Appendix A. ONC3/NFS Error Messages

This chapter explains error messages generated during NFS mounting and by the automounter. It contains these sections:

mount Error Messages

This section gives detailed descriptions of the NFS mounting failures that generate error messages.

/etc/mtab: No such file or directory  


The mounted file system table is kept in the file /etc/mtab. This file must exist before mount(1M) can succeed.

mount: ... already mounted  


The file system that you are trying to mount is already mounted or there is an incorrect entry for it in /etc/mtab.

mount: ... Block device required  


You probably left off the host name (krypton:) portion of your entry:

# mount krypton:/usr/src /krypton.src

The mount command assumes you are doing a local mount unless it sees a colon in the file system name or the file system type is nfs in /etc/fstab. See fstab(4).

mount: ... not found in /etc/fstab  


If you use mount with only a directory or file system name, but not both, it looks in /etc/fstab for an entry with file system or directory field matching the argument. For example,

# mount /krypton.src

searches /etc/fstab for a line that has a directory name field of /krypton.src. If it finds an entry, such as this,

krypton:/usr/src /krypton.src nfs rw,hard 0 0

it mounts as if you had given this command:

# mount krypton:/usr/src /krypton.src

If you see this message, it means the argument you gave mount is not in any of the entries in /etc/fstab.

/etc/fstab: No such file or directory  


mount tried to look up the name in /etc/fstab but there was no /etc/fstab.

mount: ... not in hosts database  


The host name you gave is not in the /etc/hosts database. Check the spelling and the placement of the colon in your mount call. Try to rlogin(1C) or rcp(1C) to the other system.

mount: directory path must begin with a slash (/).  


The second argument to mount is the path of the directory to be mounted. This must be an absolute path starting at /.

mount: ... server not responding: RPC: Port mapper failure  


Either the server from which you are trying to mount is inactive, or its portmap(1M) daemon is inactive or hung. Try logging in to that system. If you can log in, give this command:

# /usr/etc/rpcinfo –p hostname

This should produce a list of registered program numbers. If it does not, start the portmap daemon again. Note that starting the portmap daemon again requires that you kill and restart inetd(1M), ypbind(1M), and ypserv(1M). ypbind is active only if you are using the NIS service. The ypserv daemon only runs on NIS servers. See network(1M) for information about how to stop and restart daemons.

There are two methods for dealing with a server that is inactive or whose portmap daemon is not responding. You could reboot the server or you perform these commands:

  1. On the server, become the superuser and kill the daemons. Give this command:

    # /etc/killall portmap inetd ypbind
    

  2. Start new daemons:

    # /usr/etc/portmap
    # /usr/etc/ypbind
    # /usr/etc/inetd
    

If you cannot rlogin to the server, but the server is operational, check your network connection by trying rlogin to some other system. Also check the server's network connection.

mount: ... server not responding: RPC: Program not registered  


This means mount reached the portmap daemon but the NFS mount daemon (rpc.mountd(1M)) was not registered.

Go to the server and be sure that /usr/etc/rpc.mountd exists and that an entry appears in /etc/inetd.conf exactly like this (shown wrapped):

mountd/1 dgram rpc/udp wait root
/usr/etc/rpc.mountd mountd

Give the command ps –de to be sure that the internet daemon (inetd) is running. If you had to change /etc/inetd.conf, give this command:

# /etc/killall 1 inetd

This command informs inetd that you have changed /etc/inetd.conf.

mount: ... No such file or directory 


Either the remote directory or the local directory does not exist. Check your spelling. Use the ls(1) command for the local and remote directories. For SGI systems, check to see if you are attempting to access a hidden file or directory:

# showmount -x servername

and check for file systems exported without the nohide option.

mount: not in export list for ...  


