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

Re: 20030818: "Interrupt" in dcgrib2 log



Steve,

On Thu, 28 Aug 2003, Unidata Support wrote:

> Art,
>
> The SIGINT signal is being caught in $GEMPAK/source/bridge/dc/dcsgnl.c
> which causes the exit.
>
> I have found 1 possible cause, being a bulletin that did not have the
> trailing 7777 (signalling the end of the grib) and therefore the decoder
> kept cycling through the input buffer.....when I believe it should
> stop at the specified length of the data stated in the PDS block
> of the decoder.
>
> I have added some debugging to try to catch an example of the
> grib product that is missing the correct 7777, since asside from
> handling the erroneous condition in dcgrib2, NCEP/NWS should
> fix the problem if it exists.
>
> Do you know the product that you missed?

Not at the moment.  After further investigating, I realized that the
gempak logs don't write time stamps in UTC but, instead, local time, so
the coincidence of the missing data and the INTERUPT was apparently not
true.  However, I'm still curious what caused the INTERUPT;  as you say,
it may be an irrelevent 7777 problem.

We've still seen a case-or-two where one of our systems decoded data that
another one didn't using identical streams.  It's very intermittent.  I
will be away next week, but after I get back, I'm going to do more
extensive testing to try and pin down what's going on with more specific
information if I can.

                                   Art.


> >From: "Arthur A. Person" <address@hidden>
> >Organization: UCAR/Unidata
> >Keywords: 200308181924.h7IJOQLd021914
>
> >Hi...
> >
> >I was wondering if you know what "Interrupt" in a dcgrib2 log might mean.
> >I got one *approximately* around the time that I lost a grid field that
> >should have been decoded in a gempak output file for the ETA model.  The
> >lines nearby in the log are:
> >
> >[23085] 030818/1244 [DM 0]
> >[23085] 030818/1249 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1249 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1249 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1249 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1249 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1250 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1250 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1250 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[23085] 030818/1250 [DCGRIB -50] griblen [1234594] > maxgribsize [1000000]
> >[23085] 030818/1319 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1320 [DC 3]
> >[13628] 030818/1320 [DC -11]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1329 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1330 [DCGRIB -50] griblen [1234594] > maxgribsize [1000000]
> >[13628] 030818/1354 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1354 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1359 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1402 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1408 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[13628] 030818/1414 [DC 2] Interrupt Signal
> >[13628] 030818/1414 [DC 5]
> >[13628] 030818/1414 [DC 2] Number of bulletins read and processed: 9330
> >[13628] 030818/1414 [DC 6]
> >[19966] 030818/1415 [DC 3]
> >[19966] 030818/1415 [DC -11]
> >[19966] 030818/1422 [DCGRIB -55] Grid too large 1830 x 918 [1679940 >
> >400000]
> >[19966] 030818/1427 [DM 0]
> >
> >
> >I'm running LDM V6.0.14 on RH9 with a dcgrib2 built on RH 7.2 (gempak
> >V5.6.h) for this.  The decoded data is being filed over NFS.  The decoder
> >seems to have run great except for this one field that was lost which has
> >made me a little nervous, so I'm trying to track it down.  I'm comparing
> >output from this system with parallel ingest/decoding on another system
> >here locally to establish correctness.
> >
> >                               Thanks.
> >
> >                                 Art.
> >
> >Arthur A. Person
> >Research Assistant, System Administrator
> >Penn State Department of Meteorology
> >email:  address@hidden, phone:  814-863-1563
> >
>
> **************************************************************************** <
> Unidata User Support                                    UCAR Unidata Program <
> (303)497-8643                                                  P.O. Box 3000 <
> address@hidden                                   Boulder, CO 80307 <
> ---------------------------------------------------------------------------- <
> Unidata WWW Service              http://my.unidata.ucar.edu/content/support  <
> **************************************************************************** <
>

Arthur A. Person
Research Assistant, System Administrator
Penn State Department of Meteorology
email:  address@hidden, phone:  814-863-1563