[gmx-users] TestBed in MPI not working

[Jones de Andrade]

Thanks Justin. ;)

Well, I'm using a simple script to run the tests. The script follows:

**********************
#!/bin/tcsh
source ../../tcshrc
source /opt/chemistry/gmx404.ompi/bin/GMXRC.csh
set path=( $path /opt/mpi/ompi131/bin )
setenv LD_LIBRARY_PATH ${LD_LIBRARY_PATH}:/opt/mpi/ompi131/lib

set NCPUS = 2

set i = 1
while ($i <= $NCPUS)
  if ($i == 1) then
echo Single Precision Tests:
    ./gmxtest.pl simple #all
    ./gmxtest.pl clean >& /dev/null
echo Double Precision Tests:
    ./gmxtest.pl -double simple #all
    ./gmxtest.pl clean >& /dev/null
  else
echo Single Precision Tests - Parallel in $i
    ./gmxtest.pl -np $i simple #all
    ./gmxtest.pl clean >& /dev/null
echo Double Precision Tests - Parallel in $i
    ./gmxtest.pl -np $i -double simple #all
#    ./gmxtest.pl clean >& /dev/null
  endif
  @ i++
end
**********************

Using just the "simple" tests now to try to find out where is the error.
Since the error goes everywhere, it just simplify things.
It outputs everything ok with the serial part, but it all gets really messy
in the mpi section.

In version 4.0.4, it's complaining that the grompp is not understanding the
"-np" flag (Invalid command line argument). Then, it doesn't even create a
mdrun.out file. I tried to google this error, but basically nothing was
found. It also can't be a mixing of compilations, since everything is done
in completelly different and independent directories.

Version 3.3.3 on the other hand already failed in so many different places
that I'm really thinking IF I'll make it available in the new cluster. :P

This has to be any sort of pretty stupid mistake so. But I'm clueless this
time. :( Any ideas?

Thanks a lot in advance! :)

Jones

On Mon, May 11, 2009 at 7:53 PM, Justin A. Lemkul <jalemkul at vt.edu> wrote:

>
>
> Jones de Andrade wrote:
>
>> Hi everyone.
>>
>> Well, lets go straight to the point: last time I've installed gromacs, it
>> was still 3.3.x and there was no testbed available yet. I've done some tests
>> of my own and everything gone fine.
>>
>> I'm doing some new installations now, both gmx3.3.3 and gmx4.0.4. They
>> weem to work pretty well in serial, but absolutelly not with mpi. A whole
>> "plethora" of different errors appear.
>>
>> Before bothering the list with error messages, my question is: someone has
>> successfully used it with OpenMPI or MPIch? Before questions arise, I have
>> different test users for each, and absolutelly independent compilations.
>>
>>
> Every installation (serial and MPI) that I have installed worked
> flawlessly. Our cluster uses mpich, and our local machines (dual-core) use
> OpenMPI.
>
> Can you provide some of the error messages?  Are the reported from Gromacs,
> or from the MPI implementations?
>
> -Justin
>
>  Thanks a lot for any answer in advance.
>>
>> Sincerally yours,
>>
>> Jones
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> gmx-users mailing list    gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at http://www.gromacs.org/search before
>> posting!
>> Please don't post (un)subscribe requests to the list. Use the www
>> interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
>
> --
> ========================================
>
> Justin A. Lemkul
> Ph.D. Candidate
> ICTAS Doctoral Scholar
> Department of Biochemistry
> Virginia Tech
> Blacksburg, VA
> jalemkul[at]vt.edu | (540) 231-9080
> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>
> ========================================
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the www interface
> or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090511/61c344a3/attachment.html>