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

[McIDAS #TAL-270196]: Not getting MD data



Hi Heather,

re:
> Tom, I just got your email this morning ...

OK.

re:
> I believe I understand the problem as you have explained it, thank you.

No worries.  The situation is a bit hard to explain...

re:
> I implemented the changes this morning, and of course everything is
> working just fine now.  I will let you know tomorrow morning if the
> xcd decoders were able to make it through to the next day!

You don't have to wait to see if things will keep working.  You can
test the change you made by:

<as 'ldm'>
ldmadmin newlog

The following is "for the files":

If you built your LDM-6.13.x distribution without specifying that
you want to use the system logging daemon (this would be done by
specifying an option to 'configure'), then the script

~ldm/bin/refresh_logging

will be run, and it sends a HUP signal to the LDM process leader
(using the negative of the process leader's PID).  This signal
will be sent to all of the processes running in the same process
group as the process leader, and this will include everything run
via an EXEC line in the LDM configuration file, ~ldm/etc/ldmd.conf.
It will be sent to all 'pqact' invocations running from EXEC lines,
but it will _not_ be passed to the processes that those 'pqact's
run as they are put into a different process group.  What was
happening is the McIDAS-XCD startup script, 'xcd_run' gets the
HUP signal and passes it along to the McIDAS-XCD startup routine,
'startxcd.k'.  'startxcd.k' does _not_ do anything to alter its
action upon receipt of a signal, so it imply exits.  'startxcd.k's
job is to start McIDAS-XCD and keep running all XCD data monitors
that are specified in the McIDAS-XCD file DECINFO.DAT (this file
lives in the ~mcidas/workdata directory by default).  All of the
data monitors started by 'startxcd.k' inherit signals received
by 'startxcd.k', so they exit also.  At this point, most of XCD
is no longer running.  The exceptions to this are 'ingetext.k'
and 'ingebin.k' which do not exit since they were started in
a process group separate from the one that the main LDM routines
run in by a 'pqact' invocation.  The addition of the 'nohup'
invocation at the beginning of the McIDAS-XCD EXEC line in
~ldm/etc/ldmd.conf prevents the HUP signal from being propagated
to the routine run ('xcd_run' and then 'startxcd.k').

I recommend that you run 'ldmadmin newlog' to test your fix
when it is convenient but before your cron entry that does the
same runs.

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: TAL-270196
Department: Support McIDAS
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.