, applications can lock files,
566 but such locks provide exclusion only against other applications
567 running on the same client.
568 Remote applications are not affected by these locks.
570 NLM locking must be disabled with the
572 option when using NFS to mount
576 contains files used by the NLM implementation on Linux.
579 option is also required when mounting exports on NFS servers
580 that do not support the NLM protocol.
582 .BR intr " / " nointr
583 Selects whether to allow signals to interrupt file operations
584 on this mount point. If neither option
588 signals do not interrupt NFS file operations. If
590 is specified, system calls return EINTR if an in-progress NFS operation is interrupted by
595 option is preferred to using the
597 option because it is significantly less likely to result in data corruption.
600 Selects whether to use close-to-open cache coherence semantics.
601 If neither option is specified (or if
603 is specified), the client uses close-to-open
604 cache coherence semantics. If the
606 option is specified, the client uses a non-standard heuristic to determine when
607 files on the server have changed.
611 option may improve performance for read-only mounts,
612 but should be used only if the data on the server changes only occasionally.
613 The DATA AND METADATA COHERENCE section discusses the behavior
614 of this option in more detail.
617 Selects whether to use the NFSACL sideband protocol on this mount point.
618 The NFSACL sideband protocol is a proprietary protocol
619 implemented in Solaris that manages Access Control Lists. NFSACL was never
620 made a standard part of the NFS protocol specification.
627 the NFS client negotiates with the server
628 to see if the NFSACL protocol is supported,
629 and uses it if the server supports it.
630 Disabling the NFSACL sideband protocol may be necessary
631 if the negotiation causes problems on the client or server.
632 Refer to the SECURITY CONSIDERATIONS section for more details.
634 .BR rdirplus " / " nordirplus
635 Selects whether to use NFS version 3 READDIRPLUS requests.
636 If this option is not specified, the NFS client uses READDIRPLUS requests
637 on NFS version 3 mounts to read small directories.
638 Some applications perform better if the client uses only READDIR requests
640 .SS "Valid options for the nfs4 file system type"
641 Use these options, along with the options in the first subsection above,
647 The transport the NFS client uses
648 to transmit requests to the NFS server for this mount point.
654 All NFS version 4 servers are required to support TCP,
655 so if this mount option is not specified, the NFS version 4 client
656 uses the TCP transport protocol.
657 Refer to the TRANSPORT METHODS section for more details.
660 The numeric value of the server's NFS service port.
661 If the server's NFS service is not available on the specified port,
662 the mount request fails.
664 If this mount option is not specified,
665 the NFS client uses the standard NFS port number of 2049
666 without first checking the server's rpcbind service.
667 This allows an NFS version 4 client to contact an NFS version 4
668 server through a firewall that may block rpcbind requests.
670 If the specified port value is 0,
671 then the NFS client uses the NFS service port number
672 advertised by the server's rpcbind service.
673 The mount request fails if the server's rpcbind service is not available,
674 the server's NFS service is not registered with its rpcbind service,
675 or the server's NFS service is not available on the advertised port.
677 .BR intr " / " nointr
678 Selects whether to allow signals to interrupt file operations
679 on this mount point. If neither option is specified (or if
681 is specified), system calls return EINTR if an in-progress NFS operation
682 is interrupted by a signal. If
684 is specified, signals do not
685 interrupt NFS operations.
689 option is preferred to using the
691 option because it is significantly less likely to result in data corruption.
694 Selects whether to use close-to-open cache coherence semantics
695 for NFS directories on this mount point.
701 the default is to use close-to-open cache coherence
702 semantics for directories.
704 File data caching behavior is not affected by this option.
705 The DATA AND METADATA COHERENCE section discusses
706 the behavior of this option in more detail.
708 .BI clientaddr= n.n.n.n
709 Specifies a single IPv4 address (in dotted-quad form)
710 that the NFS client advertises to allow servers
711 to perform NFS version 4 callback requests against
712 files on this mount point. If the server is unable to
713 establish callback connections to clients, performance
714 may degrade, or accesses to files may temporarily hang.
716 If this option is not specified, the
718 command attempts to discover an appropriate callback address automatically.
719 The automatic discovery process is not perfect, however.
720 In the presence of multiple client network interfaces,
721 special routing policies,
722 or atypical network topologies,
723 the exact address to use for callbacks may be nontrivial to determine.
