Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

metgrid.exe not created when compiling WPS

I also had this issue in WPS v4.5 while compiling with gcc/gfortran and the solution proposed by mgduda solved the problem, many thanks!

Sagar
 
mgduda said:
While both geogrid and metgrid make use of WRF's I/O API to read and write NetCDF files, the metgrid program is special in that it includes additional code for reading NetCDF files (the primary use case being the ingest of MPAS-Atmosphere output). It's purely a guess, but it may be that one of the 'nf90_' routines that are called in the metgrid/src/scan_input.F module ultimately references 'nc_set_var_chunk_cache_ints'.

@Doulgas_secchi Would you be willing to try recompiling the metgrid program after overwriting the default WPS/metgrid/src/scan_input.F file with the attached version, which comments-out calls to NetCDF Fortran90 interface routines? If compilation of the metgrid.exe program is successful, you should be able to use the program as normal, as long as you don't plan on using MPAS-Atmosphere output for WRF ICs and LBCs. The modified scan_input.F file is of course a work-around, but knowing whether it works will help in narrowing down the problem and ultimately finding a long-term solution.
Click to expand...
Hello
i was having the same issue while compiling WPS V4.5 but after trying your scan_input.F i managed to get metgrid.exe
i'm compiling with Linux x86_64, gfortran (dmpar)
thank you
 
JeremyB said:
Thank you @mgduda. The file you sent also worked for me and I compiled with ifx and icc (new intel compilers).

Thanks,
JeremyB
Click to expand...
Just a solution i put to the github for review and pull request


Code: 
########################################################################################################################
#ARCH    Linux x86_64, Intel oneAPI compiler with icx   # serial serial_NO_GRIB2 dmpar dmpar_NO_GRIB2
#
COMPRESSION_LIBS    = CONFIGURE_COMP_L
COMPRESSION_INC     = CONFIGURE_COMP_I
FDEFS               = CONFIGURE_FDEFS
SFC                 = ifx
SCC                 = icx
DM_FC               = mpif90 -f90=$(SFC)
DM_CC               = mpicc -cc=$(SCC)
FC                  = CONFIGURE_FC
CC                  = CONFIGURE_CC
LD                  = $(FC)
FFLAGS              = -FR -convert big_endian
F77FLAGS            = -FI -convert big_endian
FCSUFFIX            =
FNGFLAGS            = $(FFLAGS)
LDFLAGS             =
CFLAGS              = -w -Wno-implicit-int
CPP                 = /lib/cpp -P -traditional
CPPFLAGS            = -D_UNDERSCORE -DBYTESWAP -DLINUX -DIO_NETCDF -DIO_BINARY -DIO_GRIB1 -DBIT32 CONFIGURE_MPI
ARFLAGS             =
CC_TOOLS            =
 
mgduda said:
@Doulgas_secchi The "scan_input.f90" file is a processed version of "scan_input.F". If you modify just scan_input.f90, clean, and re-compile, that scan_input.f90 file will be overwritten as you've found. Also, if you just modify scan_input.f90 and compile metgrid a second time, that scan_input.f90 file will not be recompiled; it's modifications to scan_input.F that will trigger a recompilation of that file.

Just to make sure we're running the intended test, can you do the following in this order specifically:

1. ./clean -a
2. ./configure
3. Copy the "scan_input.F.txt" that I previously attached to "WPS/metgrid/src/scan_input.F"
4. ./compile

After compiling, you should see that WPS/metgrid/src/scan_input.f90 contains the "!" comment character at the beginning of any line referencing a NetCDF Fortran90 interface routine.

Apologies for the trouble -- I just want to make sure that the modified scan_input.F is being used as intended.
Click to expand...
Yeah, it worked.
Thanks
 
mgduda said:
While both geogrid and metgrid make use of WRF's I/O API to read and write NetCDF files, the metgrid program is special in that it includes additional code for reading NetCDF files (the primary use case being the ingest of MPAS-Atmosphere output). It's purely a guess, but it may be that one of the 'nf90_' routines that are called in the metgrid/src/scan_input.F module ultimately references 'nc_set_var_chunk_cache_ints'.

@Doulgas_secchi Would you be willing to try recompiling the metgrid program after overwriting the default WPS/metgrid/src/scan_input.F file with the attached version, which comments-out calls to NetCDF Fortran90 interface routines? If compilation of the metgrid.exe program is successful, you should be able to use the program as normal, as long as you don't plan on using MPAS-Atmosphere output for WRF ICs and LBCs. The modified scan_input.F file is of course a work-around, but knowing whether it works will help in narrowing down the problem and ultimately finding a long-term solution.
Click to expand...
Hi, thanks a lot. I have replaced the .F file and WPS compile work, and metgrid.exe be created. But, another problem, I can not 'mpirun -np 4 ./real.exe', it stucked without any error, and 'mpirun -np 1 ./real.exe' work. And I have posted as '

Mpirun stuck with no errors or interuptions​

'
So, please.
 
David Jones said:
Hi, thanks a lot. I have replaced the .F file and WPS compile work, and metgrid.exe be created. But, another problem, I can not 'mpirun -np 4 ./real.exe', it stucked without any error, and 'mpirun -np 1 ./real.exe' work. And I have posted as '

Mpirun stuck with no errors or interuptions​

'
So, please.
Click to expand...
since this is a new issue can you please open a new topic on the forum in the wps section? NCAR likes to keep the issues seperate
 
You must log in or register to reply here.
Top