[McIDAS #WPR-272383]: DMRAOB - Program terminated, segmentation violation

Hi Mike,

re:
> I've noticed for the past couple of days McIDAS can't make hodographs for
> me.  When I look at the log from those scripts, the key line I see there is
> "NO WIND DATA AVAILABLE."
> 
> I loosely understand the process working here to supply this data to
> McIDAS, and I know it's happened here in the past where something about the
> XCD Monitor stops working and an LDM restart would bring that back
> online... I don't appear to be that lucky this time though.

I have never had the experience where I needed to restart the LDM
to get a McIDAS decoder/data monitor to run.

re:
> I see there's a XCD_START.LOG file, and looking there I see many segments
> like the few below.  I suppose my first question is am I correct in
> thinking this is related to not having wind data for plotting hodographs?

You are correct.  DMRAOB is the upper air data monitor which, in turn,
runs upper air decoding.

re:
> And then if that's the case how might have this happened, and what can I do
> about it?

I just checked all three of the systems where we run McIDAS
data decoding, and none of these machines are showing the problem
that you are seeing.

As for what you can do, I would suggest the following:

- rebuild and reinstall the McIDAS-XCD decoders

  This is done by doing the following as the user 'mcidas'
  (assuming that you setup your machine following the guidelines
  that we provide for building/installing/running McIDAS-X/-XCD):

  <as 'mcidas'>
  cd ~mcidas/mcidas2019/src
  make clobber
  make all
  make install.all

But: before doing this, you should probably try to figure out
why dmraob.k is crashing.  This would be done (again, assuming
your setup follows our recommendations):

<as 'ldm'>
gdb `which dmraob.k` core.xxx

where:

core.xxx is the core file produced when dmraob.k crashed

NB:

If you do not have a core file, you need to enable core
dumping otherwise you will not be able to fire out what
has gone wrong

Also:

It is possible that the upper air MD file(s) that the
decoding action that dmraob.k is trying to run might
be munged.  In this case, I would:

<as 'ldm'>
ldmadmin stop

-- find out where the output upper air files that are being
   written by the dmraob.k-run decoding process lives

   Delete the most recent files (or rename them to something
   other than MDXX0010 - MDXX0030 (MDXX0011 - MDXX0020 are
   mandatory level upper air files in the standard installation
   and MDXX0021 - MDXX030 are significant level upper air
   files in a standard installation).

ldmadmin start

re:
> Nothing's changed with any of the scripting or McIDAS install
> for some time, and I'm not all that familiar with the XCD stuff to do much
> more troubleshooting on my own.

The running of the upper air data monitor/decoder will have nothing to
do with the scripts you run to plot data.

re:
> startxcd.k: m0monexe - restart of :DMRAOB
> 70188
> startxcd.k: m0monexe - :DMRAOB                                   new PID=
> 6107
> DMRAOB       Starting: Upper Air (TEMP) Decoder
> Program terminated, segmentation violation
> startxcd.k: m0monexe - restart of :DMRAOB
> 6107
> startxcd.k: m0monexe - :DMRAOB                                   new PID=
> 15301
> DMRAOB       Starting: Upper Air (TEMP) Decoder
> Program terminated, segmentation violation
> startxcd.k: m0monexe - restart of :DMRAOB
> 15301
> startxcd.k: m0monexe - :DMRAOB                                   new PID=
> 22934
> DMRAOB       Starting: Upper Air (TEMP) Decoder
> Program terminated, segmentation violation

The question is why the upper air data monitor DMRAOB
(dmraob.k) is seg faulting.  This can only be figured out
by using a debugger.  If a bad output file is the cause of
the problem, then deleting the file(s) causing the problem
might well clear the problem without doing anything else
(modulo stopping and restarting the LDM while deleting the
file(s)).

re:
> Thanks for your help,

Questions:

- are you using a standard McIDAS installation?

- what OS are your XCD decoders running?

Comment:

- while you are troubleshooting the dmraob.k problem,
  you can point at our ADDE servers to get the data you
  need/want

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: WPR-272383
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.