You might see the following types of NFS error messages:
The following error messages are issued
to the screen or console or sent to the
syslogd
daemon.
authget: unknown authflavor
nauthflavorExplanation: Each NFS request has an authentication type. This message is displayed if the type is not AUTH_UNIX.
User Action: Have the client application use the AUTH_UNIX authentication type.
fh3tovp: bad length:
nExplanation: A client sent a bad file handle to the server.
NFS request from unprivileged port, source IP address =
nExplanation: The server, performing NFS server port monitoring, received an NFS request from a nonprivileged port (greater than or equal to 1024) on a client. This might indicate a security problem.
NFS server: fs(n,n
) not mounted; client address =
n.n.n.nExplanation: The client requested a file on a file system that is not mounted or does not exist on the server. This can occur if a file system is unmounted while clients are using it or if the client passed an invalid file handle.
User Action: Make sure that the appropriate file system is mounted on the NFS server. If the file system is mounted on the same device, have the client system retry the operation. If the file system is mounted on a different device, have the client system unmount and remount the remote file system.
NFS server: stale file handle fs(n,n
) file
file
gen
n,client address =
n.n.n.n
errno
nExplanation: The client accessed a file that no longer exists. The file was deleted either by the server or by another client.
NFS server: unexported fs(n,n
) file
file,
client address
=
n.n.n.nExplanation:
A client that previously had access
to a file system can no longer access the file system, either because of changes
in the
/etc/exports
file or in the net group mapping.
User Action: Have the client system unmount the file system.
rfs_dispatch botchExplanation: The duplicate request cache routine returned an illegal value.
rfs_dispatch: bad rfs reply
nretExplanation: A server routine did not return a value or returned an incorrect value.
rfs_dispatch: dispatch error, no replyrfs_dispatch: sendreply failedExplanation: Possible reasons for this message include the following:
The server is out of memory and cannot process or reply to a request.
The server cannot find a route to the source.
There is some other network-related problem.
too many nfsdsExplanation:
There are more
nfsd
daemons registered
with NFS than were started.
This appendix provides an explanation and suggested user actions for the following classes of client error messages:
Within each section, error messages are listed alphabetically.
C.2.1 Remote Mount Error Messages
The following error messages are displayed if you are mounting directories or file systems from remote systems:
mount: unknown special file or file system
xxxExplanation:
There is no entry in the
/etc/fstab
file for
the mount point that you specified in the
mount
command
line.
User Action:
Verify that there is
an entry in the
/etc/fstab
file for the file system.
If
not, add an entry.
If one exists, look for syntax errors or typos in the entry.
See
fstab(4)
/etc/fstab: No such file or directoryExplanation:
The
/etc/fstab
file does not exist.
The
mount
command discovered
this when it tried to look up the name specified on the command line.
User Action:
Create an
/etc/fstab
file and include the appropriate entries.
See
fstab(4)
nfs_mount: Permission denied for
yyyExplanation: Your host name is not in the export list for the file system or directory you want to mount from the server.
User Action:
Get a list of your host's exported file systems and directories
by using the
showmount -e
command.
For example,
enter the following command if your server's host name is host2:
# /usr/bin/showmount -e host2
If the file system or directory you want to mount remotely
is not on the list, or if your host or network group name is not on the user
list for the file system or directory, log in to the server and look in the
/etc/exports
file for the correct file system entry.
If the file system or directory name is in the
/etc/exports
file, but not in the output from
showmount,
the failure is in the
mountd
daemon.
The
mountd
daemon could not parse that line in the file, could not find the
file system or directory, or the file system or directory name was not a locally
mounted file system.
If the file system or directory name is in the
/etc/exports
file and Network Information Service (NIS) is configured, verify
that the
ypbind
daemon is running; it might have stopped.
See
exports(4)
nfs_mount: cannot mount
xxx
on
yyy:
Mount device busyExplanation: The file system or directory you are trying to mount is already mounted.
nfs_mount: cannot mount
xxx
on
yyy:
No such file or directoryExplanation: The local mount point does not exist.
User Action: Verify that the mount point exists and that it is spelled correctly.
nfs_mount: cannot mount
xxx
on
file:
Not a directoryExplanation: Either the remote file system or the local mount point is not a directory.
User Action:
Verify that the remote
file system and the local mount point are directories (not files) by using
the
ls
command.
Verify the spelling of both directories.
nfs_mount: cannot mount
xxx
on
yyy:
Not ownerExplanation: You must mount the remote file system or directory as superuser (root) on your system.
nfs_mount: illegal file system name
xxx;
use host:pathnameExplanation:
You did not specify the name of the server when you
issued the
mount
command.
