Re: [AMBER] allocation failure in vector: problem in MMPBSA.py.MPI

From: Sushil Mishra <sushilbioinfo.gmail.com>
Date: Wed, 15 Dec 2010 21:26:05 +0100

On Wed, Dec 15, 2010 at 6:47 PM, Jason Swails <jason.swails.gmail.com>wrote:

> On Wed, Dec 15, 2010 at 10:49 AM, Sushil Mishra <sushilbioinfo.gmail.com
> >wrote:
>
> > On Wed, Dec 15, 2010 at 4:40 PM, Jason Swails <jason.swails.gmail.com
> > >wrote:
> >
> > > On Wed, Dec 15, 2010 at 6:11 AM, mish <smncbr.gmail.com> wrote:
> > >
> > > > Hi Bill,
> > > > I finished a new simulation where I used 12CPU's, with 48GB of RAM
> for
> > > > the calculation and the problems still persist. I can see the same
> > error
> > > > "allocation failure in vector: nh = 276948109" in _complex_nm.out
> file.
> > I
> > > >
> > >
> > > This indicates a malloc failure (allocation). It's trying to allocate
> > > 276948109 "double" numbers. On a typical machine I think that's 8 bytes
> > > (64-bit real), so it's trying to allocate a total of 2.063 GB of
> memory,
> > I
> > > think, and for some reason it's failing.
> > >
> > >
> > > > have ran another calculation in 4CPU's with 48 GB of RAM and
> > calculation
> > > > are
> > > > running. However, in that running_MMPBSA_complex.mdcrd.1 calculation
> I
> > > cant
> > > > see values of entropy in
> > > > the _MMPBSA_complex_nm.out.1, _MMPBSA_complex_nm.out.2,
> > > > _MMPBSA_complex_nm.out.3
> > > > files. After each minimization is message "allocation failure in
> > vector:
> > > > nh " in the place of values of entropies. What could be other source
> of
> > > > error ?? is there something wrong with my input file ?
> > > >
> > >
> > > Did you clean all temporary files before re-running? It could be you're
> > > seeing the remnants of previous attempts?. Perhaps there's a memory
> leak
> > > as
> > > well... can you determine how much memory is being used during the
> > > simulation?
> > >
> >
> > Yes I clean all the temporary file before rerunning the calculations.
> When
> > I
> > look for the memory use during the simulation its about 8GB where as I
> had
> > assigned 48GB in for the calculations. What could be other possible
> source
> > of such error ?
> >
>
> 8 GB is being used for how many threads?
> by 4 all the threads.
>
> >
> > Sincerely
> > Sushil
> >
> >
> >
> > --
> Jason M. Swails
> Quantum Theory Project,
> University of Florida
> Ph.D. Graduate Student
> 352-392-4032
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Dec 15 2010 - 12:30:04 PST
Custom Search