Your host name is not in the export list for the file system you want to mount from the server. You can get a list of the server's exported file systems with this command:

# showmount –e servername

If the file system you want is not in the list, or your host name or network group name is not in the user list for the file system, log in to the server and check the /etc/exports file for the correct file system entry. A file system name that appears in the /etc/exports file but not in the output from showmount(1M) indicates that you need to run exportfs(1M).

mount: ... Permission denied  


This message is a generic indication that some authentication failed on the server. It could simply be that you are not in the export list (see above), the server could not figure out who you are, or the server does not recognize that you are who you say you are. Check the server's /etc/exports. In the last case, check the consistency of the NIS and local host name and user ID information.

mount: ... Not a directory  


Either the remote path or the local path is not a directory. Check your spelling and use the ls command for both the local and remote directories.

mount: ... You must be root to use mount  


You must do the mount as root on your system because it affects the file system for the whole system, not just your directories.

Verbose automount Error Messages

The following error messages are likely to be displayed if the automounter fails and the verbose option is on (automount(1M) –v option). Below each error message is a description of the probable cause of the problem.

no mount maps specified  


The automounter was invoked with no maps to serve, and it cannot find the NIS auto.master map. Recheck the command, and check for the existence of an NIS auto.master map:

# ypwhich -m | grep auto.master

mapname: Not found  


The required map cannot be located. This message is produced only when the –v option is given. Check the spelling and pathname of the map name.

leading space in map entry entry text in mapname  


The automounter has discovered an entry in an automount map that contains leading spaces. This is usually an indication of an improperly continued map entry. For example:

foo
bar geez:/usr/geez

In this example, the warning is generated when the automounter encounters the second line, because the first line should be terminated with a backslash (\).

bad directory directory in indirect map mapname  


While scanning an indirect map, the automounter has found an entry directory containing a "/". Indirect map directories must be simple names, not pathnames.

bad directory directory in direct map mapname  


While scanning a direct map, the automounter has found an entry directory without a leading "/". Directories in direct maps must be full pathnames.

NIS bind failed 


The automounter was unable to communicate with the ypbind daemon. This is information only — the automounter continues to function correctly provided it requires no explicit NIS support. If you need NIS, check to see if there is a ypbind daemon running.

Couldn't create mountpoint mountpoint: reason  


The automounter was unable to create a mount point required for a mount. This most frequently occurs when attempting to hierarchically mount all of a server's exported file systems. A required mount point may exist only in a file system that cannot be mounted (it may not be exported) and it cannot be created because the exported parent file system is exported read only.

WARNING: mountpoint already mounted on  


The automounter is attempting to mount over an existing mount point. This is indicative of an automounter internal error (bug).

server:pathname already mounted on mountpoint  


The automounter is attempting to mount over a previous mount of the same file system. This could happen if an entry appears both in /etc/fstab and in an automounter map (either by accident or because the output of mount -p was redirected to /etc/fstab). Delete one of the redundant entries.

can't mount server:pathname: reason  


The mount daemon on the server refuses to provide a file handle for server:pathname. Check the server's export list.

remount server:pathname on mountpoint: server not responding  


The automounter has failed to remount a file system it previously unmounted. This message may appear at intervals until the file system is successfully remounted.

WARNING: mountpoint not empty  


The mount point is not an empty directory. The directory contains entries that are hidden while the automounter is mounted there. This is advisory only.

General automount Error Messages

This section lists error messages generated by the automounter that can occur at any time.

WARNING: default option "option" ignored for map mapname  


Where option is an unrecognized default mount option for the map mapname.

option ignored for directory in mapname  


The automounter has detected an unknown mount option. This is advisory only. Correct the entry in the appropriate map.

bad entry in map mapname "directory" map mapname, directory directory: bad  


The map entry is malformed, and the automounter cannot interpret it. Recheck the entry; perhaps there are characters in it that need a special escape sequence.

Can't get my address  


The automounter cannot find an entry for the local system in the host database.

Cannot create UDP service  


Automounter cannot establish a UDP connection.

svc_register failed  


Automounter cannot register itself as an NFS server. Check the kernel configuration file.

couldn't create pathname: reason  


Where pathname is /tmp_mnt or the argument to the –M command line option.

Can't mount mountpoint: reason  


The automounter couldn't mount its daemon at mountpoint.

Can't update pathname  


Where pathname is /etc/mtab it means that the automounter was not able to update the mount table. Check the permissions of the file.

exiting  

This is an advisory message only. The automounter has received a SIGTERM (has been killed) and is exiting.

