The "ssh cn308 ldd $WIENROOT/lapw0_mpi" is finding files for your ifort
installation like
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_scalapack_lp64.so
just fine. So your environmental variables seem to be setup and working
fine on both nodes. It looks like the
/opt/intel/impi/5.0.2.044/intel64/lib/libmpifort.so.12 exists on the
renwei node but not on the cn308 node. It looks to me that Intel MPI
(impi) is not installed on the cn308 node.
Perhaps the cn308 node is using a different partition or different
shared drive. I have read that there are different possible solutions
for the slurm cluster problem you seem to have which depend on how it is
configured [
https://lists.schedmd.com/pipermail/slurm-users/2017-December/000272.html
]. You might be able to check which partition the renwei node and cn308
node are using with sinfo [ https://slurm.schedmd.com/sinfo.html ].
Maybe you just have to have your cluster manager (administrator, help
desk, ...) install impi like what you did for ifort. To remove the
"manpath: command not found", the cluster manager probably just has to
install the man or man-db package on the cn308 node (they should be able
to check the documentation or forums for the OS that their cluster is
using on how to install manpath, typically for example: yum install man
or apt-get install man-db). I have never performed administration
functions of a slurm cluster, so for additional help with your problem
you may have to ask a slurm expert (e.g., your cluster manager or the
slurm mailing list [ https://slurm.schedmd.com/mail.html ]).
On 6/16/2018 4:28 AM, venkatesh chandragiri wrote:
Dear Prof. Marks,
I did "ssh othernode ldd $WIENROOT/lapw0_mpi".
=========
[renwei@ln3 ~]$ ssh cn308 ldd $WIENROOT/lapw0_mpi
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/bin/mklvars.sh: line
118: manpath: command not found
linux-vdso.so.1 => (0x00007fffd8fff000)
libfftw3_mpi.so.3 =>
/THFS/home/renwei/venky/soft/fftw/lib/libfftw3_mpi.so.3
(0x00007fd41621d000)
libmkl_scalapack_lp64.so =>
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_scalapack_lp64.so
(0x00007fd415947000)
libmkl_blacs_intelmpi_lp64.so =>
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.so
(0x00007fd41570a000)
libfftw3.so.3 =>
/THFS/home/renwei/venky/soft/fftw/lib/libfftw3.so.3 (0x00007fd4153fe000)
libmkl_intel_lp64.so =>
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_intel_lp64.so
(0x00007fd414cb0000)
libmkl_intel_thread.so =>
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_intel_thread.so
(0x00007fd413c90000)
libmkl_core.so =>
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/lib/intel64/libmkl_core.so
(0x00007fd41259c000)
libpthread.so.0 => /lib64/libpthread.so.0 (0x00007fd412380000)
*_ libmpifort.so.12 => not found
libmpi.so.12 => not found_*
libdl.so.2 => /lib64/libdl.so.2 (0x00007fd412172000)
librt.so.1 => /lib64/librt.so.1 (0x00007fd411f69000)
libm.so.6 => /lib64/libm.so.6 (0x00007fd411ce5000)
libiomp5.so =>
/opt/intel/composer_xe_2013_sp1.3.174/compiler/lib/intel64/libiomp5.so
(0x00007fd4119ca000)
libc.so.6 => /lib64/libc.so.6 (0x00007fd411628000)
libgcc_s.so.1 =>
/THFS/home/sh-hzw2/software/Matlab2014a//sys/os/glnxa64/libgcc_s.so.1
(0x00007fd411413000)
libimf.so =>
/opt/intel/composer_xe_2013_sp1.3.174/compiler/lib/intel64/libimf.so
(0x00007fd410f50000)
libsvml.so =>
/opt/intel/composer_xe_2013_sp1.3.174/compiler/lib/intel64/libsvml.so
(0x00007fd410354000)
libirng.so =>
/opt/intel/composer_xe_2013_sp1.3.174/compiler/lib/intel64/libirng.so
(0x00007fd41014d000)
libintlc.so.5 =>
/opt/intel/composer_xe_2013_sp1.3.174/compiler/lib/intel64/libintlc.so.5
(0x00007fd40fef7000)
/lib64/ld-linux-x86-64.so.2 (0x00007fd416436000)
=========
As it is shown here *_ libmpifort.so.12 => not found,
libmpi.so.12 => not found when I run in cn308 node_*
But these have well defined paths when run ldd at "renwei"
libmpifort.so.12 =>
/opt/intel/impi/5.0.2.044/intel64/lib/libmpifort.so.12
<http://5.0.2.044/intel64/lib/libmpifort.so.12> (0x00002b3a37c98000)
libmpi.so.12 =>
/opt/intel/impi/5.0.2.044/intel64/lib/libmpi.so.12
<http://5.0.2.044/intel64/lib/libmpi.so.12> (0x00002b3a37f21000)
===================
[renwei@ln3 ~]$ ssh cn308 $WIENROOT/lapw0_mpi
/THFS/opt/intel/composer_xe_2013_sp1.3.174/mkl/bin/mklvars.sh: line
118: manpath: command not found
/THFS/home/renwei/venky/soft/wien2k/lapw0_mpi: error while loading
shared libraries: libmpifort.so.12: cannot open shared object file: No
such file or directory
[renwei@ln3 ~]$
===============
[renwei@ln3 ~]$ ssh cn308
Last login: Sat Jun 16 17:59:04 2018 from ln3-gn0
-bash: manpath: command not found
[renwei@cn308 ~]$ $WIENROOT/lapw0_mpi
/THFS/home/renwei/venky/soft/wien2k/lapw0_mpi: error while loading
shared libraries: libmpifort.so.12: cannot open shared object file: No
such file or directory
*
*
*========================*
You also mentioned to use " use static compilation". I don't
understand this. do you meant to be static compilation of wien2k..?
how I can do it (I am sorry to ask this, as I belongs to experimental
background I don't come across these kind of issues).
thank you.
venkatesh
_______________________________________________
Wien mailing list
Wien@zeus.theochem.tuwien.ac.at
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
