1.\" Copyright (c) 1992, 1993, 1994, 1995 2.\" The Regents of the University of California. All rights reserved. 3.\" 4.\" Redistribution and use in source and binary forms, with or without 5.\" modification, are permitted provided that the following conditions 6.\" are met: 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. 12.\" 3. Neither the name of the University nor the names of its contributors 13.\" may be used to endorse or promote products derived from this software 14.\" without specific prior written permission. 15.\" 16.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 17.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 18.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 19.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 20.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 21.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 22.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 23.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 26.\" SUCH DAMAGE. 27.\" 28.Dd November 29, 2024 29.Dt MOUNT_NFS 8 30.Os 31.Sh NAME 32.Nm mount_nfs 33.Nd mount NFS file systems 34.Sh SYNOPSIS 35.Nm 36.Op Fl 23bcdiLlNPsTU 37.Op Fl a Ar maxreadahead 38.Op Fl D Ar deadthresh 39.Op Fl g Ar maxgroups 40.Op Fl I Ar readdirsize 41.Op Fl o Ar options 42.Op Fl R Ar retrycnt 43.Op Fl r Ar readsize 44.Op Fl t Ar timeout 45.Op Fl w Ar writesize 46.Op Fl x Ar retrans 47.Ar rhost : Ns Ar path node 48.Sh DESCRIPTION 49The 50.Nm 51utility calls the 52.Xr nmount 2 53system call to prepare and graft a remote NFS file system 54.Pq Ar rhost : Ns Ar path 55on to the file system tree at the point 56.Ar node . 57This command is normally executed by 58.Xr mount 8 . 59For NFSv2 and NFSv3, 60it implements the mount protocol as described in RFC 1094, Appendix A and 61RFC 1813, Appendix I. 62For NFSv4, it uses the NFSv4 protocol as described in RFC 7530, RFC 5661 and 63RFC 7862. 64.Pp 65By default, 66.Nm 67keeps retrying until the mount succeeds. 68This behaviour is intended for file systems listed in 69.Xr fstab 5 70that are critical to the boot process. 71For non-critical file systems, the 72.Cm bg 73and 74.Cm retrycnt 75options provide mechanisms to prevent the boot process from hanging 76if the server is unavailable. 77.Pp 78If the server becomes unresponsive while an NFS file system is 79mounted, any new or outstanding file operations on that file system 80will hang uninterruptibly until the server comes back. 81To modify this default behaviour, see the 82.Cm intr 83and 84.Cm soft 85options. 86.Pp 87The options are: 88.Bl -tag -width indent 89.It Fl o 90Options are specified with a 91.Fl o 92flag followed by a comma separated string of options. 93See the 94.Xr mount 8 95man page for possible options and their meanings. 96The following NFS specific options are also available: 97.Bl -tag -width indent 98.It Cm acregmin Ns = Ns Aq Ar seconds 99.It Cm acregmax Ns = Ns Aq Ar seconds 100.It Cm acdirmin Ns = Ns Aq Ar seconds 101.It Cm acdirmax Ns = Ns Aq Ar seconds 102When attributes of files are cached, a timeout calculated to determine 103whether a given cache entry has expired. 104These four values determine the upper and lower bounds of the timeouts for 105.Dq directory 106attributes and 107.Dq regular 108(ie: everything else). 109The default values are 3 -> 60 seconds 110for regular files, and 30 -> 60 seconds for directories. 111The algorithm to calculate the timeout is based on the age of the file. 112The older the file, 113the longer the cache is considered valid, subject to the limits above. 114.It Cm actimeo Ns = Ns Aq Ar seconds 115Set four cache timeouts above to specified value. 116.It Cm allgssname 117This option can be used along with 118.Fl o Cm gssname 119to specify that all operations should use the host-based initiator 120credential. 121This may be used for clients that run system daemons that need to 122access files on the NFSv4 mounted volume. 