do you know if the exchange actually occurred? check the remlog and
see if it did the exchange. maybe the problem is in a different process
opening the file created by the other. if they did exchange, then increase
the difference in the temperatures in order to force them to not have a
successful exchange (like 200K and 500K) and see if it still crashes.
obviously that isn't the simulation you want to do but it will help
narrow down the
problem.
===================================================================
Carlos L. Simmerling, Ph.D.
Associate Professor Phone: (631) 632-1336
Center for Structural Biology Fax: (631) 632-1555
Stony Brook University Web: http://comp.chem.sunysb.edu/carlos
Stony Brook, NY 11794-5115 E-mail: carlos.simmerling.stonybrook.edu
===================================================================
Jordi Rodrigo wrote:
>Hi Guangiei,
>
>we have tried to run the job under various locations. The TMPDIR is a
>temporary directory created automatically and specificaly for submitted job
>and destroyed at the end of the job. The space is local for performance
>reasons (the i/o are way fasters). But we have tried also to run the job on
>GPFS shared directories. GPFS is an IBM proprietary NFS-like. But in both
>cases we continue to get the same error:
>
> Unit 6 Error on OPEN: mdout.rem.001
>
> Unit 6 Error on OPEN: mdout.rem.000
>
>Your help is very appreciated,
>
>Gilles and Jordi
>
>
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Mon Jan 17 2005 - 17:53:00 PST
