From: Lorenzo Dematté (dematte_at_[hidden])
Date: 2008-06-10 09:47:53
I don't know about linux installation, but on the execution side is the
same as on all other MPI implementation: on every machine in the
cluster, there are one or more services/daemons running and waiting for
requests. When you use the mpiexec program is used to launch your
executable, this contacts the other nodes and starts the application on
all of them; the usual stuff...
Carmen Navarrete Navarrete wrote:
> Dear list readers,
> I have downloaded the 0.8.0 version of the MPI.NET for Unix platforms. I
> have any configure file, so I tried to execute the autogen.sh, but this
> is the result I obtained:
> --> Running aclocal
> configure.ac:205: error: AC_SUBST: `MCS ILASM MONO MPIEXEC PERL
> CSHARP_FLAGS CSHARP_DEFINES MPI_HEADER
> MPI_SHARED_LIB_NAME MPI_FUNC_PREFIX' is not a valid shell variable
> configure.ac:205: the top level
> autom4te-2.62: /usr/bin/m4 failed with exit status: 1
> aclocal-1.10: autom4te failed with exit status: 1
> *** aclocal failed
> What does it means?
> On the other hand. I'm confused about how to execute a parallel
> application with this implementation of MPI in more than one machine.
> Thanks in advance and best regards.