BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES

Lumoss · Dec 18, 2023

jedwards said:
Those are pretty old requirements - I'll try to get that web page updated.
python >= 3.7
netcdf-c 4.9.2, netcdf-fortran 4.6.1
hdf5 1.12.3 or newer
esmf 8.6.0

Thank you for your reply. I will try to port CESM using these versions, and are there any requirements for the compiler? I am currently using Intel 2021.1 and Intel MPI 2021.1.

jedwards · Dec 19, 2023

The intel compiler is the best choice, but newer versions are available and since they are free perhaps you should consider updating.

Lumoss · Dec 19, 2023

jedwards said:
The intel compiler is the best choice, but newer versions are available and since they are free perhaps you should consider updating.

Thank you again for your reply and suggestions. My machine is running a slightly older version of CentOS7, and installing the latest version of the intel compiler may be problematic. Maybe I need to think about updating, which may take a lot of time. I hope this will help me with my problem. Thanks again.

Lumoss · Dec 19, 2023

Lumoss said:
Thank you again for your reply and suggestions. My machine is running a slightly older version of CentOS7, and installing the latest version of the intel compiler may be problematic. Maybe I need to think about updating, which may take a lot of time. I hope this will help me with my problem. Thanks again.

I found that the latest version of intel compiler Intel 2024.0 no longer supports icc and uses icx, will this affect the later installation?
Transitioning to ifx

jedwards · Dec 20, 2023

We have found the icx compiler to be as good or better than icc and recommend it.
The ifx compiler works but still lags ifort a little with respect to performance and we recommend staying with ifort for now.

Lumoss · Dec 20, 2023

Thank you for your continued help. After careful consideration, I have decided to utilize oneapi2022.3, which is also compatible with icx. However, I encountered difficulties while compiling cmake. Perhaps you could provide me with some guidance. The log is enclosed

jedwards · Dec 20, 2023

It's not required to use the same compiler to build cmake as you are using to build the model - you probably already have a working copy of cmake available to use - if not, perhaps try using the gnu compiler for cmake.

Lumoss · Dec 20, 2023

jedwards said:
It's not required to use the same compiler to build cmake as you are using to build the model - you probably already have a working copy of cmake available to use - if not, perhaps try using the gnu compiler for cmake.

Aha, I remember, I used to compile with gnu, so I'll use the previous version, thank you!

Lumoss · Dec 20, 2023

One more question, what have mpiicc, mpiicpc, and mpiifort become in the new oneapi?

Lumoss · Dec 21, 2023

I am compiling hdf5-1.14.3 using oneapi2022.3. There will be a warning when make:

Bash:

icx: warning: -Wl,-s: 'linker' input unused [-Wunused-command-line-argument]

Does that make a difference?
In addition, when make check, there will be an error:

Bash:

Testing: t_2Gio
============================
Test log for t_2Gio
============================
===================================
2 GByte IO TESTS START
2 MPI ranks will run the tests...
===================================
MPI-process 0. hostname=login
Test filenames are:
    ParaTest.h5
    Hugefile.h5


For help use: /public/home/LIBRARIES/intel2022/hdf5-1.14.3/testpar/.libs/t_2Gio -help
Linked with hdf5 version 1.14 release 3
Testing  -- dataset independent write (idsetw)
Testing  -- dataset independent read (idsetr)
Testing  -- dataset collective write (cdsetw)
Testing  -- dataset collective read (cdsetr)
Testing  -- extendible dataset independent write #2 (eidsetw2)
Testing  -- chunked dataset with none-selection (selnone)


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 0 PID 207616 RUNNING AT login
=   KILLED BY SIGNAL: 14 (Alarm clock)
===================================================================================


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 1 PID 207617 RUNNING AT login
=   KILLED BY SIGNAL: 14 (Alarm clock)
===================================================================================


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 2 PID 207618 RUNNING AT login
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 3 PID 207619 RUNNING AT login
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 4 PID 207620 RUNNING AT login
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================


===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 5 PID 207621 RUNNING AT login
=   KILLED BY SIGNAL: 9 (Killed)
===================================================================================
2469.66user 3954.87system 20:07.50elapsed 532%CPU (0avgtext+0avgdata 2695456maxresident)k
0inputs+0outputs (1159major+2047482minor)pagefaults 0swaps
make[4]: *** [t_2Gio.chkexe_] Error 1
make[4]: Leaving directory `/public/home/LIBRARIES/intel2022/hdf5-1.14.3/testpar'
make[3]: *** [build-check-p] Error 1
make[3]: Leaving directory `/public/home/LIBRARIES/intel2022/hdf5-1.14.3/testpar'
make[2]: *** [test] Error 2
make[2]: Leaving directory `/public/home/LIBRARIES/intel2022/hdf5-1.14.3/testpar'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory `/public/home/LIBRARIES/intel2022/hdf5-1.14.3/testpar'
make: *** [check-recursive] Error 1

It looks like the same error as before when running CESM.
My steps for building hdf5:

