[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 20000626: steps in the binary installation of the LDM (fwd)

"Patrick S. Market" wrote:

> Anne:
>
> I double checked my /etc/syslog.conf, and the only error I could discern was
> a <space> in place of a <tab>.  I went over the other steps in the section
> you mentioned, and they seemed to be right on also.
>
> I had stopped the LDM, and so restarted it again per your instructions
> from last week.  Apparently, the LDM tried starting up for about 38 minutes
> before giving up the ghost:
>
> I am LDM =>ldmadmin clean
> I am LDM =>ldmadmin delqueue
> Jul 3 18:01:56 UTC bergeron.snr.missouri.edu : delete_pq: 
> /usr/local/ldm/data/ldm.pq does not exist
> I am LDM =>ldmadmin mkqueue
> I am LDM =>dlmadmin start
> dlmadmin: Command not found
> I am LDM =>ldmadmin start
> starting the LDM server...
> Jul 3 18:39:08 UTC bergeron.snr.missouri.edu : start_ldm: Server not started 
> or registered.
> I am LDM =>
>
> Nothing was written to the ldmd.log file:
>
> I am LDM =>ls -ltr /var/data/ldm/logs
> total 0
> -rw-r--r--   1 ldm             0 Jun 30 07:56 ldmd.log
> I am LDM =>
>
> But some processes continue to run, apparently:
>
> I am LDM =>!!
> ps -aux | grep ldm
> ldm       7898  0.1  0.2  944  704 pts/3    S 08:27:23  0:00 grep ldm
> ldm       6152  0.0  0.3 1416 1152 pts/8    S   Jul 03  0:00 -csh
> root      6245  0.0  0.3 1944 1104 ?        S   Jul 03  0:00 rpc.ldmd -q 
> /usr/l
> ldm       6247  0.0  0.3101768 1560 ?        S   Jul 03  0:00 pqbinstats
> ldm       6248  0.0  0.4102280 2072 ?        S   Jul 03  0:00 pqact
> ldm       6322  0.0  0.3 1416 1128 pts/3    S   Jul 03  0:00 -csh
> I am LDM =>
>
> I will go and recheck my steps in the "Configuring the Operating System as 
> root"
> section, but I'm fairly sure they're correct.  Maybe a day away will have
> cleared my head.
>
> Thanks.
>
> Pat
>
> --
> ========================================================================
> Dr. Patrick S. Market
> Dept. of Soil & Atmospheric Sciences
> University of Missouri-Columbia                 Phone:  (573) 882 - 1496
> 203 Gentry Hall                                 Fax:    (573) 884 - 5133
> Columbia, MO 65211 USA          E-mail: address@hidden
> ========================================================================

Good Morning Pat,

I still suspect that there's an error in one of your conf files, but we really 
need to have logging
working to confirm that.   And, of course, you will need the logs down the 
road.  Let's focus on
getting the logging working.

1. Make sure that hupsyslog are set SUID root:
    (anne) imogene:/local/ldm/bin 127 % ls -l hupsyslog rpc.ldmd
    -rwsr-xr-x    1 root     ustaff      24041 Jun 29 10:18 hupsyslog*
    -rwsr-xr-x    1 root     ustaff     428774 Jun 29 10:21 rpc.ldmd*
If not, set them appropriately:
    cd ~ldm/bin
    su            <<--- become root
    chown root rpc.ldmd
    chown root hupsyslog
    chmod 4775 rpc.ldmd
    chmod 4775 hupsyslog

2. Check that the link to the logs directory is pointing to the right place.  
Also, check that the
directory is writable by ldm:
    (anne) imogene:/local/ldm 133 % cd ~ldm
    (anne) imogene:/local/ldm 134 % ls -l logs
    lrwxrwxrwx    1 ldm      ustaff         22 Jun  5 13:44 logs -> 
/imogene/data/ldm/logs

3. Check that log file is writable by ldm and readable by at least ldm:
    (anne) imogene:/local/ldm/logs 137 % cd ~ldm/logs
    (anne) imogene:/local/ldm/logs 138 % ls -l ldmd.*
    -rw-r--r--    1 ldm      ustaff      28654 Jun 30 10:36 ldmd.log
    -rw-r--r--    1 ldm      ustaff       1508 Jun 29 11:04 ldmd.log.1
    -rw-r--r--    1 ldm      ustaff        432 Jun 29 10:52 ldmd.log.2
    -rw-r--r--    1 ldm      ustaff        432 Jun 29 10:48 ldmd.log.3
    -rw-rw-r--    1 ldm      ustaff       1507 Mar 29 14:55 ldmd.log.4
(I don't know why my ldmd.log.4 is group writable...??)

4. See if 'hupsyslog' is successful:
    cd~ldm/bin
    ./hupsyslog
If it returns with no error, it probably succeeded.  Confirm this by looking at 
the system log (and
see item #6 below):
   sudo tail /var/log/messages
This should produce an entry in the system log similar to:
    Jul  5 10:44:43 imogene syslogd 1.3-3: restart.

5. As ldm, try this 'foreach' command at the OS command line:
      foreach lev (err warning notice info)
     echo $lev | logger -i -t test -p local0.$lev
     end
This should add something like the following to the end of your ldmd.log file:
    Jul  5 10:25:57 imogene test[20413]: err
    Jul  5 10:25:57 imogene test[20415]: warning
    Jul  5 10:25:57 imogene test[20417]: notice
    Jul  5 10:25:57 imogene test[20419]: info

6. Could it be that the ldm messages are being sent to the system log?   My 
system log is in
/var/log/messages.   You could grep for some ldm messages from that file.  
(Note, you may have to be
root to view this file, hence the 'sudo'):
  sudo grep rpc.ldmd /var/log/messages
Also, grep for other ldm processes: pqact, pqexpire, pqbinstats, actually "pq*" 
should do it.

Try these and let me know what happens.

Anne

--
***************************************************
Anne Wilson                     UCAR Unidata Program
address@hidden                  P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************