123.It Cm bg 124If an initial attempt to contact the server fails, fork off a child to keep 125trying the mount in the background. 126Useful for 127.Xr fstab 5 , 128where the file system mount is not critical to multiuser operation. 129.It Cm bgnow 130Like 131.Cm bg , 132fork off a child to keep trying the mount in the background, 133but do not attempt to mount in the foreground first. 134This eliminates a 13560+ second timeout when the server is not responding. 136Useful for speeding up the boot process of a client when the server is 137likely to be unavailable. 138This is often the case for interdependent servers 139such as cross-mounted servers (each of two servers is an NFS client of 140the other) and for cluster nodes that must boot before the file servers. 141.It Cm deadthresh Ns = Ns Aq Ar value 142Set the 143.Dq "dead server threshold" 144to the specified number of round trip timeout intervals before a 145.Dq "server not responding" 146message is displayed. 147.It Cm dumbtimer 148Turn off the dynamic retransmit timeout estimator. 149This may be useful for UDP mounts that exhibit high retry rates, 150since it is possible that the dynamically estimated timeout interval is too 151short. 152.It Cm fg 153Same as not specifying 154.Cm bg . 155.It Cm gssname Ns = Ns Aq Ar service-principal-name 156This option can be used with the KerberosV security flavors for NFSv4 mounts 157to specify the 158.Dq "service-principal-name" 159of a host-based entry in the default 160keytab file that is used for system operations. 161It allows the mount to be performed by 162.Dq "root" 163and avoids problems with 164cached credentials for the system operations expiring. 165The 166.Dq "service-principal-name" 167should be specified without instance or domain and is typically 168.Dq "host" , 169.Dq "nfs" 170or 171.Dq "root" , 172although the form 173.Sm off 174.Aq Ar service 175@ 176.Aq Ar fqdn 177.Sm on 178can also be used if the local system's 179.Xr gethostname 3 180value does not match the host-based principal in the keytab. 181.It Cm hard 182Same as not specifying 183.Cm soft . 184.It Cm intr 185Make the mount interruptible, which implies that file system calls that 186are delayed due to an unresponsive server will fail with EINTR when a 187termination signal is posted for the process. 188To avoid leaving file locks in an indeterminate state on the NFS 189server, it is recommended that the 190.Cm nolockd 191option be used with this option. 192.It Cm maxgroups Ns = Ns Aq Ar value 193Set the maximum size of the group list for the credentials to the 194specified value. 195This should be used for mounts on old servers that cannot handle a 196group list size of 16, as specified in RFC 1057. 197Try 8, if users in a lot of groups cannot get response from the mount 198point. 199.It Cm mountport Ns = Ns Aq Ar value 200Specify the port number to be used to communicate with 201.Xr mountd 8 202on the NFS server. 203This option allows an NFSv2 or NFSv3 mount to be done without 204the need to run the 205.Xr rpcbind 8 206service. 207This option is meaningless for an NFSv4 mount, since NFSv4 208does not use the Mount protocol. 209.It Cm mntudp 210Force the mount protocol to use UDP transport, even for TCP NFS mounts. 211(Necessary for some old 212.Bx 213servers.) 214.It Cm nametimeo Ns = Ns Aq Ar value 215Override the default of NFS_DEFAULT_NAMETIMEO for the timeout (in seconds) 216for positive name cache entries. 217If this is set to 0 it disables positive name caching for the mount point. 218.It Cm negnametimeo Ns = Ns Aq Ar value 219Override the default of NFS_DEFAULT_NEGNAMETIMEO for the timeout (in seconds) 220for negative name cache entries. 221If this is set to 0 it disables negative name caching for the mount point. 222.It Cm nconnect Ns = Ns Aq Ar value 223Specify the number of TCP connections (1-16) to be used 224for an NFS Version 4, minor version 1 or 2 mount. 