User Action:
For example, to mount
the file system
/usr/src
from the server host2, enter the
following command:
# mount host2:/usr/src /host2/usr/src
nfs_mount: invalid directory name
xxxdirectory pathname must begin with '/'.Explanation:
The mount point on the local (client)
system must be an absolute path starting at the root directory (/
nfs_mount: RPC: Authentication error;why=Client credential too weakExplanation:
The server is allowing client superuser mounts only
and you are not a superuser.
See
mountd(8)
nfs_mount: RPC: Authentication error;why=Server rejected credentialExplanation: Possible reasons for this error message include the following:
The server is running with Internet address verification turned
on and it cannot resolve your Internet address.
If your system has multiple
network interfaces configured, the server must be able to resolve all IP addresses,
either using the local
/etc/hosts
file or the distributed
hosts
file.
The server is running with domain or subdomain verification turned on and your system is not in the same domain or subdomain as the server.
See
mountd(8)
nfs_mount:
xxx server not responding:
port mapper failurerpc timed out Giving
up on
yyyExplanation: The server you are trying to mount from is down, or its port mapper is inoperative.
User Action:
Log in remotely to the server. If you are able to log in, the network is working.
Execute the
rpcinfo
command from the server.
For example, for a server named host2, you would enter the following command:
# /usr/sbin/rpcinfo -p host2
If the port mapper is running properly on the server, the
rpcinfo
command lists the registered program numbers.
If it does
not, restart the port mapper on the server.
You also need a port mapper running
on the client host; if it is not running there, start it.
See
portmap(8)
After you restart the port mapper, stop the NFS daemons by entering the following command:
# /sbin/init.d/nfs stop
If NIS is running, stop the
ypbind
daemon on the server.
Use the
ps
command
to obtain the process ID (PID) and the
kill
command to
stop the process:
# ps -A | grep ypbind 439 ?? I 0:00.02 /usr/sbin/ypbind ... 170866 pts/3 S + 0:00.01 grep ypbind # kill -9 439
If you stopped the
ypbind
daemon, restart
it by entering the following command:
# /usr/sbin/ypbind &
Restart the NFS daemons on the server by entering the following command:
# /sbin/init.d/nfs start
nfs_mount:
xxx
server not responding: rpc prog not registeredExplanation:
The
mount
command got through to the port mapper, but the NFS
mountd
daemon was not registered.
User Action:
Log in to the server.
Verify that the
/usr/sbin/mountd
file exists
by using the
ls
command.
Run the
ps
command to see if the
mountd
daemon is running.
If it is not running, restart it by entering
the following command:
# /usr/sbin/mountd
Can't get net id for hostExplanation:
There is no entry in the
/etc/hosts
file for the NFS server specified in the
mount
command line.
If NIS is running, there is no entry in the
hosts
NIS map for the host name specified.
If BIND is running, there
is no entry in the
hosts
database for the host name specified.
C.2.2 Automount Error Messages
The
following error messages are issued to the screen or console or sent to the
syslogd
daemon by the
automount
program:
bad entry in map
mapnameExplanation:
The map entry
in
mapname
is malformed and the
automount
program cannot interpret it.
User Action: Verify the entry; you might need to include escape characters.
Can't mount
mountpoint:
reasonExplanation:
The
automount
program cannot mount itself at
mountpoint.
The error is indicated in the
reason
statement.
couldn't create directory:
reasonExplanation:
The
automount
program could not create a directory.
The error is indicated
in the
reason
statement.
dir
mountpoint
must start with '/'Explanation: The mountpoint must have a full pathname.
User Action: Verify both the spelling and path name of the mount point.
hierarchical mountpoint:
mountpointExplanation:
The
automount
program will not allow itself to be mounted within an
automounted directory.
User Action: Use another strategy to mount the directory.
host
hostname
not respondingExplanation:
The
automount
program attempted to mount from
hostname
but received no response or failed.
These errors could
indicate a server or network problem.
server not
respondingExplanation:
The
automount
program attempted to mount from
hostname
but received no response or failed.
These errors could
indicate a server or network problem.
hostname: exports:
rpc_errExplanation:
The
automount
program encountered an error
while attempting to get the list of exported file systems and directories
that it is allowed to mount from
hostname.
This error occurs when a user attempts to access a mount point that
has the
-hosts
map associated with it.
This error
indicates a server or network problem.
already mounted
on
mountpointExplanation:
The
automount
program
is attempting to mount a file system on a mount point that has already been
mounted with that file system.
map
mapname,
key
key:
badExplanation:
The map entry
in
mapname
is malformed and the
automount
program cannot interpret it.
User Action: Verify the entry; you might need to include escape characters.
mapname: Not foundExplanation:
The
automount
program cannot locate the map it requires.
