Open FANG-FZJ opened 4 months ago
Hi, even though it is hard to get the full picture without the complete error-log, I would estimate based on your description that there is not enough time dedicated to the job (UGRID_1km-merge-10min_HYDRO1K-merge-nomask_c130402.nc is the last file, which is processed).
Could you please try to enhance the wall-time, e.g. by increasing the values in this line and run again?
Hi, I actually have tried to increase the wall clock time to 30min, but the job failed after 3min. Please find attached the error report. Also even when I tried to process only the UGRID_1km-merge-10min_HYDRO1K-merge-nomask_c130402.nc file, it also failed in very short time, without generating the corresponding file.
When running the runscript_mkmapdata.sh, there is always an error for the file UGRID_1km-merge-10min_HYDRO1K-merge-nomask_c130402.nc to execute the command: srun $ESMFBIN_PATH/./ESMF_RegridWeightGen --ignore_unmapped -s ${rawpath}/${file} -d $GRIDFILE -m conserve -w ${OUTPUT}/${OUTFILE} --dst_regional --netcdf4. The report from the jobscript showed that no map_UGRID_1km-merge-10min_HYDRO1K-merge-nomask_c130402.ncto${GRIDNAME}_nomask_aave_da_c${CDATE}.nc file was generated.
An alternative way to avoid this error is to run the commands from runscript_mkmapdata.sh one by one on the JSC Linux terminal. This may take longer than submitting the job to JSC.