WARNING: pathname: line line_number: bad entry  


Where pathname is /etc/mtab it means that the automounter has detected a malformed entry in the /etc/mtab file.

server:pathname no longer mounted  


The automounter is acknowledging that server:pathname which it mounted earlier has been unmounted by the umount(1M) command. The automounter notices this within 1 minute of the unmount or immediately if it receives a SIGHUP.

trymany: servers not responding: reason 


No server in a replicated list is responding. This may indicate a network problem.

server:pathnamelinkname : dangerous symbolic link  


The automounter is trying to use server:pathname as a mount point but it is a symbolic link that resolves to a pathname referencing a mount point outside of /tmp_mnt (or the mount point set with the –M option). The automounter refuses to do this mount because it could cause problems in the system's file system, e.g. mounting on /usr rather than in /tmp_mnt.

host server not responding  


The automounter attempted to contact server but received no response.

Mount of server:pathname on mountpoint: reason  


The automounter failed to do a mount. This may indicate a server or network problem.

pathconf: server: server not responding 


The automounter is unable to contact the mount daemon on the server that provides portable operating systems based on UNIX (POSIX) pathconf(2) information.

pathconf: no info for server:pathname  


The automounter failed to get pathconf information for pathname.

hierarchical mountpoints: pathname1 and pathname2  


The automounter does not allow its mount points to have a hierarchical relationship. An automounter mount point must not be contained within another automounted file system.

mountpoint: Not a directory  


The automounter cannot mount itself on mountpoint because mountpoint is not a directory. Check the spelling and pathname of the mount point.

dir mountpoint must start with '/'  


Automounter mount point must be given as full pathname. Check the spelling and pathname of the mount point.

mapname: yp_err  


Error in looking up an entry in an NIS map. May indicate NIS problems.

hostname: exports: rpc_err  


Error getting export list from hostname. This indicates a server or network problem.

nfscast: cannot send packet: reason  


The automounter cannot send a query packet to a server in a list of replicated file system locations.

nfscast: cannot receive reply: reason  


The automounter cannot receive replies from any of the servers in a list of replicated file system locations.

nfscast:select: reason Cannot create socket for nfs: rpc_err  


These error messages indicate problems attempting to contact servers for a replicated file system. This may indicate a network problem.

NFS server (pid@mountpoint) not responding; still trying  


An NFS request made to the automount daemon with process identifier pid serving mountpoint has timed out. The automounter may be temporarily overloaded or dead. Wait a few minutes. If the condition persists, reboot the client or use fuser(1M) to find and kill all processes that use automounted directories (or change to a non-automounted directory in the case of a shell), kill the current automount process, and restart it again from the command line.

General CacheFS Errors

This section describes the error messages that may be generated from commands used to administer the CacheFS file system.

cfsadmin Error Messages

This section gives detailed descriptions of the CacheFS cfsadmin command failures that generate error messages.

cfsadmin: must be run by root  


You must be logged in as root to run cfsadmin(1M).

cfsadmin: Cache name is in use and cannot be modified. 


This error occurs when you attempt to remove a cache ID from the cache name, if the cache is active (has mounted file systems).

cfsadmin: cachepath already exists. 


The cache directory path cachepath specified with the cachedir option already exists. The last component of the path cachepath must not exist when creating a cache. All other path components must exist.

cfsadmin: creating labelpath failed. 


The cache label file labelpath could not be created. This message always appears with one of the following:

Could not remove labelpath: errmsg
Error creating labelpath: errmsg
Writing labelpath failed: errmsg
Writing labelpath failed on sync: errmsg

 

In each case, the system error is given in errmsg.

cfsadmin: create resource failed: errmsg 


The cache resource file resource could not be created. The system error is given in errmsg.

cfsadmin: Cache cachedir is in use and cannot be modified. 