This message
is returned only when you specify the
-v
option.
Explanation:
The
automount
program encountered an error when looking up an NIS map
entry.
Mount of
hostname:filesystem
on
mountpoint:
reasonExplanation:
The
automount
program attempted to mount from
hostname
but received no response or failed.
These errors could indicate
a server or network problem.
mountpoint: Not a directoryExplanation: The mountpoint exists but is not a directory.
User Action: Verify both the spelling and pathname of the mount point.
pathname from
hostname:
absolute symbolic linkExplanation:
The
automount
program detected
that
mountpoint
is an absolute symbolic link (begins
with /).
The content of the link is
pathname.
Because this might have undesired consequences on the client, the
automount
program will not mount on absolute symbolic links.
no mount maps specifiedExplanation:
The
automount
program
cannot find any maps to serve, nor can it find any NIS maps.
This message
is returned only when you specify the
-v
option.
WARNING:
hostname:file system
already mounted on mountpointExplanation:
The
automount
program is mounting
itself on top of an existing mount point.
This message is a warning only.
WARNING:
mountpoint
not empty!Explanation:
The
mountpoint
directory is not empty.
This message is returned only when you specify the
-v
option.
It is warning you that the previous contents of
mountpoint
will not be accessible while the mount is in effect.
The following error messages can occur when a file system
is exported from multiple servers as specified in a multiple-server map entry.
They indicate possible network problems that can occur when the
automount
daemon requests a response from the servers.
Cannot create socket for broadcast rpc:
rpc_errExplanation: No server in a multiple-server map entry is responding. This indicates that the replicated file system could not be reached on any of the specified servers.
Cannot receive reply to many_cast:
rpc_errExplanation: No server in a multiple-server map entry is responding. This indicates that the replicated file system could not be reached on any of the specified servers.
Cannot send broadcast packet:
rpc_errExplanation: No server in a multiple-server map entry is responding. This indicates that the replicated file system could not be reached on any of the specified servers.
Many_cast select problem:
rpc_errExplanation: No server in a multiple-server map entry is responding. This indicates that the replicated file system could not be reached on any of the specified servers.
NFS server (pid
n@mountpoint)
not responding still tryingExplanation:
An NFS request to the automount daemon with PID
n
serving mount point has timed out.
The
automount
daemon might be overloaded or not running.
User Action: If the condition persists, reboot the client. You can also do the following:
Exit all processes that are using automounted directories.
Kill the current
automount
process.
Restart the
automount
process from the
command line.
Remount
hostname:filesystem
on
mountpoint
server not respondingExplanation:
The
automount
program was attempting to remount
filesystem
because it discovered that a part of the automounted
hierarchy at the
mountpoint
was busy.
The remote
file system's server,
hostname, did not respond
to the mount request.
This error indicates a server problem.
trymany: servers not responding:
reasonExplanation: No server in a multiple-server map entry is responding. This indicates that the replicated file system could not be reached on any of the specified servers.
The following sections describe error messages for the two components
of AutoFS: the
autofsd
daemon and the
autofsmount
command.
C.2.3.1 autofsd Messages
The following error messages are issued to the screen or console or
sent to the
syslogd
daemon by the
autosfd
program:
autofs not configured
Explanation: AutoFS is not properly configured in the kernel.
User Action:
If necessary, add the
AUTOFS
option to the kernel configuration file and rebuild the kernel.
See
System Administration
for more information on modifying and rebuilding the
kernel.
Cannot create socket for nfs:
reasonExplanation: Network socket creation failed due to reason.
can't mount
hostnameExplanation:
A mount request
was rejected by the
mountd
daemon on
hostname.
This error usually indicates a permissions problem or that
the file system does not exist.
User Action:
Verify the export permissions
in the
/etc/exports
file on the server and verify that
the file system exists.
Can't ping mountd version
NFS-version
at server
hostname
reasonExplanation:
The
autofsd
daemon attempted to communicate with the
mountd
daemon on the
hostname
server, but received
no response.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
cfg_subsys_state returned
errorcode
Explanation: AutoFS is not properly configured in the kernel.
User Action:
If necessary, add the
AUTOFS
option to the kernel configuration file and rebuild the kernel.
See
System Administration
for more information on modifying and rebuilding the
kernel.
host
hostname
not respondingExplanation:
The
autofsd
daemon attempted to mount from
hostname
but it received no response or the request failed.
These errors could indicate a server or network problem.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
exports:
rpc_errExplanation:
The
autofsd
daemon encountered an error while
attempting to get the list of exported file systems and directories that is
allowed to mount from
hostname.
