My Community

Hi,
I have compiled and executed parallel vasp on several machines. But I am not able to execute it on one cluster, It gives me the following error. I have no clue, even sys admin has no clue. please help out.
Machines:
node051
node054
node053
node058
node046
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
/export/local/etc/std-cshrc: No such file or directory.
FORTRAN STOP
FORTRAN STOP
Error reading item 'IMAGES' from file INCAR.
Error reading item 'IMAGES' from file INCAR.
FORTRAN STOP
FORTRAN STOP
Error reading item 'IMAGES' from file INCAR.
Error reading item 'IMAGES' from file INCAR.
FORTRAN STOP
Error reading item 'IMAGES' from file INCAR.

hmmm, strange. It seems that the INCAR file cannot be opened and/or read correctly (according to the vasp- error output "Error reading item 'IMAGES' from file INCAR."
please check if INCAR has been copied to the approporiate machine and batch worker directory by the job script. It has to be the directory that vasp starts from: you can simply add a line which copies the contents of INCAR to the job.stdout file like 'cat INCAR' in your job-script in order to do so.

However, I rather think that the real problem is due to the error that causes that /export/local/etc/std-cshrc error message. Though I do not use the csh, this looks like if some shell environment variables are not set correctly because some run-commands-file is not found. Maybe it helps to explicitely start the jobscript with a line like (the csh -equivalent of)
#!/bin/bash --rcfile=$HOME/.bashrc
(you will have to set the $MPI environment variable and the $PATH, including the $LD_LIBRARY_PATH in the .xxxrc file (either .bashrc or .cshrc)
Please ask your system admin again if this might be the case, these details depend on your local system, how the parallel cluster is set up,....