The cache cachedir was in use when an attempt was made to modify the contents of the cache label. This operation may only be performed when the cache has no mounted file systems.

cfsadmin: Cache size cannot be reduced, maxblocks current p%, requested n% 


An attempt was made to reduce the maximum file system block allocation percentage from p% to n%. The allocation can only be increased.

cfsadmin: Cache size cannot be reduced, maxfiles current p% requested n% 


An attempt was made to reduce the maximum number of files allocated from p% to n%. The allocation can only be increased.

cfsadmin: cacheid is not a valid cache id. 


The cache identifier given by cacheid is not valid. You may have specified an invalid cache identifier on the command line for cfsadmin. This can occur when deleting a cache.

cfsadmin: lowblocks can't be >= hiblocks. 


The block allocation specified by minblocks is greater than or equal to that specified by maxblocks. minblocks must be less than maxblocks.

cfsadmin: lowfiles can't be >= hifiles.  


The file allocation specified by minfiles is greater than or equal to that specified by maxfiles. minfiles must be less than maxfiles.

cfsadmin: Could not open resource: errmsg, run fsck 


The cache resource file resource could not be opened. The error message is given by errmsg. cachefs_fsck(1M) should be run.

cfsadmin: Could not read cache_usage, val, run fsck 


The cache usage structure could not be read from the resource file. val is the return value from read(2).

cfsadmin: Could not open option file optpath 


The cache option file optpath could not be opened. The entire cache should be removed and reconstructed.

cfsadmin: Could not read option file optpath 


The cache option file optpath could not be read. The entire cache should be removed and reconstructed.

cfsadmin: Reading cachelabel failed. 


The cache label file cachelabel could not be read. This message appears with one of the following messages:

Cannot stat file cachelabel: errmsg
File cachelabel does not exist.
Cache label file cachelabel corrupted
Cache label file cachelabel wrong size
Error opening cachelabel: errmsg
Reading cachelabel failed: errmsg

 

The above messages occur when the cache label file is not a regular file or the label contains the incorrect cache version. The system error is given in errmsg.

cfsadmin: Could not open resource: errmsg, run fsck  


The resource file resource could not be opened in order to enlarge it or mark it as dirty. The error is given in errmsg. cachefs_fsck(1M) should be run.

cfsadmin: Resource file has wrong size cursize expected, run fsck  


The size of the resource file is incorrect. Its size is cursize when it should be expected.

cfsadmin: Could not write cache_usage, val, run fsck  


The cache usage structure could not be written to the resource file. val is the return value from write(2).

cfsadmin: Could not write file, val, run fsck  


The expanded resource file could not be initialized. val is the return value from write(2).

mount_cachefs Error Messages

This section gives detailed descriptions of the CacheFS mounting failures that generate error messages.

mount_cachefs(1M) is normally executed from mount(1M).

mount_cachefs: must be run by root  


You must be logged in as root to run mount(1M).

mount_cachefs: mount failed, options do not match. 


The mount options supplied on the mount command are not compatible with the mount options currently set for the cache. This occurs when there are multiple mount points for one cache. All mount points must have the same options.

mount_cachefs: suid and nosuid are mutually exclusive 


Both of the options suid and nosuid have been specified. Only one is allowed.

mount_cachefs: rw and ro are mutually exclusive  


Both of the options rw and ro have been specified. Only one is allowed.

mount_cachefs: acregmin cannot be greater than acregmax  


The specified acregmin option has a value greater than that for acregmax.

mount_cachefs: acdirmin cannot be greater than acdirmax  


The specified acdirmin option has a value greater than that for acdirmax.

mount_cachefs: only one of non-shared or write-around may be specified  


Both of the options non-shared and write-around have been specified. Only one is allowed.

umount_cachefs Error Messages

This section gives detailed descriptions of the CacheFS unmounting failures that generate error messages.

umount_cachefs(1M) is normally executed from umount(1M).

umount_cachefs: must be run by root  


You must be logged in as root to run umount(1M).

umount_cachefs: warning: dir not in mtab  


The mount point directory dir has no entry in the mount table. This means that the mount table has been corrupted. The umount command can still be successful; however, the back file system can not be unmounted.

umount_cachefs: could not exec /sbin/umount on back file system errmsg  


An attempt was made to run umount on the back file system and failed. The system error is given in errmsg.