Hi Sadhna,
Can you check that the sander.MPI executable is actually running in parallel
and not just running 2 serial copies. Take a look at
$AMBERHOME/test/dhfr/mdout.dhfr from the last parallel set of test cases you
ran. The key lines you are looking for are:
| Atom division among processors:
| 0 11467 22930
Sum of charges from parm topology file = -0.00000006
Forcing neutrality...
| Running AMBER/MPI version on 2 nodes
If this is missing then it implies something didn't work properly during the
parallel build.
Also please check that the TI test cases (ti_eth2meth_gas and ti_ggcc)
passed in parallel.
All the best
Ross
> -----Original Message-----
> From: owner-amber.scripps.edu [mailto:owner-amber.scripps.edu] On Behalf
> Of Sadhna Rana
> Sent: Thursday, October 02, 2008 9:08 AM
> To: amber.scripps.edu
> Cc: baker.ccb.wustl.edu
> Subject: RE: AMBER: thermodynamic integration with Amber9
>
> Prof Ross Walker,
> I am using the right executable,
> /usr/bin/mpirun -np 2 /usr/local/amber9/exe/sander.MPI -ng 2
> Error: specified more groups ( 2 ) than the number of
> processors (
>
> 1 ) !
>
> I've also verified that the sander.MPI does execute in parallel when I
> don't
> use the -ng option.(When I use it for MD without TI)
>
> I did run the parallel test cases.
> Most tests pass, but there are a few failures:
>
> export TESTsander=/Users/srana/AmberTest/amber9/exe/sander.MPI; make
> test.sander
> .BASIC
> cd dmp; ./Run.dmp
> This test not set up for parallel
> cannot run in parallel with #residues < #pes
> cd adenine; ./Run.adenine
> This test not set up for parallel
> cannot run in parallel with #residues < #pes
> ==============================================================
> cd cytosine; ./Run.cytosine
> diffing cytosine.out.save with cytosine.out
> PASSED
> ...
> cd dhfr; ./Run.dhfr.min
> diffing mdout.dhfr.min.save with mdout.dhfr.min
> possible FAILURE: check mdout.dhfr.min.dif
> ...
> cd dna_pol; ./Run.dna_pol
> diffing mdout.dna_pol.save with mdout.dna_pol
> possible FAILURE: check mdout.dna_pol.dif
> ...
> cd rdc; ./Run.dip
> ./Run.dip: Program error
>
> The tests exit on the last error.
>
> This is on an OSX 10.5.5 system, but we've verified the same runtime error
> with the -ng flag on other architectures.
>
>
> thanks
> sadhna
>
>
> > Hi Sadhna,
>
>
> >
> >> Error: specified more groups ( 2 ) than the number of
> >> processors ( 1 ) !
> >>
> >> inspite of specifying 2 processors through -np 2 option
> >>
> >> I am not sure where I am going wrong. If anyone can give some hints
> >> that
> >> would be great.
> >
> > Are you using sander.MPI as the executable? You must do this which means
> > you
> > must have compiled the parallel version of the code. So you would do
> > something like:
> >
> > mpirun -np 2 $AMBERHOME/exe/sander.MPI .....
>
> >
> > Also did you run the parallel test cases? You should really do this -
> make
> > sure you set DO_PARALLEL='mpirun -np 2' and then do
> >
> > cd $AMBERHOME/test/
> > make test.parallel
> >
> > If this gives errors then please post these to the list since this is
> the
> > best way for us to help you debug.
> >
> > All the best
> > Ross
> >
> >
> > /\
> > \/
> > |\oss Walker
> >
> > | Assistant Research Professor |
> > | San Diego Supercomputer Center |
> > | Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
> > | http://www.rosswalker.co.uk | PGP Key available on request |
> >
> > Note: Electronic Mail is not secure, has no guarantee of delivery, may
> not
> > be read every day, and should not be used for urgent or sensitive
> issues.
> >
> >
> >
> >
> > -----------------------------------------------------------------------
> > The AMBER Mail Reflector
> > To post, send mail to amber.scripps.edu
> > To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
> > to majordomo.scripps.edu
> >
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
> to majordomo.scripps.edu
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
to majordomo.scripps.edu
Received on Fri Oct 03 2008 - 05:10:53 PDT