725 To mount an export using NFS version 2,
728 file system type and specify the
731 To mount using NFS version 3,
734 file system type and specify the
737 To mount using NFS version 4,
743 mount option is not supported for the
747 The following example from an
749 file causes the mount command to negotiate
750 reasonable defaults for NFS behavior.
753 .TA 2.5i +0.7i +0.7i +.7i
754 server:/export /mnt nfs defaults 0 0
757 Here is an example from an /etc/fstab file for an NFS version 2 mount over UDP.
760 .TA 2.5i +0.7i +0.7i +.7i
761 server:/export /mnt nfs nfsvers=2,proto=udp 0 0
764 Try this example to mount using NFS version 4 over TCP
765 with Kerberos 5 mutual authentication.
768 .TA 2.5i +0.7i +0.7i +.7i
769 server:/export /mnt nfs4 sec=krb5 0 0
772 This example can be used to mount /usr over NFS.
775 .TA 2.5i +0.7i +0.7i +.7i
776 server:/export /usr nfs ro,nolock,nocto,actimeo=3600 0 0
778 .SH "TRANSPORT METHODS"
779 NFS clients send requests to NFS servers via
780 Remote Procedure Calls, or
782 The RPC client discovers remote service endpoints automatically,
783 handles per-request authentication,
784 adjusts request parameters for different byte endianness on client and server,
785 and retransmits requests that may have been lost by the network or server.
786 RPC requests and replies flow over a network transport.
790 command, NFS client, and NFS server
791 can automatically negotiate proper transport
792 and data transfer size settings for a mount point.
793 In some cases, however, it pays to specify
794 these settings explicitly using mount options.
796 Traditionally, NFS clients used the UDP transport exclusively for
797 transmitting requests to servers. Though its implementation is
798 simple, NFS over UDP has many limitations that prevent smooth
799 operation and good performance in some common deployment
800 environments. Even an insignificant packet loss rate results in the
801 loss of whole NFS requests; as such, retransmit timeouts are usually
802 in the subsecond range to allow clients to recover quickly from
803 dropped requests, but this can result in extraneous network traffic
806 However, UDP can be quite effective in specialized settings where
807 the network’s MTU is large relative to NFS’s data transfer size (such
808 as network environments that enable jumbo Ethernet frames). In such
809 environments, trimming the
813 settings so that each
814 NFS read or write request fits in just a few network frames (or even
815 in a single frame) is advised. This reduces the probability that
816 the loss of a single MTU-sized network frame results in the loss of
817 an entire large read or write request.
819 TCP is the default transport protocol used for all modern NFS
820 implementations. It performs well in almost every conceivable
821 network environment and provides excellent guarantees against data
822 corruption caused by network unreliability. TCP is often a
823 requirement for mounting a server through a network firewall.
825 Under normal circumstances, networks drop packets much more
826 frequently than NFS servers drop requests. As such, an aggressive
827 retransmit timeout setting for NFS over TCP is unnecessary. Typical
828 timeout settings for NFS over TCP are between one and ten minutes.
829 After the client exhausts its retransmits (the value of the
831 mount option), it assumes a network partition has occurred,
832 and attempts to reconnect to the server on a fresh socket. Since
833 TCP itself makes network data transfer reliable,
837 can safely be allowed to default to the largest values supported by
838 both client and server, independent of the network's MTU size.
839 .SS "Using the mountproto mount option"
840 This section applies only to NFS version 2 and version 3 mounts
841 since NFS version 4 does not use a separate protocol for mount
844 The Linux NFS client can use a different transport for
845 contacting an NFS server's rpcbind service, its mountd service,
846 its Network Lock Manager (NLM) service, and its NFS service.
847 The exact transports employed by the Linux NFS client for
848 each mount point depends on the settings of the transport
849 mount options, which include
852 .BR udp ", and " tcp .
854 The client sends Network Status Manager (NSM) notifications
855 via UDP no matter what transport options are specified, but
856 listens for server NSM notifications on both UDP and TCP.
857 The NFS Access Control List (NFSACL) protocol shares the same
858 transport as the main NFS service.
860 If no transport options are specified, the Linux NFS client
861 uses UDP to contact the server's mountd service, and TCP to
862 contact its NLM and NFS services by default.
864 If the server does not support these transports for these services, the
866 command attempts to discover what the server supports, and then retries
867 the mount request once using the discovered transports.
868 If the server does not advertise any transport supported by the client
869 or is misconfigured, the mount request fails.
