On 16 Jul, 19:59, SteveB wrote:
> the response to the followin:
> #ps -ef
> returns
> #ps -ef
> UID PID PPID C STIME TTY TIME CMD
>
> with no process listed..
>
> Also I noticed that /var/syslog is empty. Today I moved the original
> out of the way and recreated it. Since then there have bin know
> entries into the syslog.
>
> I have rebooted the server, and the condition still exists.
> I have run both integrity and fixmog. I also ran "/etc/custom -V
> symlinks -x" and nothing has changed.
>
> You suggestions will appreciated.

This sounds bit like:

http://www.sco.com/ta/109457

If this TA does not help then I think you should contact
your SCO Support provider to help get to the bottom
of the problem.

John

SteveB wrote:
> the response to the followin:
> #ps -ef
> returns
> #ps -ef
> UID PID PPID C STIME TTY TIME CMD
>
> with no process listed..

So much for an industry-strength computer system like OpenServer 6...

Pepe typed (on Thu, Jul 17, 2008 at 11:01:38PM +0200):
> SteveB wrote:
>> the response to the followin:
>> #ps -ef
>> returns
>> #ps -ef
>> UID PID PPID C STIME TTY TIME CMD
>>
>> with no process listed..
>
> So much for an industry-strength computer system like OpenServer 6...

If I run the same command on my OSR 6 with MP3, I get 149 lines of
output.

--
JP

On Jul 17, 5:33 pm, Jean-Pierre Radley wrote:
> Pepe typed (on Thu, Jul 17, 2008 at 11:01:38PM +0200):
>
> > SteveB wrote:
> >> the response to the followin:
> >> #ps -ef
> >> returns
> >> #ps -ef
> >> UID PID PPID C STIME TTY TIME CMD
>
> >> with no process listed..
>
> > So much for an industry-strength computer system like OpenServer 6...
>
> If I run the same command on my OSR 6 with MP3, I get 149 lines of
> output.
>
> --
> JP

Well JP I'm not surprised. Until yesterday it worked.

With more poking around with mount I have found that the /proc and /
process filesystems are not mounting at all.

I ahve tried the following:
truss -o /tmp/mount.truss -aef mnt /proc

and get an error:
truss: /proc is not a PROC directory

I have also tried using mount and mountall with basically the same
result.

If I do the following:

echo proc | crash get followiing (I've edit the output for brevity)

46 s 1726 1 1726 0 0 80 1 e-d10243cc cupsd
48 s 1731 1709 1709 28 0 80 1 c-cdc66d50 httpd
49 s 1732 1709 1709 28 0 80 1 c-cdc66d50 httpd
50 s 1733 1709 1709 28 0 80 1 c-cdc66d50 httpd
52 s 1768 1 1768 0 0 80 1 e-d3025ebc smbd
52 s 1 59 0 u-ddac2000 smbd
54 s 1781 1768 1768 0 0 70 1 e-c04e7b24 smbd
54 s 1 59 0 u-d45e4200 smbd
55 s 19368 1 0 0 0 70 0 c-c81168cc uadmin
59 s 19370 19369 0 0 0 70 0 c-cc7cbe4c sh
64 s 8417 7347 7347 0 0 80 0 c-ebaa33c0 vi
65 s 19157 19033 19033 0 0 70 1 c-e7d2b8ec vi
67 s 8478 1556 1556 0 0 80 0 e-d1e54ebc sendmail
67 s 1 59 0 u-dd824bc0 sendmail
69 s 14301 1 14301 0 0 80 0 e-e85e187c statd
70 p 8490 19183 19033 0 0 59 0 crash
74 s 19033 19032 19033 0 0 70 1 c-cc1748ec sh
78 s 19183 19157 19033 0 0 70 0 c-d8a583ac sh

Again thanks!

