I've added this mount in vfstab:
Kaapstad:/admin - /Kaapstadadmin nfs - yes -
I've added the /Kaapstadadmin directory as well. Now when I boot the
system, I get the message that the mountpoint cannot be determined, and
several other error-messages (can't remember which) regarding this
mount. Oddly (luckily?) enough, the mount does work flawlessly.
Why do I get these messages?
Thanks
Huub
Hold on a second while I peer into my crystal ball and read
your mind ... wait a sec ... I've almost got it ... it says ...
"here's 25 cents, go buy yourself a clue". Apparently either
my crystal ball doesn't work across the Atlantic or the transmission
problem is on your end with the bioware unit sitting in front of
your terminal rather than with your computer's software or hardware.
If you can remember (and post) the *actual* text of those
"error-messages" then someone might remember where the error
message is documented and/or tell you why you get them.
Also, it would help to see the actual output of the "mount" command
to make sense out of your (contradictory) conclusion that "the mount
does work flawlessly" (or maybe they've changed the definitions of
"error-messages" and "flawlessly" since I last checked a dictionary).
Thanks
Huub
Does your /etc/hosts file contain the ip address for Kaapstad?
If it does is that directory exported with the correct permissions?
Mike
Huub
> > RTFM: http://docs.sun.com/db/doc/806-4074/6jd68pq38?a=view
Wrong section. Read the next section about consadm. The messages
from the rc scripts are "special" and aren't handled via syslogd.
That makes it slightly harder to "find" and/or capture them.
To capture them you have to use consadm to setup a terminal with
some sort of logging capability. Personally I use headless systems
permanently set up with serial consoles rather than consadm but
serial consoles are a completely different kettle of fish (many
PC motherboards can't handle not having a keyboard and display).
OTOH, for problems like this, copying the text right off the screen
usually works just fine unless you've got some sort of really hairy
error message text scrolling by (in which case you have to go the
consadm route, to ensure completeness and accuracy). The message text
for most startup script errors is usually simple enough that most
people seem to be able to handle just writing it down off of the
screen as it scrolls by. Perhaps disabling the Desktop Login might
make it easier for you to read the messages as they scroll by? And since
you say it happens on every reboot, you can try rebooting a couple of
times and write down a portion of the message(s) on each reboot.
What, no mention of dmesg?
-Greg
--
::::::::::::: Greg Andrews ::::: ge...@panix.com :::::::::::::
I have a map of the United States that's actual size.
-- Steven Wright
Surely you don't expect me to exhaustively enumerate every wrong
answer in addition to supplying the right answer (which was "RTFM").
Or did you mean to imply that you thought "dmesg" was the right answer?
You really ought to make at least a minimal effort to check your
facts before you try to catch me in a blunder. If you bother to
read the Sys Admin Guide I pointed to, I think you'll find that
dmesg is covered in the same section as /var/adm/messages. In
other words, it's also the *wrong* place to look for rc script
messages. In fact nowadays, if you look at the internals, dmesg
and syslogd are implemented on top of the exact same underlying
kernel functions.
mount: mount point cannot be determined
nfs mount: mount: /mnt/Kaapstadadmin: no such file or directory
nfs mount: mount: device busy
nfs mount: mount: device busy
Actual situation is: I've add this line in /etc/vfstab:
Kaapstad:/admin - /mnt/Kaapstadadmin nfs - yes -
And created the Kaapstadadmin directory in /mnt
Now when I enter the mount I can access all directories and files on
that server. So, if this is in any 'rtfm', please tell me where..
Thanks for you patience,
Huub
Huub> I've rebooted 5 times now and written it down:
Huub> mount: mount point cannot be determined
Huub> nfs mount: mount: /mnt/Kaapstadadmin: no such file or directory
Huub> nfs mount: mount: device busy
Huub> nfs mount: mount: device busy
Huub> Actual situation is: I've add this line in /etc/vfstab:
Huub> Kaapstad:/admin - /mnt/Kaapstadadmin nfs - yes -
Huub> And created the Kaapstadadmin directory in /mnt
Huub> Now when I enter the mount I can access all directories and files on
Huub> that server. So, if this is in any 'rtfm', please tell me where..
It seems like this get automounted at boot time.
What does /etc/auto_master look like? Is the system set up to be a
NIS-client? In that case what does 'ypcat auto_master' look like? Or
is it NIS+? Then it's 'niscat auto_master.org_dir'
If *either* of the above contains specifications for your mount your
mount will not succeed since the automounter then has control of that
mount point. And if everything gets mounted the way you want it using
the automounter, then the line in /etc/vfstab is useless.
/jb
Why on earth would you say in your original message that your
mountpoint was in the "/" directory when it's actually in "/mnt"?
Surely it's obvious that details like that matter.
> And created the Kaapstadadmin directory in /mnt
>
> Now when I enter the mount I can access all directories and files on
> that server. So, if this is in any 'rtfm', please tell me where..
That's only the first half of what I said was necessary to diagnose
your problem. In my original response I said:
"Also, it would help to see the actual output
of the "mount" command ..."