225Multiple TCP connections can provide more client to server network 226bandwidth for certain network configurations such as: 227.Bd -literal 228- Multiple network interfaces that are aggregated together. 229- A fast network interface that uses multiple queues. 230.Ed 231.sp 232The first TCP connection will be used for all RPCs that consist 233entirely of small RPC messages. 234The RPCs that can have large RPC messages (Read/Readdir/Write) are 235distributed over the additional TCP connections in a round robin 236fashion. 237This option will result in more IP port#s being used. 238This option requires the 239.Cm nfsv4 240option. 241Note that for NFS servers such as AmazonEFS, where each new TCP 242connection can connect to a different cluster that maintains lock 243state separately, this option cannot be used. 244.It Cm nfsv2 245Use the NFS Version 2 protocol (the default is to try version 3 first 246then version 2). 247Note that NFS version 2 has a file size limit of 2 gigabytes. 248.It Cm nfsv3 249Use the NFS Version 3 protocol. 250.It Cm nfsv4 251Use the NFS Version 4 protocol. 252This option will force the mount to use 253TCP transport. 254By default, the highest minor version of NFS Version 4 that is 255supported by the NFS Version 4 server will be used. 256See the 257.Cm minorversion 258option. 259Make sure that all your NFS Version 4 clients have unique 260values in 261.Pa /etc/hostid . 262.It Cm minorversion Ns = Ns Aq Ar value 263Use the specified minor version for a NFS Version 4 mount, 264overriding the default. 265The minor versions supported are 0, 1, and 2. 266This option is only meaningful when used with the 267.Cm nfsv4 268option. 269.It Cm oneopenown 270Make a minor version 1 or 2 of the NFS Version 4 protocol mount use a single 271OpenOwner for all Opens. 272This may be useful for a server with a very low limit on OpenOwners, such as 273AmazonEFS. 274It may be required when an accumulation of NFS version 4 Opens occurs, 275as indicated by the 276.Dq Opens 277count displayed by 278.Xr nfsstat 1 279with the 280.Fl c 281and 282.Fl E 283command-line options. 284A common case for an accumulation of Opens is a shared library within 285the NFS mount that is used by several 286processes, where at least one of these processes is always running. 287This option cannot be used for an NFS Version 4, minor version 0 mount. 288It may not work correctly when Delegations are being issued by a server, 289but note that the AmazonEFS server does not issued delegations at this time. 290This option is only meaningful when used with the 291.Cm nfsv4 292option. 293.It Cm pnfs 294Enable support for parallel NFS (pNFS) for minor version 1 or 2 of the 295NFS Version 4 protocol. 296This option is only meaningful when used with the 297.Cm nfsv4 298option. 299.It Cm noac 300Disable attribute caching. 301.It Cm noconn 302For UDP mount points, do not do a 303.Xr connect 2 . 304This must be used if the server does not reply to requests from the standard 305NFS port number 2049 or replies to requests using a different IP address 306(which can occur if the server is multi-homed). 307Setting the 308.Va vfs.nfs.nfs_ip_paranoia 309sysctl to 0 will make this option the default. 310.It Cm nocto 311Normally, NFS clients maintain the close-to-open cache coherency. 312This works by flushing at close time and checking at open time. 313Checking at open time is implemented by getting attributes from 314the server and purging the data cache if they do not match 315attributes cached by the client. 316.Pp 317This option disables checking at open time. 318It may improve performance for read-only mounts, 319but should only be used if the data on the server changes rarely. 320Be sure to understand the consequences before enabling this option. 321.It Cm noinet4 , noinet6 322Disables 323.Dv AF_INET 324or 325.Dv AF_INET6 326connections. 327Useful for hosts that have 328both an A record and an AAAA record for the same name. 329.It Cm nolockd 330Do 331.Em not 332forward 333.Xr fcntl 2 334locks over the wire via the NLM protocol for NFSv3 mounts 335or via the NFSv4 protocol for NFSv4 mounts. 