On Mon, Mar 06, 2006, B. Lachele Foley wrote:
>
> It doesn't look like an input issue because when it isn't
> successful, the job goes for a few thousand steps and seems to
> just hang (using cpu according to top, but not writing any
> output for, like, 20 hours).
Is this reproducible? It sounds like an mpi problem -- either from MPI
itself, or because one of the two processes crashed, and the other process
is just sitting there, waiting for a return message. Is there anything in
the two output files for the job that crashed that is helpful or suggests
a problem?
I'm not sure that I've ever tested a cap water problem with icfe in Amber
9,
but it does seem odd that it would run for thousands of steps. What
happens
if set nstlim to a number shorter than the value at which the simulation
stopped? Do you get a full output with reasonable looking results?
...dac
Received on Wed Apr 05 2006 - 23:49:41 PDT
