wrf 3.8.1 real.exe issues with ERA5 data

[avipwrfhelp]

Hello,
I am having trouble running real.exe with ERA5 interim data.
All of the rsl.error.* files display the following message

taskid: 0 hostname: r4i1n7
module_io_quilt_old.F 2931 T
Ntasks in X 18 , ntasks in Y 24
--- WARNING: traj_opt is zero, but num_traj is not zero; setting num_traj to zero.
--- NOTE: both grid_sfdda and pxlsm_soil_nudge are 0 for domain 1, setting sgfdda interval and ending time to 0 for that domain.
--- NOTE: obs_nudge_opt is 0 for domain 1, setting obs nudging interval and ending time to 0 for that domain.
--- NOTE: bl_pbl_physics /= 4, implies mfshconv must be 0, resetting
--- NOTE: RRTMG radiation is in use, setting: levsiz=59, alevsiz=12, no_src_types=6
--- NOTE: num_soil_layers has been set to 4
REAL_EM V3.8.1 PREPROCESSOR
*************************************
Parent domain
ids,ide,jds,jde 1 1320 1 1220
ims,ime,jms,jme -4 81 -4 58
ips,ipe,jps,jpe 1 74 1 51
*************************************
DYNAMICS OPTION: Eulerian Mass Coordinate
alloc_space_field: domain 1 , 320861796 bytes allocated
metgrid input_wrf.F first_date_input = 2018-04-02_00:00:00
metgrid input_wrf.F first_date_nml = 2018-04-02_00:00:00
Timing for input 158 s.
flag_soil_layers read from met_em file is 1
Max map factor in domain 1 = 1.07. Scale the dt in the model accordingly.
Using sfcprs to compute psfc
flag value of flag_qnwfa is 0
No average surface temperature for use with inland lakes
Assume Noah LSM input

Only rsl.error.0017 displays the additional
:
Max map factor in domain 1 = 1.07. Scale the dt in the model accordingly.
Using sfcprs to compute psfc
flag value of flag_qnwfa is 0
No average surface temperature for use with inland lakes
Assume Noah LSM input
forcing artificial silty clay loam at 22 points, out of 3723
Timing for processing 26 s.
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 707
real: error opening wrfinput for writing
-------------------------------------------
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 17
In: PMI_Abort(1, application called MPI_Abort(MPI_COMM_WORLD, 1) - process 17)

Any ideas what could be causing this write error?

I had previously run real.exe from a later version (4.1.3) without any issues on the same filesystem.I am also attaching the namelist.input file for WRF 3.8.1.

Thank you -- Avi

 

[avipwrfhelp]

Some needed changes from the original namelist.input were made, such as
io_form_history = 2
io_form_restart = 2
io_form_input = 2
io_form_boundary = 2

These were needed since real.exe was built with netcdf and not pnetcdf. However the original unexplained issue of writing out the wrfinput file remains.

taskid: 0 hostname: r2i3n13
module_io_quilt_old.F 2931 T
Ntasks in X 18 , ntasks in Y 24
--- WARNING: traj_opt is zero, but num_traj is not zero; setting num_traj to zero.
--- NOTE: both grid_sfdda and pxlsm_soil_nudge are 0 for domain 1, setting sgfdda interval and ending time to 0 for that domain.
--- NOTE: obs_nudge_opt is 0 for domain 1, setting obs nudging interval and ending time to 0 for that domain.
--- NOTE: bl_pbl_physics /= 4, implies mfshconv must be 0, resetting
--- NOTE: RRTMG radiation is in use, setting: levsiz=59, alevsiz=12, no_src_types=6
--- NOTE: num_soil_layers has been set to 4
REAL_EM V3.8.1 PREPROCESSOR
*************************************
Parent domain
ids,ide,jds,jde 1 1320 1 1220
ims,ime,jms,jme -4 81 -4 58
ips,ipe,jps,jpe 1 74 1 51
*************************************
DYNAMICS OPTION: Eulerian Mass Coordinate
alloc_space_field: domain 1 , 320861796 bytes allocated
metgrid input_wrf.F first_date_input = 2018-04-02_00:00:00
metgrid input_wrf.F first_date_nml = 2018-04-02_00:00:00
Timing for input 43 s.
flag_soil_layers read from met_em file is 1
Max map factor in domain 1 = 1.07. Scale the dt in the model accordingly.
Using sfcprs to compute psfc
flag value of flag_qnwfa is 0
No average surface temperature for use with inland lakes
Assume Noah LSM input
Timing for processing 0 s.
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 707
real: error opening wrfinput for writing
-------------------------------------------

 

[Ming Chen]

is there any special reason you want to run WRFv3.8.1? Can you update to newer WRF version like WRFV4.3? It is not our top priority to debug any possible problems in older version of WRF (older than V4.0).

 

[avipwrfhelp]

I'm doing some tests on A100 GPUs and their current working version is 3.8.1 hence the need for that.

 

[Ming Chen]

I am suspicious this might be a library issue. I cannot repeat your error in our Mac and the super computer Cheyenne, which makes it hard for us to figure out exactly what is wring. Please take a look at this post and hopefully it could help:
https://forum.mmm.ucar.edu/phpBB3/viewtopic.php?p=18108#p18108