Difference between revisions of "Gaussian/Batch"

From ScientificComputing
Jump to: navigation, search
Line 1: Line 1:
 
You can submit a Gaussian computation using the input file <tt>test000.com</tt> in batch mode with the following command.
 
You can submit a Gaussian computation using the input file <tt>test000.com</tt> in batch mode with the following command.
  
  bsub [LSF parameters] g09 test000
+
  bsub [LSF parameters] -R "rusage[scratch=500]" g09 test000
  
The results of the simulation will be stored in a file called test000.log. LSF parameters are documented on the wiki page about [[Using_the_batch_system|using the batch system]].
+
The results of the simulation will be stored in a file called test000.log. LSF parameters are documented on the wiki page about [[Using_the_batch_system|using the batch system]]. Please note that you need to request scratch space in any case. The batch system automatically sets
 +
 
 +
GAUSS_SCRDIR=$TMPDIR
 +
 
 +
which forces Gaussian to write temporary file during the calculation into the local scratch of the compute node. Using a '''local''' scratch directory offers some significant advantages:
 +
 
 +
* no risk to fill up your home directory
 +
* very high I/O performance

Revision as of 09:11, 25 August 2016

You can submit a Gaussian computation using the input file test000.com in batch mode with the following command.

bsub [LSF parameters] -R "rusage[scratch=500]" g09 test000

The results of the simulation will be stored in a file called test000.log. LSF parameters are documented on the wiki page about using the batch system. Please note that you need to request scratch space in any case. The batch system automatically sets

GAUSS_SCRDIR=$TMPDIR

which forces Gaussian to write temporary file during the calculation into the local scratch of the compute node. Using a local scratch directory offers some significant advantages:

  • no risk to fill up your home directory
  • very high I/O performance