pio_support::pio_die:: myrank=-1 : ERROR: ionf_mod.F90: 235 : Feature is not yet supported. NC_FORMAT_NETCDF4 is not yet supported

[tobeawarrior]

Recently, I run a case with modification on POP SourceMods - forcing_coupled.F90

allocate(Hosing_fw_flux(nx_block,ny_block,max_blocks_clinic))

forcing_file = construct_file('nc', &
full_name=trim(hosing_nc_file), &
record_length = rec_type_dbl, &
recl_words=nx_global*ny_global)
call data_set(forcing_file,'open_read')

i_dim = construct_io_dim('i',nx_global)
j_dim = construct_io_dim('j',nx_global)

io_hosing = construct_io_field( &
trim('PREC_F'), &
dim1=i_dim, dim2=j_dim, &
field_loc = field_loc_center, &
field_type = field_type_scalar, &
d2d_array = Hosing_fw_flux)

call data_set(forcing_file,'define',io_hosing)
call data_set(forcing_file,'read' ,io_hosing)
call destroy_io_field(io_hosing)

call data_set(forcing_file,'close')
call destroy_file(forcing_file)

if (my_task == master_task) then
write(stdout,*) ' Hosing experiment:'
write(stdout,blank_fmt)
write(stdout,hosing_nml)
write(stdout,blank_fmt)
endif

However, the case crushed like the attach file - cesm.log. There is no waring in other log file.
If I do not modify this, the case can run normally.
What's more, I can run this case in another HPC normally with same modification, However, the remain memory of another HPC is almost zero.

I feel very doubtful, why other nc file from Inputdata folder can be read into the case normally? How to solve this problem?
Anthor idea ---- Is it OK, if I change this nc file to bin file?

 

[tobeawarrior]

I also doubt about how to change the nc file to bin file