336All locks will be local and not seen by the server 337and likewise not seen by other NFS clients for NFSv3 or NFSv4 mounts. 338This removes the need to run the 339.Xr rpcbind 8 340service and the 341.Xr rpc.statd 8 342and 343.Xr rpc.lockd 8 344servers on the client for NFSv3 mounts. 345Note that this option will only be honored when performing the 346initial mount, it will be silently ignored if used while updating 347the mount options. 348Also, note that NFSv4 mounts do not use these daemons. 349The NFSv4 protocol handles locks, 350unless this option is specified. 351.It Cm noncontigwr 352This mount option allows the NFS client to 353combine non-contiguous byte ranges being written 354such that the dirty byte range becomes a superset of the bytes 355that are dirty. 356This reduces the number of writes significantly for software 357builds. 358The merging of byte ranges is not done if the file has been file 359locked, since most applications modifying a file from multiple 360clients will use file locking. 361As such, this option could result in a corrupted file for the 362rare case of an application modifying the file from multiple 363clients concurrently without using file locking. 364.It Cm principal 365For the RPCSEC_GSS security flavors, such as krb5, krb5i and krb5p, 366this option sets the name of the host based principal name expected 367by the server. 368This option overrides the default, which will be ``nfs@<server-fqdn>'' 369and should normally be sufficient. 370.It Cm noresvport 371Do 372.Em not 373use a reserved socket port number (see below). 374.It Cm port Ns = Ns Aq Ar port_number 375Use specified port number for NFS requests. 376The default is to query the portmapper for the NFS port. 377.It Cm proto Ns = Ns Aq Ar protocol 378Specify transport protocol version to use. 379Currently, they are: 380.Bd -literal 381udp - Use UDP over IPv4 382tcp - Use TCP over IPv4 383udp6 - Use UDP over IPv6 384tcp6 - Use TCP over IPv6 385.Ed 386.It Cm rdirplus 387Used with NFSV3 to specify that the \fBReaddirPlus\fR RPC should 388be used. 389For NFSV4, setting this option has a similar effect, in that it will make 390the Readdir Operation get more attributes. 391This option reduces RPC traffic for cases such as 392.Dq "ls -l" , 393but tends to flood the attribute and name caches with prefetched entries. 394Try this option and see whether performance improves or degrades. 395Probably 396most useful for client to server network interconnects with a large bandwidth 397times delay product. 398.It Cm readahead Ns = Ns Aq Ar value 399Set the read-ahead count to the specified value. 400This may be in the range of 0 - 4, and determines how many blocks 401will be read ahead when a large file is being read sequentially. 402Trying a value greater than 1 for this is suggested for 403mounts with a large bandwidth * delay product. 404.It Cm readdirsize Ns = Ns Aq Ar value 405Set the readdir read size to the specified value. 406The value should normally 407be a multiple of 408.Dv DIRBLKSIZ 409that is <= the read size for the mount. 410.It Cm resvport 411Use a reserved socket port number. 412This flag is obsolete, and only retained for compatibility reasons. 413Reserved port numbers are used by default now. 414(For the rare case where the client has a trusted root account 415but untrustworthy users and the network cables are in secure areas this does 416help, but for normal desktop clients this does not apply.) 417.It Cm retrans Ns = Ns Aq Ar value 418Set the retransmit timeout count for soft mounts to the specified value. 419.It Cm retrycnt Ns = Ns Aq Ar count 420Set the mount retry count to the specified value. 421The default is a retry count of zero, which means to keep retrying 422forever. 423There is a 60 second delay between each attempt. 424.It Cm rsize Ns = Ns Aq Ar value 425Set the read data size to the specified value. 