872 option is in effect, the mount command backgrounds itself and continues
873 to attempt the specified mount request.
881 option is specified but the
883 option is not, the specified transport is used to contact
884 both the server's mountd service and for the NLM and NFS services.
888 option is specified but none of the
889 .BR proto ", " udp " or " tcp
890 options are specified, then the specified transport is used for the
891 initial mountd request, but the mount command attempts to discover
892 what the server supports for the NFS protocol, preferring TCP if
893 both transports are supported.
896 .BR mountproto " and " proto
899 options are specified, then the transport specified by the
901 option is used for the initial mountd request, and the transport
905 .BR udp " or " tcp " options)"
906 is used for NFS, no matter what order these options appear.
907 No automatic service discovery is performed if these options are
911 .BR proto ", " udp ", " tcp ", "
914 options are specified more than once on the same mount command line,
915 then the value of the rightmost instance of each of these options
917 .SH "DATA AND METADATA COHERENCE"
918 Some modern cluster file systems provide
919 perfect cache coherence among their clients.
920 Perfect cache coherence among disparate NFS clients
921 is expensive to achieve, especially on wide area networks.
922 As such, NFS settles for weaker cache coherence that
923 satisfies the requirements of most file sharing types. Normally,
924 file sharing is completely sequential:
925 first client A opens a file, writes something to it, then closes it;
926 then client B opens the same file, and reads the changes.
928 .SS "Close-to-open cache consistency"
929 When an application opens a file stored on an NFS server,
930 the NFS client checks that it still exists on the server
931 and is permitted to the opener by sending a GETATTR or ACCESS request.
932 When the application closes the file,
933 the NFS client writes back any pending changes
934 to the file so that the next opener can view the changes.
935 This also gives the NFS client an opportunity to report
936 any server write errors to the application
937 via the return code from
939 The behavior of checking at open time and flushing at close time
940 is referred to as close-to-open cache consistency.
941 .SS "Weak cache consistency"
942 There are still opportunities for a client's data cache
943 to contain stale data.
944 The NFS version 3 protocol introduced "weak cache consistency"
945 (also known as WCC) which provides a way of efficiently checking
946 a file's attributes before and after a single request.
947 This allows a client to help identify changes
948 that could have been made by other clients.
950 When a client is using many concurrent operations
951 that update the same file at the same time
952 (for example, during asynchronous write behind),
953 it is still difficult to tell whether it was
954 that client's updates or some other client's updates
955 that altered the file.
956 .SS "Attribute caching"
959 mount option to achieve attribute cache coherence
960 among multiple clients.
961 Almost every file system operation checks
962 file attribute information.
963 The client keeps this information cached
964 for a period of time to reduce network and server load.
967 is in effect, a client's file attribute cache is disabled,
968 so each operation that needs to check a file's attributes
969 is forced to go back to the server.
970 This permits a client to see changes to a file very quickly,
971 at the cost of many extra network operations.
973 Be careful not to confuse the
975 option with "no data caching."
978 mount option prevents the client from caching file metadata,
979 but there are still races that may result in data cache incoherence
980 between client and server.
982 The NFS protocol is not designed to support
983 true cluster file system cache coherence
984 without some type of application serialization.
985 If absolute cache coherence among clients is required,
986 applications should use file locking. Alternatively, applications
987 can also open their files with the O_DIRECT flag
988 to disable data caching entirely.
989 .SS "The sync mount option"
990 The NFS client treats the
992 mount option differently than some other file systems
995 for a description of the generic
1004 is specified (or if the
1006 option is specified),
1007 the NFS client delays sending application
1008 writes to the server
1009 until any of these events occur:
1011 Memory pressure forces reclamation of system memory resources.
1013 An application flushes file data explicitly with
1019 An application closes a file with
1022 The file is locked/unlocked via
1025 In other words, under normal circumstances,
1026 data written by an application may not immediately appear
1027 on the server that hosts the file.
1031 option is specified on a mount point,
1032 any system call that writes data to files on that mount point
1033 causes that data to be flushed to the server
1034 before the system call returns control to user space.
1035 This provides greater data cache coherence among clients,
1036 but at a significant performance cost.
1038 Applications can use the O_SYNC open flag to force application
1039 writes to individual files to go to the server immediately without
1043 .SS "Using file locks with NFS"
1044 The Network Lock Manager protocol is a separate sideband protocol
1045 used to manage file locks in NFS version 2 and version 3.
