[Pw_forum] igk files

Axel Kohlmeyer akohlmey at cmm.chem.upenn.edu
Sun Jul 13 22:21:49 CEST 2008


On Sun, 13 Jul 2008, alan chen wrote:

AC> Dear PWSCF users,
AC>      In my calculation, I set wfcdir to be '/scratch' in order to save disk
AC> space. I have found some xxx.igk files are dumped in the scratch direction.
AC> I have two questions.
AC> 1) I have a few jobs running at the same time. Unfortunately they have the
AC> same 'prefix' so that these xxx.igk files may overwrite each other. Does
AC> that matter? Does the code read in these files?

yes, it matters. that is the reason for 'prefix' in the first place. 
i'm usually too paranoid for just using 'prefix' and alway create 
separate directories for each job in the scratch using the jobid 
string from the batch system (which is guaranteed to be unique in 
the cluster and i don't have to worry about somebody else accidentally
having created a file that would have the same name as mine).

AC> 2) Recently when I calculate large jobs, I got the following error message:
AC> 'input statement requires too much data, unit 16, file
AC> /home1/hc336/scratch/IF-relax.igk19'. Does that mean the xxx.igk19 file is
AC> too large for the code to read in?

how should we know? nobody here can read the mind of your computer(s).
what is the exact sequence of actions/inputs to reproduce this? 

cheers,
   axel.

AC> 
AC> Thank you very much.
AC> 
AC> Hanghui
AC> Department of Physics
AC> Yale University
AC> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.



More information about the users mailing list