426It should normally be a power of 2 greater than or equal to 1024. 427This should be used for UDP mounts when the 428.Dq "fragments dropped due to timeout" 429value is getting large while actively using a mount point. 430(Use 431.Xr netstat 1 432with the 433.Fl s 434option to see what the 435.Dq "fragments dropped due to timeout" 436value is.) 437.It Cm sec Ns = Ns Aq Ar flavor 438This option specifies what security flavor should be used for the mount. 439Currently, they are: 440.Bd -literal 441krb5 - Use KerberosV authentication 442krb5i - Use KerberosV authentication and 443 apply integrity checksums to RPCs 444krb5p - Use KerberosV authentication and 445 encrypt the RPC data 446sys - The default AUTH_SYS, which uses a 447 uid + gid list authenticator 448.Ed 449.It Cm soft 450A soft mount, which implies that file system calls will fail 451after 452.Ar retrycnt 453round trip timeout intervals. 454.It Cm syskrb5 455This option specifies that a KerberosV NFSv4 minor version 1 or 2 mount 456uses AUTH_SYS for system operations. 457Using this option avoids the need for a KerberosV mount to have a 458host-based principal entry in the default keytab file 459(no 460.Cm gssname 461option) or a requirement for the user doing the mount to have a 462valid KerberosV ticket granting ticket (TGT) when the mount is done. 463This option is intended to be used with the 464.Cm sec Ns = Ns krb5 465and 466.Cm tls 467options and can only be used for 468NFSv4 mounts with minor version 1 or 2. 469.It Cm tcp 470Use TCP transport. 471This is the default option, as it provides for increased reliability on both 472LAN and WAN configurations compared to UDP. 473Some old NFS servers do not support this method; UDP mounts may be required 474for interoperability. 475.It Cm timeout Ns = Ns Aq Ar value 476Set the initial retransmit timeout to the specified value, 477expressed in tenths of a second. 478May be useful for fine tuning UDP mounts over internetworks 479with high packet loss rates or an overloaded server. 480Try increasing the interval if 481.Xr nfsstat 1 482shows high retransmit rates while the file system is active or reducing the 483value if there is a low retransmit rate but long response delay observed. 484(Normally, the 485.Cm dumbtimer 486option should be specified when using this option to manually 487tune the timeout 488interval.) 489.It Cm timeo Ns = Ns Aq Ar value 490Alias for 491.Cm timeout . 492.It Cm tls 493This option specifies that the connection to the server must use TLS 494per RFC 9289. 495TLS is only supported for TCP connections and the 496.Xr rpc.tlsclntd 8 497daemon must be running for an NFS over TCP connection to use TLS. 498.It Cm tlscertname Ns = Ns Aq Ar name 499This option specifies the name of an alternate certificate to be 500presented to the NFS server during TLS handshake. 501The default certificate file names are 502.Dq cert.pem 503and 504.Dq certkey.pem . 505When this option is specified, 506.Ar name 507replaces 508.Dq cert 509in the above file names. 510For example, if the value of 511.Ar name 512is specified as 513.Dq other 514the certificate file names to be used will be 515.Dq other.pem 516and 517.Dq otherkey.pem . 518These files are stored in 519.Pa /etc/rpc.tlsclntd 520by default. 521This option is only meaningful when used with the 522.Cm tls 523option and the 524.Xr rpc.tlsclntd 8 525is running with the 526.Fl m 527command line flag set. 528.It Cm udp 529Use UDP transport. 530.It Cm vers Ns = Ns Aq Ar vers_number 531Use the specified version number for NFS requests. 532See the 533.Cm nfsv2 , 534.Cm nfsv3 , 535and 536.Cm nfsv4 537options for details. 538.It Cm wcommitsize Ns = Ns Aq Ar value 539Set the maximum pending write commit size to the specified value. 540This determines the maximum amount of pending write data that the NFS 541client is willing to cache for each file. 542.It Cm wsize Ns = Ns Aq Ar value 543Set the write data size to the specified value. 