[McIDAS #BYN-431727]: LDM Dying

Hi Martha,

re:
> I did copy xcd_run to the decoders subdirectory of the ldm home directory.  
> And no, it is not on
> an automount file system  that suddenly goes offline.

OK.  Why an exec of xcd_run would fail sometimes is beyond me.  I asked about 
automounted
file systems since that was the only thing I could conjure up that might cause 
the
results you reported.

re:
> I am at home right now, but was able to get on via our VPN and print out some 
> settings for
> you to see that I am attaching.

OK.  What is not in your email is the ~ldm/etc/ldmd.conf file (so I can review 
your
'exec' line for 'xcd_run'.

re:
> As I say, the system ran successfully from last Thursday till today, then it 
> suddenly died.
> xcd_run is found, then suddenly it isn't.

Yup, this one is a puzzler.  Since the LDM simply inherits the environment 
in-scope
when it starts, it should exit because 'xcd_run' is "suddenly not found".

Hmm... perhaps something being run from within 'xcd_run' is suddenly not
available?  This would have the effect of 'xcd_run' exiting with error which,
in turn, would cause the LDM to exit.

re:
> Also I made sure I set
> 
> kernel.shmall = 536870912
> kernel.shmmax = 536870912
> net.ipv4.ipfrag_max_dist = 0

None of these should make any difference for LDM execution/exiting.

Again musing...

The error message:

Aug 10 16:30:38 noaapxcd pqact[19201] ERROR: Couldn't exec(xcd_run): No such 
file or directory

is very strange since it says that it is an error from pqact.  'xcd_run' should
be started from an EXEC line from ~ldm/etc/ldmd.conf, not from an entry in
a pattern-action file.  The typical 'xcd_run' invocation line will look like:

exec    "xcd_run MONITOR"

Do you have this entry in your ~ldm/etc/ldmd.conf file?

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: BYN-431727
Department: Support McIDAS
Priority: Normal
Status: Closed