On Jul 18, 7:49*am, SteveB wrote:
> On Jul 17, 5:33 pm, Jean-Pierre Radley wrote:
>
>
>
>
>
> > Pepe typed (on Thu, Jul 17, 2008 at 11:01:38PM +0200):
>
> > > SteveB wrote:
> > >> the response to the followin:
> > >> #ps -ef
> > >> returns
> > >> #ps -ef
> > >> * * *UID * PID *PPID *C * *STIME TTY * * * ** * *TIME CMD
>
> > >> with no process listed..
>
> > > So much for an industry-strength computer system like OpenServer 6...
>
> > If I run the same command on my OSR 6 with MP3, I get 149 lines of
> > output.
>
> > --
> > JP
>
> Well JP I'm not surprised. *Until yesterday it worked.
>
> With more poking around *with mount I have found that the /proc and /
> process filesystems are not mounting at all.
>
> I ahve tried the following:
> truss -o /tmp/mount.truss -aef mnt /proc
>
> and get an error:
> truss: /proc is not a PROC directory
>
> I have also tried using mount and mountall with basically the same
> result.
>
> If I do the following:
>
> echo proc | crash *get *followiing *(I've edit the output for brevity)
>
> *46 s *1726 * * 1 *1726 * * 0 * *0 * 80 * 1 *e-d10243cc *cupsd
> * 48 s *1731 *1709 *1709 * *28 * *0 * 80 * 1 *c-cdc66d50 *httpd
> * 49 s *1732 *1709 *1709 * *28 * *0 * 80 * 1 *c-cdc66d50 *httpd
> * 50 s *1733 *1709 *1709 * *28 * *0 * 80 * 1 *c-cdc66d50 *httpd
> * 52 s *1768 * * 1 *1768 * * 0 * *0 * 80 * 1 *e-d3025ebc *smbd
> * 52 s * * * * * * * * * * * * * *1 * 59 * 0 *u-ddac2000 *smbd
> * 54 s *1781 *1768 *1768 * * 0 * *0 * 70 * 1 *e-c04e7b24 *smbd
> * 54 s * * * * * * * * * * * * * *1 * 59 * 0 *u-d45e4200 *smbd
> * 55 s 19368 * * 1 * * 0 * * 0 * *0 * 70 * 0 *c-c81168cc *uadmin
> * 59 s 19370 19369 * * 0 * * 0 * *0 * 70 * 0 *c-cc7cbe4c *sh
> * 64 s *8417 *7347 *7347 * * 0 * *0 * 80 * 0 *c-ebaa33c0 *vi
> * 65 s 19157 19033 19033 * * 0 * *0 * 70 * 1 *c-e7d2b8ec *vi
> * 67 s *8478 *1556 *1556 * * 0 * *0 * 80 * 0 *e-d1e54ebc *sendmail
> * 67 s * * * * * * * * * * * * * *1 * 59 * 0 *u-dd824bc0 *sendmail
> * 69 s 14301 * * 1 14301 * * 0 * *0 * 80 * 0 *e-e85e187c *statd
> * 70 p *8490 19183 19033 * * 0 * *0 * 59 * 0 * * * * * * *crash
> * 74 s 19033 19032 19033 * * 0 * *0 * 70 * 1 *c-cc1748ec *sh
> * 78 s 19183 19157 19033 * * 0 * *0 * 70 * 0 *c-d8a583ac *sh
>
> Again thanks!- Hide quoted text -
>
> - Show quoted text -

Just to clarify - did this happen when you loaded MP3 (or any other
system change) or was it perceived as a random event with unknown
cause after a period of successful operation?

I also note in your first e-mail that you stated that /var/syslog was
empty. Could you please check /var/adm/syslog as there is no /var/
syslog on OSR6. Assuming you made a typo - when you say it is empty
what is the reported file size of /var/adm/syslog and does it show the
correct link to /var/opt/K/SCO/Unix/6.0.0Ni/usr/adm/syslog?

Regards

James

James_Szabadics typed (on Thu, Jul 17, 2008 at 06:35:11PM -0700):
| I also note in your first e-mail that you stated that /var/syslog was
| empty. Could you please check /var/adm/syslog as there is no /var/
| syslog on OSR6. Assuming you made a typo - when you say it is empty
| what is the reported file size of /var/adm/syslog and does it show the
| correct link to /var/opt/K/SCO/Unix/6.0.0Ni/usr/adm/syslog?

If /var/adm/syslog is NOT symlinked to any other file: so what?

