Ah, this is MPICH2 or some variant thereof. It's mpich2 version 1.2.1p1 or
something like that, right?
In that case, you'll need to start an mpd daemon on all of the nodes that
you want to run a parallel job on, so you'll need to run something like
"mpd &" or "mpdboot". I'd personally suggest the latter, since it launches
all of the mpd processes as daemons and doesn't need to be thrown to the
background.
You should also throw in an "mpdallexit" after the job is done to kill the
daemons. If you can, I'd actually suggest upgrading to MPICH2 1.4.1p1 (the
current stable version, last time I checked). It doesn't require an mpd
daemon at all, and removes that bit of instability (it's been a bit buggy
in my experience).
All the best,
Jason
On Tue, Dec 27, 2011 at 6:42 PM, David Condon <dec986.gmail.com> wrote:
> Hi Jason,
>
> I figured this out. The solution is simple, given by standard output of
> one of the mpiexec commands:
>
> cd $HOME
> touch .mpd.conf
> chmod 600 .mpd.conf
> vim .mpd.conf #insert your secretword here:
> SECRETWORD=your_secret_word_here
> mpd &
> ./Run &
>
> Just in case anyone else runs into this error in the future.
>
> Thanks so much for your help Jason!
>
> Have a happy 2012,
> -Dave
>
> On Tue, Dec 27, 2011 at 6:00 PM, Jason Swails <jason.swails.gmail.com
> >wrote:
>
> > On Tue, Dec 27, 2011 at 12:20 PM, David Condon <dec986.gmail.com> wrote:
> >
> > > Hello,
> > >
> > > I am trying to make AMBER11 parallel for PMEMD production runs on my
> > lab's
> > > 8-cpu desktop computer, following instructions on
> > > http://ambermd.org/ubuntu.html The installation completed.
> > >
> > > However, I am trying to run PMEMD with the following:
> > >
> > > mpirun -np 8 pmemd -O -i md3.in -o 18.out -p prmtop -c 17.rst -r
> 18.rst
> > -x
> > > > 18.mdcrd
> > >
> >
> > This doesn't sound like it'll solve the issue you're having, but pmemd is
> > the serial version of pmemd in Amber 11. Now, pmemd.MPI is the parallel
> > version.
> >
> > >
> > >
> > > as I have done successfully on other AMBER11 installations.
> > >
> > > This time, however, I get the error:
> > >
> > > mpiexec failed: gethostbyname_ex failed for zip.chem.rochester.edu
> > > >
> > >
> > > I don't think this is an AMBER-specific problem but something wrong
> with
> > > /etc/hosts; but I have no idea how to fix it.
> > >
> > > What changes can I make so I run PMEMD in parallel?
> > >
> > > thanks for your expertise,
> > > -Dave
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > >
> >
> >
> >
> > --
> > Jason M. Swails
> > Quantum Theory Project,
> > University of Florida
> > Ph.D. Candidate
> > 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
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Dec 27 2011 - 17:30:02 PST
