mount(1M-nfs) mount(1M-nfs)
NAME
mount - mount remote NFS resources
SYNOPSIS
mount [-F nfs] [-r] [-o specifoptions] [resource mountpoint]
DESCRIPTION
The mount command is used to mount a specified resource in the file
system hierarchy at the mountpoint pathname position. This position
must already exist. If mountpoint already contains data before the
mount operation, this is suppressed until the resource is unmounted
again.
If the resource is referenced in the vfstab(4) file, either resource
or mountpoint can be specified in the command line and mount will
search vfstab for further specifications. If the -F option is not
specified, mount uses the file system type specified in vfstab. mount
maintains a table of mounted file systems in /etc/mnttab, as described
in mnttab(4).
OPTIONS
-r Mounts the specified resource with read authorization.
-o specifoptions
Specifies the file system options in a list of words separated by
commas. These words are listed below:
acdirmax=n
Holds attributes in the cache buffer for up to n seconds
after a directory has been updated. The default value is 60
seconds.
acdirmin=n
Holds attributes in the cache buffer for at least n seconds
after a directory has been updated. The default value is 30
seconds.
acregmax=n
Holds attributes in the cache buffer for up to n seconds
after the file has been changed. The default value is 60
seconds.
acregmin=n
Holds attributes in the cache buffer for at least n seconds
after the file has been changed. The default value is 3
seconds.
actimeo=n
Extends the flushing interval for file attributes in the
cache for regular files and directories by n seconds (see
also the section on file attributes).
Page 1 Reliant UNIX 5.44 Printed 11/98
mount(1M-nfs) mount(1M-nfs)
bg | fg
Starts a new attempt in the foreground or background if the
first attempt fails. The default is fg.
devacc
Allows access to special files. If this option is not speci-
fied, no special files can be opened. This can be desirable
under certain circumstances, since the special file is
opened on the client rather than the server via NFS.
fewgrps
A maximum of 8 group IDs instead of 16 is entered in the RPC
credentials of the NFS client. This option is needed in the
case of servers where more than 8 group IDs causes an
authentication error. A file system on such a server cannot
be mounted without this option.
grpid
Files are created using the group ID (GID) of the effective
GID of the calling process. This setting can be deactivated
for each directory by setting the s-bit for the group of the
parent directory; in this case, the group ID is the same as
the ID of the parent directory [see open(2) and mkdir(2)].
Files created in file systems which are not mounted with the
grpid grpid option are subject to BSD semantics, i.e. the
GID must be accepted by the GID of the parent directory.
intr Enables a hung process to be killed using keyboard inter-
rupts while it is waiting for a response on a hard mounted
file system.
locallocks
Locks on files in the mounted NFS file system are only set
locally on the client (and not on the server). This is use-
ful if the mounting client is the only client using the
mounted NFS, for example a client which has no hard disk.
noac Suppresses buffering of attributes in the cache buffer.
port=n
The server IP port number. The default value is NFSPORT
(2049).
remount
If a file system is mounted with read authorization only, it
is remounted with read/write authorization.
retrans=n
Sets the number of NFS retransmissions to n. The default
value is 5.
Page 2 Reliant UNIX 5.44 Printed 11/98
mount(1M-nfs) mount(1M-nfs)
retry=n
Number of retries for the mount operation. The default is
10000.
rsize=n
Defines the size of the read buffer at n bytes. The default
value is 8 KB.
rw | ro
resource is mounted with write or read authorization. The
default is rw.
soft | hard
Specifies whether the mount procedure is to be soft or hard.
An NFS job for a file system mounted with the hard option is
repeated until the server responds. A job for a file system
mounted with the soft option is aborted if the server does
not respond after the timelimit elapses (timeo option) or
after an adjustable number of retransmissions (retrans
option).
suid | nosuid
Specifies whether set uid bits are regarded for executable
files. Setuid bits are taken into account during execution
if suid is specified and ignored if nosuid is specified. (If
you do not specify any option, s-bits are taken into account
by default.) It is recommended that nosuid is used with
directories mounted with rw.
timeo=n
Specifies a value for the maximum time (in tenths of a
second) that a client should wait for an NFS request to be
executed. For NFS communication over a longer period of
time, this value is constantly recalculated by the client on
the basis of the server's response time behavior. The
default setting is 1.1 seconds.
uforce
A file system mounted using the uforce can be unmounted
quickly by specifying umount, even if the server cannot be
accessed.
v2 This command mounts server resources using the version 2 NFS
protocol, even if the servers support version 3.
wsize=n
Defines the size of the write buffer at n bytes. The default
value is 8 KB.
Page 3 Reliant UNIX 5.44 Printed 11/98
mount(1M-nfs) mount(1M-nfs)
NFS FILE SYSTEMS
Background compared to foreground
If file systems are mounted using the bg option, this means that mount
is to repeat the mount operation in the background if the mount daemon
of the server [see mountd(1M)] does not respond. mount repeats the
query as often as is specified in the retry=n option. As soon as the
file system is mounted, NFS jobs are executed and the system waits for
a response. If no response is received, the wait time is multiplied by
2 and the request is transmitted again. If the number of retries has
reached the amount specified in the retrans=n option or if the timeout
is exceeded, a file system which is mounted with the soft option
returns an error for the job; if the file system was mounted with the
hard option, a warning is output and the request is retried.
Read/write access compared to read access
You should use the hard option for file systems mounted with rw
(read/write access).
Program execution with NFS
If programs are to be executed on file systems mounted with NFS, the
file system must not have been mounted with the soft or intr option.
File attributes
File attributes for the client are temporarily stored in the cache
buffer. File attributes are deleted after a certain period. If a file
is changed before the attribute buffer is emptied, the emptying inter-
val is extended by the time since the last change; it is assumed that
recently changed files will soon be changed again. There are minimum
and maximum values for extending the emptying intervals for normal
files and directories. actimeo=n extends the emptying interval for
regular files and directories by n seconds.
EXAMPLES
Mounting a remote file system:
mount -o hard serv:/usr/src /usr/src
Permanently mounting a remote file system:
mount -o hard serv:/usr/src /usr/src
NOTES
If the directory in which a file system is to be mounted is a symbolic
link, the file system is mounted in the directory to which the sym-
bolic link refers and not in the link itself.
Page 4 Reliant UNIX 5.44 Printed 11/98
mount(1M-nfs) mount(1M-nfs)
All client NFS file systems can be made interruptible by setting the
nfsintr kernel variable to the value 1, for example with crash(1M).
This is advantageous if processes are left hanging permanently when
accessing mounted NFS file systems which are uninterruptible, because
for example the server cannot be accessed.
FILES
/etc/mnttab
Table of mounted file systems
/etc/vfstab
Table of automatically mounted resources
SEE ALSO
mountall(1M), mount(2), umount(2), fstypes(4), mnttab(4), vfstab(4).
Page 5 Reliant UNIX 5.44 Printed 11/98