544Ditto the comments w.r.t.\& the 545.Cm rsize 546option, but using the 547.Dq "fragments dropped due to timeout" 548value on the server instead of the client. 549Note that both the 550.Cm rsize 551and 552.Cm wsize 553options should only be used as a last ditch effort at improving performance 554when mounting servers that do not support TCP mounts. 555.El 556.El 557.Sh IMPLEMENTATION NOTES 558When neither the 559.Cm rsize 560nor 561.Cm wsize 562options are specified, the I/O size will be set to the largest value 563supported by both the NFS client and server. 564The largest value supported by the NFS client is defined by 565the tunable 566.Cd vfs.maxbcachebuf 567which can be set to a power of two up to 568.Cd kern.maxphys . 569.Pp 570The 571.Xr nfsstat 1 572command with the 573.Ic -m 574command line option will show what 575.Nm 576option settings are actually in use for the mount. 577.Sh COMPATIBILITY 578The following command line flags are equivalent to 579.Fl o 580named options and are supported for compatibility with older 581installations. 582.Bl -tag -width indent 583.It Fl 2 584Same as 585.Fl o Cm nfsv2 586.It Fl 3 587Same as 588.Fl o Cm nfsv3 589.It Fl D 590Same as 591.Fl o Cm deadthresh 592.It Fl I 593Same as 594.Fl o Cm readdirsize Ns = Ns Aq Ar value 595.It Fl L 596Same as 597.Fl o Cm nolockd 598.It Fl N 599Same as 600.Fl o Cm noresvport 601.It Fl P 602Use a reserved socket port number. 603This flag is obsolete, and only retained for compatibility reasons. 604(For the rare case where the client has a trusted root account 605but untrustworthy users and the network cables are in secure areas this does 606help, but for normal desktop clients this does not apply.) 607.It Fl R 608Same as 609.Fl o Cm retrycnt Ns = Ns Aq Ar value 610.It Fl T 611Same as 612.Fl o Cm tcp 613.It Fl U 614Same as 615.Fl o Cm mntudp 616.It Fl a 617Same as 618.Fl o Cm readahead Ns = Ns Aq Ar value 619.It Fl b 620Same as 621.Fl o Cm bg 622.It Fl c 623Same as 624.Fl o Cm noconn 625.It Fl d 626Same as 627.Fl o Cm dumbtimer 628.It Fl g 629Same as 630.Fl o Cm maxgroups 631.It Fl i 632Same as 633.Fl o Cm intr 634.It Fl l 635Same as 636.Fl o Cm rdirplus 637.It Fl r 638Same as 639.Fl o Cm rsize Ns = Ns Aq Ar value 640.It Fl s 641Same as 642.Fl o Cm soft 643.It Fl t 644Same as 645.Fl o Cm retransmit Ns = Ns Aq Ar value 646(deprecated) 647.It Fl w 648Same as 649.Fl o Cm wsize Ns = Ns Aq Ar value 650.It Fl x 651Same as 652.Fl o Cm retrans Ns = Ns Aq Ar value 653.El 654.Pp 655The following 656.Fl o 657named options are equivalent to other 658.Fl o 659named options and are supported for compatibility with other 660operating systems (e.g., Linux, Solaris, and OSX) to ease usage of 661.Xr autofs 4 662support. 663.Bl -tag -width indent 664.It Fl o Cm vers Ns = Ns 2 665Same as 666.Fl o Cm nfsv2 667.It Fl o Cm vers Ns = Ns 3 668Same as 669.Fl o Cm nfsv3 670.It Fl o Cm vers Ns = Ns 4 671Same as 672.Fl o Cm nfsv4 673.El 674.Sh SEE ALSO 675.Xr nfsstat 1 , 676.Xr nmount 2 , 677.Xr unmount 2 , 678.Xr lagg 4 , 679.Xr nfsv4 4 , 680.Xr fstab 5 , 681.Xr gssd 8 , 682.Xr mount 8 , 683.Xr nfsd 8 , 684.Xr nfsiod 8 , 685.Xr rpcbind 8 , 686.Xr rpc.tlsclntd 8 , 687.Xr showmount 8 688.Sh HISTORY 689A version of the 690.Nm 691utility appeared in 692.Bx 4.4 . 693.Sh BUGS 694Since NFSv4 performs open/lock operations that have their ordering strictly 695enforced by the server, the options 696.Cm intr 697and 698.Cm soft 699cannot be safely used. 700For NFSv4 minor version 1 or 2 mounts, the ordering is done 701via session slots and the NFSv4 client now handles broken session slots 702fairly well. 703As such, if the 704.Cm nolockd 705option is used along with 706.Cm intr 707and/or 708.Cm soft , 709an NFSv4 minor version 1 or 2 mount 710should work fairly well, although still not completely correctly. 711For NFSv4 minor version 0 mounts, 712.Cm hard 713mounts without the 714.Cm intr 715mount option is strongly recommended. 716