This occurs during
attempted access to a mount point with the
-hosts
map.
It
indicates a server or network problem.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
: mountd not responding
reasonExplanation:
The
autofsd
daemon attempted to
communicate with the
mountd
daemon on the
hostname
server, but received no response.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
: server's portmap not respondingExplanation:
The
autofsd
daemon attempted to communicate with the
portmap
daemon on the
hostname
server,
but received no response.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
lookup_addr: gethostbyname failed
error
for
hostnameExplanation:
The
autofsd
daemon
was unable to obtain a network address for the named host.
User Action: Verify the address for the host in the local hosts file and the DNS or NIS database. Verify that the DNS or NIS server is up and running.
match
mapname:keyname
failed:
reasonExplanation:
The
autofsd
daemon is having a problem reading the map file
mapname
to find key
keyname.
The
error is indicated in the
reason
statement.
Mount of
hostname:filesystem
on
mountpoint:
reasonExplanation:
The
autofsd
daemon attempted to mount from
hostname,
but it received no response or the request failed.
These errors could indicate
a server or network problem.
User Action: Verify the status of the network and verify that the server is properly configured and running NFS services.
Unable to locally serve
filesystemExplanation: Locally serving the file system would result in a circular symbolic link.
User Action: Choose a different mount point for the file system, or specify a different host to serve the file system.
autofs is not configured or not enabled
Explanation: AutoFS is not properly configured in the kernel.
User Action:
If necessary, add the
AUTOFS
option to the kernel configuration file and rebuild the kernel.
See
System Administration
for more information on modifying and rebuilding the
kernel.
Intercept
filesystem
mount failed:
reason
Explanation:
The attempt to create an intercept mount point for
filesystem
has failed due to
reason.
The
autofsmount
command issues this error message for direct
map entries, and when running in verbose mode, for indirect map entries as
well.
Map
mapname
does not existExplanation:
The
autofsmount
command could not find the specified direct or indirect
map file.
User Action: Ensure that you have specified the proper location for the map files on the command line or in your master map file.
Note: Indirect entry in map
mapname
with key
keyname
cannot be locally served with
the mounton and mountfrom directories as defined.
Explanation: An external server will be chosen to avoid a circular symbolic link.
Note: The hierarchical entry in map
mapname
for
keyname
cannot be served locally
Explanation: An external server will be chosen to avoid a circular symbolic link.
Note: The shared map entry in map
mapname
with key
keyname
will be converted to a non-shared
entry
Explanation: AutoFS does not support the shared mount syntax of Automount. It converts all shared map entries to their nonshared counterparts.
Unmount
filesystem:
reason
Explanation: An attempt to unmount filesystem has failed with reason.
Warning: Cannot support the hierarchy in map
mapname
with key
keyname
with the
mounton and mountfrom directories as defined.
Explanation:
The hierarchical direct
map entry for subdirectory
/
cannot be supported, as no
external servers are listed and locally serving it would create a circular
symbolic link.
The file systems in the map entry will be treated as though
they are individual map entries.
User Action: Specify an external server or change the key and/or the location of the file system in question to avoid a circular symoblic link.
Warning: Skipping entry in map
mapname
with key
keyname
Explanation: The file system will be locally served. No intercept mount point will be created, only a symbolic link.
Warning: The hierarchical entry in map
mapname
for
keyname
will not work.
Explanation:
A hierarchical
direct map entry for some subdirectory other than
/
cannot
be supported, as no external servers are listed and locally serving it would
create a circular symbolic link.
The file systems in the map entry will be
treated as though they are individual map entries.
User Action: Specify an external server or change the key and/or the location of the file system in question to avoid a circular symoblic link.
Warning: There are no servers available for this entry
Explanation: In the context of previous messages for this map entry, this error message indicates that locally serving the file system would create a circular symbolic link, and no external servers are specified.
User Action: Specify an external server or change the key and/or the location of the file system in question to avoid a circular symoblic link.
The following error messages might be displayed on the NFS client system console and in the error logger. They note an NFS file access failure.
NFS server
hostname
not responding, still tryingExplanation: File operations in a hard-mounted file system are suspended because communication between the client and the server has stopped.
NFS server
hostname
okExplanation: File operations have resumed.
NFS
file operation
failed for server
hostname:
reasonExplanation: If the operation is in a soft-mounted file system and the server is inoperable, the reason for the failure is that the operation timed out.
NFS write error, server
hostname,
remote file system fullExplanation: A write operation failed because the remote file system is full.
NFS write error
errno,
server
hostname,
fs(n,n),
file
fileExplanation:
A write operation
was refused by the server.
The
fs
and
file
variables are parts of the file handle (fhandle).
See
errno(2)