By "it would help" I meant "it's necessary".
I can't be certain what you've done until I see that output, but I
strongly suspect that you've got multiple entries in your vfstab that
refer to "/mnt". You've probably got NFS mounts on top of NFS mounts
and assumed that the NFS mounts would be processed in the order they
appear in vfstab. They're not, they're processed in parallel.
I suggest you do the following to understand what's really happening
on your system:
Comment out the "Kaapstad" entry in your vfstab.
Reboot your system into single-user mode (type "b -s" at the
"Select (b)oot ..." prompt). In that state, what's the output
of the commands "ls -la /mnt" and "mount -v"?
Type control-D to let your system continue to multi-user mode.
In multi-user mode, what's the output of the commands "ls -la /mnt"
and "mount -v"?
Finally, what's the output of the command "grep /mnt /etc/vfstab"?
Of course, I'm just guessing the problem is in your vfstab. There are
lots of other ways you could creatively bugger the /mnt directory.
For example, you could modify your automount config to point to
the "/mnt" directory. Or you could have a custom rc script that
temporarily mounts something on /mnt during the transition to
multi-user mode (at the same time as your NFS mounts). Or you could
have installed or configured some sort of background daemon to make
use of /mnt.
IMHO, experienced sys-admins don't usually permanently mount
anything on "/mnt", since every example in the docs that involves
a mount command uses "/mnt" as the mountpoint. In other words, only
use "/mnt" for temporary short-lived mounts that you expect to almost
immediately unmount. For permanent mounts, create a new mountpoint
in "/" using some name your absolutely certain isn't used for any
other purpose. (If that advice isn't in some guide somewhere, it's
probably only because it's so obvious).
> I suggest you do the following to understand what's really happening
> on your system:
>
> Comment out the "Kaapstad" entry in your vfstab.
>
> Reboot your system into single-user mode (type "b -s" at the
> "Select (b)oot ..." prompt). In that state, what's the output
> of the commands "ls -la /mnt" and "mount -v"?
#ls -la /mnt
total 6
drwxr-xr-x 3 root sys 512 Apr 14 12:58 .
drwxr-xr-x 28 root root 1024 Apr 15 08:44 ..
drwxr-xr-x 2 root other 512 Apr 14 12:1 Kaapstadadmin
# mount -v
/dev/dsk/c0d0s0 on / type ufs
read/write/setuid/intr/largefiles/onerror=panic/dev=1980000
on Tue Apr 15 08:46:51 2003
/dev/dsk/c0d0p0:boot on /boot type pcfs
read/write/setuid/nohidden/nofoldcase/dev=19a3010 on Tue Apr 15 08:46:50
2003
/proc pn /proc type proc read/write/setuid/dev=2d240000 on Tue Apr 15
08:46:51 2003
fd on /dev/fd type fd read/write/setuid/dev=2e00000 on Tue Apr 15
08:46:52 2003
mnttab on etc/mnttab type mntfs read/write/setuid/dev=2f00000 on Tue apr
15 08:45:54 2003
swap on /var/run type tmpfs read/write/setuid/dev=1 on Tue Apr 08:46:54 2003
>
> Type control-D to let your system continue to multi-user mode.
> In multi-user mode, what's the output of the commands "ls -la /mnt"
> and "mount -v"?
# ls -la /mnt
total 6
drwxr-xr-x 3 root sys 512 Apr 14 12:58 .
drwxr-xr-x 28 root root 1024 Apr 15 08:59 ..
drwxr-xr-x 2 root other 512 Apr 14 12:58 Kaapstadadmin
# mount -v
/dev/dsk/c0d0s0 on / type ufs
read/write/setuid/intr/largefiles/onerror=panic/dev=1980000 on Tue Apr
15 08:58:18 2003
/dev/dsk/c0d0p0:boot on /boot type pcfs
read/write/setuid/nohidden/nofoldcase/dev=19a3010 on Tue Apr 15 08:58:17
2003
/proc on /proc type proc read/write/setuid/dev=2d40000 on Tue Apr 15
08:58:18 2003
fd on /dev/fd type fd read/write/setuid/dev=2e00000 on Tue Apr 15
08:58:19 2003
mnttab on /etc/mnttab type mntfs read/write/setuid/dev=2f00000 on Tue
Apr 15 08:58:21 2003
swap on /var/run type tmpfs read/write/setuid/dev=1 on Tue Apr 15
08:58:21 2003
swap on /tmp type tmpfs read/write/setuid/dev=2 on Tue Apr 15 08:59:11 2003
/dev/dsk/c0d0s7 on /export/home type ufs
read/write/setuid/intr/largefiles/onerror=panic/dev=1980007 on Tue Apr
15 08:59:11 2003
#
>
> Finally, what's the output of the command "grep /mnt /etc/vfstab"?
>
# grep /mnt /etc/vfstab
#Kaapstad:/admin - /mnt/Kaapstadadmin nfs -
yes -#
Hope this help determining the cause. Thanks you.