Dear John Thanks for the reply. I'll need help of you people to solve this problem. I am not expert in HPC and this would be my learning as well. Let me add that the cluster is based on Platform Cluster Manager (PCM) by IBM Computing. The compute nodes are NFS mounted with the installer node. Therefore the directory containing binary rca.x is also present in the compute nodes. Unfortunately I was trying to copy gfortran libraries from installer node to compute nodes using rsync but something went wrong and the model binary rca.x stopped working. I have recompiled the binary after reinstalling hdf as well as netcdf which model uses during compilation. All path are set in bashrc as well. Below is the output of ldd on master as well as compute nodes
[pmdtest@pmd HadGEM]$ ldd rca.x libstdc++.so.6 => /usr/local/lib64/libstdc++.so.6 (0x00002b6a9503c000) libnetcdff.so.5 => /usr/local/lib/libnetcdff.so.5 (0x00002b6a95344000) libnetcdf.so.7 => /usr/local/lib/libnetcdf.so.7 (0x00002b6a95798000) libhdf5.so.7 => /usr/local/lib/libhdf5.so.7 (0x00002b6a95aa1000) libhdf5_hl.so.7 => /usr/local/lib/libhdf5_hl.so.7 (0x00002b6a95f5c000) libsz.so.2 => /usr/local/lib/libsz.so.2 (0x00002b6a9618b000) libz.so.1 => /usr/local/lib/libz.so.1 (0x00002b6a9639f000) libmpi_f90.so.0 => /home/openmpi/lib/libmpi_f90.so.0 (0x00002b6a965b4000) libmpi_f77.so.0 => /home/openmpi/lib/libmpi_f77.so.0 (0x00002b6a967b7000) libmpi.so.0 => /home/openmpi/lib/libmpi.so.0 (0x00002b6a969ee000) libopen-rte.so.0 => /home/openmpi/lib/libopen-rte.so.0 (0x00002b6a96cb6000) libopen-pal.so.0 => /home/openmpi/lib/libopen-pal.so.0 (0x00002b6a96f16000) libdl.so.2 => /lib64/libdl.so.2 (0x00000033e0e00000) libnsl.so.1 => /lib64/libnsl.so.1 (0x00000033e2200000) libutil.so.1 => /lib64/libutil.so.1 (0x00000033ee400000) libm.so.6 => /lib64/libm.so.6 (0x00000033e1200000) libpthread.so.0 => /lib64/libpthread.so.0 (0x00000033e1600000) libc.so.6 => /lib64/libc.so.6 (0x00000033e0a00000) libgcc_s.so.1 => /usr/local/lib64/libgcc_s.so.1 (0x00002b6a971a0000) /lib64/ld-linux-x86-64.so.2 (0x00000033e0600000) librt.so.1 => /lib64/librt.so.1 (0x000000362ac00000) libifport.so.5 => /opt/intel/Compiler/11.1/064/lib/intel64/libifport.so.5 (0x00002b6a973b5000) libifcore.so.5 => /opt/intel/Compiler/11.1/064/lib/intel64/libifcore.so.5 (0x00002b6a974ef000) libimf.so => /opt/intel/composer_xe_2013.0.079/compiler/lib/intel64/libimf.so (0x00002b6a97765000) libsvml.so => /opt/intel/composer_xe_2013.0.079/compiler/lib/intel64/libsvml.so (0x00002b6a97c2f000) libintlc.so.5 => /opt/intel/composer_xe_2013.0.079/compiler/lib/intel64/libintlc.so.5 (0x00002b6a984f5000) libifcoremt.so.5 => /opt/intel/Compiler/11.1/064/lib/intel64/libifcoremt.so.5 (0x00002b6a98743000) libirng.so => /opt/intel/composer_xe_2013.0.079/compiler/lib/intel64/libirng.so (0x00002b6a989e8000) [pmdtest@pmd HadGEM]$ ssh compute-01-18 ssh: connect to host compute-01-18 port 22: No route to host [pmdtest@pmd HadGEM]$ ssh compute-01-13 Last login: Mon Jan 28 07:48:08 2013 from pmd-eth0.private.dns.zone [pmdtest@compute-01-13 ~]$ ldd rca.x ldd: ./rca.x: No such file or directory [pmdtest@compute-01-13 ~]$ ls /home/pmdtest/RCA4_CORDEX/RCA4_CORDEX_SAsia/HadGEM/rca.x Regards Ahsan On Thu, Feb 7, 2013 at 7:40 PM, John Hearns <hear...@googlemail.com> wrote: > ldd rca.x > > Try logging in to each node and run this command. > Even better use pdsh > > _______________________________________________ > users mailing list > us...@open-mpi.org > http://www.open-mpi.org/mailman/listinfo.cgi/users >
