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

[McIDAS #EIJ-330346]: McIDAS GRID file problem

Hi Mark,

re:
> I rebuild mcidas 2006 according to your recommendation and it does seem to
> have eliminated the problem with strange names for our GRID files.

I wish I could tell you that I understand why rebuilding works, but I can't.

> The build
> went smoothly, there were no errors and all of the tests were successful. We
> still have a problem which is likely related.  dmgrid.k hangs very frequently
> consuming any free cycles on the cpu.  It also ceases to process any GRID
> data once this happens.  We have not been able to get more than one model run
> processed before it hangs and more often it does not stay running that long.

This is the exact same problem that I referred to in an earlier email.  I have
not yet been able to deduce what the problem is.  One of the problems in
troubleshooting this problem is McIDAS programs turn off dumping of core
files by default.  In order to see where the program is hung, we need a
core file.  This will require either running dmgrid.k manually from a
session in which core dumping is turned on, or modifying the source code
to turn the ability to dump a core file back on.

> Here is what I'm seeing in XCD_START.LOG:
> 
> ---------------------------------------------
> DMMISC       Starting: NAMMOS Decoder
> DMMISC       Starting: GFSMOS Decoder
> DMMISC       Starting: PIREP/AIREP Decoder
> DMMISC       Starting: Terminal Forecast Decoder
> DMMISC: TIRDEC       is labeled inactive:  TIROS Navigation Decoder
> DMMISC: MDRDEC       is labeled inactive:  MDR Grid Decoder
> Reading /software/mcidas/workdata/RTMODELS.CFG...
> ingebin.k: Done
> Starting HRS at 06291.164536
> High Resolution Data Service
> Program terminated, segmentation violation
> DMGRID in free(): warning: junk pointer, too high to make sense
> DMGRID in free(): warning: junk pointer, too high to make sense
> DMGRID in free(): warning: modified (chunk-) pointer
> -------------------------------------------------------------------------

I suspect that there is a pointer that has been incorrectly sized,
or there is a product that arrives in the afternoon that is being
unpacked incorrectly (the hands I see are always in the afternoon).

> We are running FreeBSD 4.10

We are also.  When I find the problem (which does not occur on other
OSes), I will create an Addendum with the fix.

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: EIJ-330346
Department: Support McIDAS
Priority: Normal
Status: Open