[SIESTA-L] Problems Running in Parallel

[Thomas Helander]

Hi,








Thanks,
Thomas Helander

<u9>/home/thelande/SIESTA-1.3/h2o [23] mpiexec -n 3 siesta < h2o.fdf
SIESTA 1.3.7  -- [1.3f1 release] (04 Feb 2004)
Architecture  : pgf90-mpich
Compiler flags: mpif90 -g -i4
PARALLEL version

* Running on    3 nodes in parallel
sname=
sname=
>> Start of run:   2-MAY-2007  13:26:49

                           ***********************
                           *  WELCOME TO SIESTA  *
                           ***********************

reinit: Reading from standard input

SystemName          Water molecule
SystemLabel         h2o
NumberOfAtoms       3
NumberOfSpecies     2
%block ChemicalSpeciesLabel
1  8  O      # Species index, atomic number, species label
2  1  H
%endblock ChemicalSpeciesLabel
AtomicCoordinatesFormat  Ang
%block AtomicCoordinatesAndAtomicSpecies
0.000  0.000  0.000  1
0.757  0.586  0.000  2
-0.757  0.586  0.000  2
%endblock AtomicCoordinatesAndAtomicSpecies

I am attempting to run siesta in parallel but every time I run the program, it hangs right after displaying the input file and when I check top, it reports that siesta is using no processor time. Siesta compiled with no warnings or errors and runs fine in serial. Likewise, I receive no errors when running in parallel except that it never proceeds past the input file, even with the simple h2o example. Below is the output I do receive and my arch.make is attached in case it is needed. As a note, I am using 1.3 due to restrictions placed on me by other software, otherwise I would be using 2.0. ************************** Dump of input data file ****************************

arch.make
Description: Binary data

smime.p7s
Description: S/MIME cryptographic signature