1046 To support lock recovery after a client or server reboot,
1047 a second sideband protocol --
1048 known as the Network Status Manager protocol --
1051 file locking is supported directly in the main NFS protocol,
1052 and the NLM and NSM sideband protocols are not used.
1054 In most cases, NLM and NSM services are started automatically,
1055 and no extra configuration is required.
1056 Configure all NFS clients with fully-qualified domain names
1057 to ensure that NFS servers can find clients to notify them of server reboots.
1059 NLM supports advisory file locks only.
1060 To lock NFS files, use
1062 with the F_GETLK and F_SETLK commands.
1063 The NFS client converts file locks obtained via
1067 When mounting servers that do not support the NLM protocol,
1068 or when mounting an NFS server through a firewall
1069 that blocks the NLM service port,
1072 mount option. NLM locking must be disabled with the
1074 option when using NFS to mount
1078 contains files used by the NLM implementation on Linux.
1082 option may also be advised to improve the performance
1083 of a proprietary application which runs on a single client
1084 and uses file locks extensively.
1085 .SS "NFS version 4 caching features"
1086 The data and metadata caching behavior of NFS version 4
1087 clients is similar to that of earlier versions.
1088 However, NFS version 4 adds two features that improve
1090 .I change attributes
1092 .IR "file delegation" .
1096 is a new part of NFS file and directory metadata
1097 which tracks data changes.
1098 It replaces the use of a file's modification
1099 and change time stamps
1100 as a way for clients to validate the content
1102 Change attributes are independent of the time stamp
1103 resolution on either the server or client, however.
1107 is a contract between an NFS version 4 client
1108 and server that allows the client to treat a file temporarily
1109 as if no other client is accessing it.
1110 The server promises to notify the client (via a callback request) if another client
1111 attempts to access that file.
1112 Once a file has been delegated to a client, the client can
1113 cache that file's data and metadata aggressively without
1114 contacting the server.
1116 File delegations come in two flavors:
1122 delegation means that the server notifies the client
1123 about any other clients that want to write to the file.
1126 delegation means that the client gets notified about
1127 either read or write accessors.
1129 Servers grant file delegations when a file is opened,
1130 and can recall delegations at any time when another
1131 client wants access to the file that conflicts with
1132 any delegations already granted.
1133 Delegations on directories are not supported.
1135 In order to support delegation callback, the server
1136 checks the network return path to the client during
1137 the client's initial contact with the server.
1138 If contact with the client cannot be established,
1139 the server simply does not grant any delegations to
1141 .SH "SECURITY CONSIDERATIONS"
1142 NFS servers control access to file data,
1143 but they depend on their RPC implementation
1144 to provide authentication of NFS requests.
1145 Traditional NFS access control mimics
1146 the standard mode bit access control provided in local file systems.
1147 Traditional RPC authentication uses a number
1148 to represent each user
1149 (usually the user's own uid),
1150 a number to represent the user's group (the user's gid),
1151 and a set of up to 16 auxiliary group numbers
1152 to represent other groups of which the user may be a member.
1154 Typically, file data and user ID values appear unencrypted
1155 (i.e. "in the clear") on the network.
1156 Moreover, NFS versions 2 and 3 use
1157 separate sideband protocols for mounting,
1158 locking and unlocking files,
1159 and reporting system status of clients and servers.
1160 These auxiliary protocols use no authentication.
1162 In addition to combining these sideband protocols with the main NFS protocol,
1163 NFS version 4 introduces more advanced forms of access control,
1164 authentication, and in-transit data protection.
1165 The NFS version 4 specification mandates NFSv4 ACLs,
1166 RPCGSS authentication, and RPCGSS security flavors
1167 that provide per-RPC integrity checking and encryption.
1168 Because NFS version 4 combines the
1169 function of the sideband protocols into the main NFS protocol,
1170 the new security features apply to all NFS version 4 operations
1171 including mounting, file locking, and so on.
1172 RPCGSS authentication can also be used with NFS versions 2 and 3,
1173 but does not protect their sideband protocols.
1177 mount option specifies the RPCGSS security mode
1178 that is in effect on a given NFS mount point.
1181 provides cryptographic proof of a user's identity in each RPC request.
1182 This provides strong verification of the identity of users
1183 accessing data on the server.
1184 Note that additional configuration besides adding this mount option
1185 is required in order to enable Kerberos security.
1188 man page for details.