Bash:

export CC="mpiicc -cc=icx"
export CXX="mpiicpc -cxx=icpx"
export FC=mpiifort
export F90=mpiifort
export F77=mpiifort
export CPPFLAGS=-I/public/home/compiler/intel2022/oneapi/mpi/2021.7.0/include
export LDFLAGS="-L/public/home/compiler/intel2022/oneapi/mpi/2021.7.0/lib"
export CFLAGS="-O3 -fPIC"
export FCFLAGS="-O3 -fPIC"
export CXXFLAGS="-O3 -fPIC"
./configure --with-zlib=$DIR/zlib1.3 --with-szlib=$DIR/szip2.1.1 --prefix=$DIR/hdf51.14.3 --enable-fortran --enable-parallel
make
make check

Perhaps you could provide me with some guidance. Thanks again!

jedwards · Dec 21, 2023

Sorry - I can't offer support for hdf5.

Lumoss · Dec 21, 2023

jedwards said:
Sorry - I can't offer support for hdf5.

Thank you again for your reply. I will go to the hdf5 forum to ask my questions. Thanks again!

Lumoss · Dec 24, 2023

Hi,
When I run./case.build, pio reports the following error:

Bash:

cat: Filepath: No such file or directory
cat: Srcfiles: No such file or directory
/bin/sh: line 1: -cc=icx: command not found
make: *** [/public/home/cesm2_3_alpha016g/projects/scratch/aqua_mus/bld/intel/intelmpi/nodebug/nothreads/nuopc/pio/pio2/Makefile] Error 127

my config_compilers.xml:

Bash:

<SFC> ifort </SFC>
<SCC> icx </SCC>
<SCXX> icpx </SCXX>
<MPIFC> mpiifort </MPIFC>
<MPICC> "mpiicc -cc=icx"  </MPICC>
<MPICXX> "mpiicpc -cxx=icpx" </MPICXX>

What should I do? Thanks again!

All the best,
Lumos

Lumoss · Dec 24, 2023

Lumoss said:

Hi,
When I run./case.build, pio reports the following error:

Bash:

cat: Filepath: No such file or directory
cat: Srcfiles: No such file or directory
/bin/sh: line 1: -cc=icx: command not found
make: *** [/public/home/cesm2_3_alpha016g/projects/scratch/aqua_mus/bld/intel/intelmpi/nodebug/nothreads/nuopc/pio/pio2/Makefile] Error 127

my config_compilers.xml:

Bash:

<SFC> ifort </SFC>
<SCC> icx </SCC>
<SCXX> icpx </SCXX>
<MPIFC> mpiifort </MPIFC>
<MPICC> "mpiicc -cc=icx"  </MPICC>
<MPICXX> "mpiicpc -cxx=icpx" </MPICXX>

What should I do? Thanks again!

All the best,
Lumos

I think maybe I solved the problem. Thanks again!

Code:

set(MPICC "icx")
set(MPICXX "icpx")
set(MPIFC "mpiifort")
set(SCC "icx")
set(SCXX "icpx")
set(SFC "ifort")

jedwards · Dec 24, 2023

Make sure that icx is in your path as set in config_machines.xml - are you loading a module?

Lumoss · Dec 26, 2023

jedwards said:
Make sure that icx is in your path as set in config_machines.xml - are you loading a module?

Thank you for your prompt reply, the problem has been solved. But I found can't visit https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata today. What's going on here?

jedwards · Dec 26, 2023

Thank you for letting us know about inputdata, I can confirm that the server is down and will report the issue to our support staff.

Lumoss · Dec 26, 2023

jedwards said:
Thank you for letting us know about inputdata, I can confirm that the server is down and will report the issue to our support staff.

Thanks for replying, it is now available.

Lumoss · Dec 28, 2023

Great! I was able to run it successfully after updating the compiler and various library versions. Thank you so much!!

There will be some warnings during the run, what caused this? This doesn't seem to affect how the mode works.

Code:

(ice_comp_nuopc):(InitializeAdvertise)WARNING: atmiter_ from driver   0.10000D-
 01 is overwritting atmiter_conv from cice namelist   0.00000D+00
......
Warning: incorrect ice nucleation number (nuci reset =0)
   10479.1850686492        194.898301991199       0.580253623425064
  2.000000000000000E-002  0.000000000000000E+000  0.000000000000000E+000
  0.000000000000000E+000  0.000000000000000E+000  0.000000000000000E+000
  1.625526213166524E-008   560.609355150535

jedwards · Dec 28, 2023

You would need to ask the cice model developers about warnings from cice - each model component prints warnings to the output file.
Far too many in my opinion.

BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES

Member

CSEG and Liaisons

Member

Member

CSEG and Liaisons

Member

Attachments

CSEG and Liaisons

Member

Member

Member

CSEG and Liaisons

Member

Member

Member

CSEG and Liaisons

Member

CSEG and Liaisons

Member

Member

CSEG and Liaisons