I seem to be having problems with automountd. I had posted a previous
message describing this problem, it started since my last patch update
[mount problem since last patch update]. I am running Solaris 8 and
NIS. I now am sure it has something to do with autofs. As mentioned I
tested mounting NFS directories and it worked but no automount.
I checked /var/adm/messages and found a series of messages:
[ID 995402 kern.notice] /kernel/fs/sparcv9/autofs: undefined symbol
'mntfs_optprint'
[ID 472681 kern.notice] WARNING: mod_load: cannot load module 'autofs'

All that was done was download the recommended and security patches
from sunsolve.sun.com and update the system [Nov 17] and now autofs
does not load?
I found a similar mentioned problem on newsgroup "comp.unix.solaris"
subject: autofs problem an entry from Dan Stromberg
(strombrg@dcs.nac.uci.edu), who had the same problem with the same
error messages but he appears to have solved the problem by just
updating the recommended patches. So I did this again [using the
latest Dec 1 release], but still the same problem.
I have run automount -v and this is what I got:
automount: mount /home: Invalid argument
automount: mount /net: Invalid argument
automount: mount /xfn: Invalid argument
automount: no mounts
automount: no unmounts

I have checked the auto_master, auto_home and nsswitch.conf files and
all are the same as before [and my other machines that I have not
updated because of this].
> more /etc/auto_master
# Master map for automounter
#
+auto_master
/net -hosts -nosuid
/home auto_home
/xfn -xfn
>

> more /etc/auto_home
# Home directory map for automounter
#
+auto_home
>

> cat /etc/nsswitch.conf | grep automount
automount: files nis
>

I also tried to start autofs and this is what I get:
#/etc/init.d/autofs start
#automount: mount /home: Invalid argument
automount: mount /net: Invalid argument
automount: mount /xfn: Invalid argument

One thing I did notice is that my clock and timezone had been changed.
I have now corrected the timezone and clock but still have autofs
problem.
This really is driving me nuts I really do not want to re-install the
system from scratch and its also holding up patch updates for my other
machines because I can not risk this happening to other systems too.
Please let me know if some of you have seen this before and if you
know how to fix it.

-James

-Thanx Tim but unfortunatly I do not have a support contract.
-Thanx Jim for the link but it lists both my postings and the one from
Dan Stromberg which I mentioned, unfortuantly he fixes his problem by
running the latest update to recommended patches. I just did that and
I still have the same problem. In the same stream I did read the
posting by Beardy (beardy@beardy.net)
This is what I get:
# /usr/ccs/bin/nm /kernel/fs/sparcv9/mntfs | grep optprint
[44] | 1388| 336|FUNC |LOCL |0 |1
|mntfs_optprint

There was another posting suggesting that LOCL might be wrong and the
system might have been compromised and to check mntfs against the
Solaris Figerprint Database. I did that and it is OK.

I could not run the second suggested command "pkgchk -v SUNWcarx 2>&1
| grep mntfs" but this is what I got when I tried a few others:
# pkgchk -v SUNWcarx 2 > & 1 | grep mntfs
Ambiguous output redirect.
# pkginfo | grep SUNWcarx
system SUNWcarx Core Architecture, (Root) (64-bit)
# pkgchk SUNWcarx
ERROR: /kernel/fs/sparcv9/mntfs
file size <18528> expected <18504> actual
file cksum <45052> expected <45998> actual
ERROR: /kernel/misc/sparcv9/klmmod
file size <91520> expected <91832> actual
file cksum <48817> expected <58056> actual
ERROR: /kernel/misc/sparcv9/klmops
file size <89976> expected <89984> actual
file cksum <50856> expected <51056> actual
ERROR: /platform/SUNW,Sun-Fire-880/kernel/drv/sparcv9/hpc3130
file size <26928> expected <27000> actual
file cksum <11782> expected <1842> actual

Beardy (beardy@beardy.net) then says
"And if the suitable one returns a line saying "ERROR", then the file
is
broke.
If /kernel/fs/autofs is whingeing at boot about a symbol that should
be
defined in /kernel/fs/mntfs (or their 64-bit equivalents), then I
suspect that the mntfs modules are broke somehow, and that there is no
problem with 108993-25. You should be able to get them back from the
CD
distro."
I am getting errors but my problem is I don't exactly know how to get
them back from CD. Do I have to get the whole Core Architecture or can
I just get mntfs? and how do I do it?

-Thanx Casper, I think you might be right. I does look like maybe a
partially installed patch. Do you know how I can check for this? or
how to check the integrity of the patch database and see if it has
corrupted patches.

-Thanx Thomas but I do not use disksuit nor do I mirror my root disk
on this system.

well I'll keep at it but still no dice on fixing my system, thanx all
you guys for answering my posting. I'll be sure to post a fix if I
find one. In the mean time still loooking for that Solaris Guru that
might be able to help me.

cheers

-James