1190 Two additional flavors of Kerberos security are supported:
1196 security flavor provides a cryptographically strong guarantee
1197 that the data in each RPC request has not been tampered with.
1200 security flavor encrypts every RPC request
1201 to prevent data exposure during network transit; however,
1202 expect some performance impact
1203 when using integrity checking or encryption.
1204 Similar support for other forms of cryptographic security (such as lipkey and SPKM3)
1207 The NFS version 4 protocol allows
1208 clients and servers to negotiate among multiple security flavors
1209 during mount processing.
1210 However, Linux does not yet implement such negotiation.
1211 The Linux client specifies a single security flavor at mount time
1212 which remains in effect for the lifetime of the mount.
1213 If the server does not support this flavor,
1214 the initial mount request is rejected by the server.
1215 .SS "Mounting through a firewall"
1216 A firewall may reside between an NFS client and server,
1217 or the client or server may block some of its own ports via IP
1219 It is still possible to mount an NFS server through a firewall,
1222 command's automatic service endpoint discovery mechanisms may not work; this
1223 requires you to provide specific endpoint details via NFS mount options.
1225 NFS servers normally run a portmapper or rpcbind daemon to advertise
1226 their service endpoints to clients. Clients use the rpcbind daemon to determine:
1228 What network port each RPC-based service is using
1230 What transport protocols each RPC-based service supports
1232 The rpcbind daemon uses a well-known port number (111) to help clients find a service endpoint.
1233 Although NFS often uses a standard port number (2049),
1234 auxiliary services such as the NLM service can choose
1235 any unused port number at random.
1237 Common firewall configurations block the well-known rpcbind port.
1238 In the absense of an rpcbind service,
1239 the server administrator fixes the port number
1240 of NFS-related services so that the firewall
1241 can allow access to specific NFS service ports.
1242 Client administrators then specify the port number
1243 for the mountd service via the
1248 It may also be necessary to enforce the use of TCP or UDP
1249 if the firewall blocks one of those transports.
1250 .SS "NFS Access Control Lists"
1251 Solaris allows NFS version 3 clients direct access
1252 to POSIX Access Control Lists stored in its local file systems.
1253 This proprietary sideband protocol, known as NFSACL,
1254 provides richer access control than mode bits.
1255 Linux implements this protocol
1256 for compatibility with the Solaris NFS implementation.
1257 The NFSACL protocol never became a standard part
1258 of the NFS version 3 specification, however.
1260 The NFS version 4 specification mandates a new version
1261 of Access Control Lists that are semantically richer than POSIX ACLs.
1262 NFS version 4 ACLs are not fully compatible with POSIX ACLs; as such,
1263 some translation between the two is required
1264 in an environment that mixes POSIX ACLs and NFS version 4.
1272 option is not fully supported.
1273 Generic options, such as
1275 can be modified using the
1278 but NFS-specific options are not all supported.
1279 The underlying transport or NFS version
1280 cannot be changed by a remount, for example.
1281 Performing a remount on an NFS file system mounted with the
1283 option may have unintended consequences.
1286 option is a mixture of a generic option,
1288 and an NFS-specific option
1291 Before 2.4.7, the Linux NFS client did not support NFS over TCP.
1293 Before 2.4.20, the Linux NFS client used a heuristic
1294 to determine whether cached file data was still valid
1295 rather than using the standard close-to-open cache coherency method
1298 Starting with 2.4.22, the Linux NFS client employs
1299 a Van Jacobsen-based RTT estimator to determine retransmit
1300 timeout values when using NFS over UDP.
1302 Before 2.6.0, the Linux NFS client did not support NFS version 4.
1304 Before 2.6.8, the Linux NFS client used only synchronous reads and writes
1306 .BR rsize " and " wsize
1307 settings were smaller than the system's page size.
1309 The Linux NFS client does not yet support
1310 certain optional features of the NFS version 4 protocol,
1311 such as security negotiation, server referrals, and named attributes.
1324 .BR rpc.svcgssd (8),
1327 RFC 768 for the UDP specification.
1329 RFC 793 for the TCP specification.
1331 RFC 1094 for the NFS version 2 specification.
1333 RFC 1813 for the NFS version 3 specification.
1335 RFC 1832 for the XDR specification.
1337 RFC 1833 for the RPC bind specification.
1339 RFC 2203 for the RPCSEC GSS API protocol specification.
1341 RFC 3530 for the NFS version 4 specification.