--
JP

On Jul 17, 9:35 pm, James_Szabadics wrote:
> On Jul 18, 7:49 am, SteveB wrote:
>
>
>
> > On Jul 17, 5:33 pm, Jean-Pierre Radley wrote:
>
> > > Pepe typed (on Thu, Jul 17, 2008 at 11:01:38PM +0200):
>
> > > > SteveB wrote:
> > > >> the response to the followin:
> > > >> #ps -ef
> > > >> returns
> > > >> #ps -ef
> > > >> UID PID PPID C STIME TTY TIME CMD
>
> > > >> with no process listed..
>
> > > > So much for an industry-strength computer system like OpenServer 6...
>
> > > If I run the same command on my OSR 6 with MP3, I get 149 lines of
> > > output.
>
> > > --
> > > JP
>
> > Well JP I'm not surprised. Until yesterday it worked.
>
> > With more poking around with mount I have found that the /proc and /
> > process filesystems are not mounting at all.
>
> > I ahve tried the following:
> > truss -o /tmp/mount.truss -aef mnt /proc
>
> > and get an error:
> > truss: /proc is not a PROC directory
>
> > I have also tried using mount and mountall with basically the same
> > result.
>
> > If I do the following:
>
> > echo proc | crash get followiing (I've edit the output for brevity)
>
> > 46 s 1726 1 1726 0 0 80 1 e-d10243cc cupsd
> > 48 s 1731 1709 1709 28 0 80 1 c-cdc66d50 httpd
> > 49 s 1732 1709 1709 28 0 80 1 c-cdc66d50 httpd
> > 50 s 1733 1709 1709 28 0 80 1 c-cdc66d50 httpd
> > 52 s 1768 1 1768 0 0 80 1 e-d3025ebc smbd
> > 52 s 1 59 0 u-ddac2000 smbd
> > 54 s 1781 1768 1768 0 0 70 1 e-c04e7b24 smbd
> > 54 s 1 59 0 u-d45e4200 smbd
> > 55 s 19368 1 0 0 0 70 0 c-c81168cc uadmin
> > 59 s 19370 19369 0 0 0 70 0 c-cc7cbe4c sh
> > 64 s 8417 7347 7347 0 0 80 0 c-ebaa33c0 vi
> > 65 s 19157 19033 19033 0 0 70 1 c-e7d2b8ec vi
> > 67 s 8478 1556 1556 0 0 80 0 e-d1e54ebc sendmail
> > 67 s 1 59 0 u-dd824bc0 sendmail
> > 69 s 14301 1 14301 0 0 80 0 e-e85e187c statd
> > 70 p 8490 19183 19033 0 0 59 0 crash
> > 74 s 19033 19032 19033 0 0 70 1 c-cc1748ec sh
> > 78 s 19183 19157 19033 0 0 70 0 c-d8a583ac sh
>
> > Again thanks!- Hide quoted text -
>
> > - Show quoted text -
>
> Just to clarify - did this happen when you loaded MP3 (or any other
> system change) or was it perceived as a random event with unknown
> cause after a period of successful operation?
>
> I also note in your first e-mail that you stated that /var/syslog was
> empty. Could you please check /var/adm/syslog as there is no /var/
> syslog on OSR6. Assuming you made a typo - when you say it is empty
> what is the reported file size of /var/adm/syslog and does it show the
> correct link to /var/opt/K/SCO/Unix/6.0.0Ni/usr/adm/syslog?
>
> Regards
>
> James

Sylog is now working.

Though not in production, the server has been working for 6 months.

The "ps" problem started a few days ago.

James directed me to TA#109457. The /etc/vsftab is linked to /etc/
default/filesystems. But /proc & /prcessorfs will not mount. In
additions none of the filestems from the production server will mount
via nfs on the new server (which is the one I am having the problems
with).

On Jul 17, 10:29 pm, Jean-Pierre Radley wrote:
> James_Szabadics typed (on Thu, Jul 17, 2008 at 06:35:11PM -0700):
> | I also note in your first e-mail that you stated that /var/syslog was
> | empty. Could you please check /var/adm/syslog as there is no /var/
> | syslog on OSR6. Assuming you made a typo - when you say it is empty
> | what is the reported file size of /var/adm/syslog and does it show the
> | correct link to /var/opt/K/SCO/Unix/6.0.0Ni/usr/adm/syslog?
>
> If /var/adm/syslog is NOT symlinked to any other file: so what?
>
> --
> JP

JP
syslog is now performin properly. The problem was that syslogd was
not running, which I missed because I can not see the process table.

Thanks again

SteveB typed (on Fri, Jul 18, 2008 at 09:23:54AM -0700):
| The "ps" problem started a few days ago.
|
| James directed me to TA#109457. The /etc/vsftab is linked to /etc/
| default/filesystems. But /proc & /prcessorfs will not mount. In
| additions none of the filestems from the production server will mount
| via nfs on the new server (which is the one I am having the problems
| with).

Please be a little more accurate in your typing. Earlier you referred
to /var/syslog, when you really meant /var/adm/syslog.

Now you tell us /etc/vfstab is linked to /etc/default/filesystems, which
is highly unlikely; it ought to be linked to /etc/default/filesys.

TA 109457 discusses the Unixware variant of /etc/vfstab, but its OSR6
cousin is in fact relevant to your problem; if you read /etc/mountall,
you'll see that 'mountall' reads that file to know what to do. You
(inaccurately) reported vfstab's symlink, but told us nothing about its
contents.

So please now show us its contents. No typos, please import the file as
is into your email editor when you reply. :-)

--
JP