This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Hi Randy, re: > in /data/conduit/GEFS we now have about 1887 individual grib files. I > created a gribm.csh which runs wgrib2 on each file and piped the output > to gribout.txt in the same directory. There were no ERRORS found in this > file. VERY interesting! > I also wgrib2 the GFS_Global_1p0deg_Ensemble_12_2009022400.grib2 file > located in /data/pub/native/grid/NCEP/GFS/Global_1p0deg_Ensemble/. > > I get the *** FATAL ERROR: rd_grib2_msg, missing end section ('7777') *** > after > just a 126 records. > > Does this mean there is an issue with cating the files together? Perhaps. It could also mean that 'wgrib2' has a problem that manifests itself when looking through all of the GRIB2 messages in the monolithic output file. It could also mean that there is some problem related to the LDM or OS that results in one ore more GRIB2 messages being stepped on in the process of writing them to the end of the monolithic output file. Here is what I would do to test this: - get a full listing of the files that got put into one of the monolithic output files ** in the same order that they were written by the LDM into the file created by the FILE action ** - create a new monolithic output file by 'cat'ting each GRIB2 file to the end ** in the exact same order as in the file created by the LDM FILE action ** - run 'wgrib2' on the newly created output file If 'wgrib2' does not show an error for the new file you create, then there is some problem with the LDM appending new GRIB2 messages to the end of the output file. If this is the case, I would add the '-flush' flag to the FILE action: CONDUIT data/nccf/com/gens/prod/gefs\.(........)/(..)/pgrb2a FILE -flush /data/pub/native/grid/NCEP/GFS/Global_1p0deg_Ensemble/GFS_Global_1p0deg_Ensemble_\2_\100.grib2 By the way, I just noticed that the copy of this action that you included in a previous email contained the '-close' flag: CONDUIT data/nccf/com/gens/prod/gefs\.(........)/(..)/pgrb2a FILE -close /data/pub/native/grid/NCEP/GFS/Global_1p0deg_Ensemble/GFS_Global_1p0deg_Ensemble_\2_\100.grib2 I would remove the '-close' flag! > I did notice that some individual files are named: > UREL;VREL_925Pa_20090224_1200_378.grib2 > Because of the ; in the file name one must put quotes "" around the > file name in order for wgrib2 to work correctly. I dont think this is the > issue but wanted to throw that out. > > comments? I was a bit concerned about the product IDs that contained the 'UREL;VREL' field as the semi-colon has a special meaning for the *nix shell. I ran a test that proved to myself that one can create a valid output file name, and I stopped worrying. I am glad you recognized that you had to quote the file name to get 'wgrib2' to work. You are correct in your assumption that this should _not_ be a problem. 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: WFA-619667 Department: Support IDD Priority: Normal Status: Closed