On Fri, 9 Jan 2004 louis.gonzales_at_[hidden] wrote:
> Hello everyone, i'm new to the mailing list - and to cluster
> configuration - I've currently got 4 nodes - redhat 9.0 kernel 2.4.20 -
> lam 6.5.9 installed and after issuing:
> [snipped]
>
> However, when I run mpirun like this:
> # mpirun -np 4 /some/mpi/program/from/LSTC info
> I get the following error:
> --------------------------------------------------------------------------------------
> It seems that [ at least ] one of processes that was started with mpirun
> did not invoke MPI_INIT before quitting ( it is possible that more than
> one process did not invoked MPI_INIT -- mpirun was only notified of the
> first one, which was on node n0).
>
> mpirun can *only* be used with MPI programs (i.e., programs that invoke
> MPI_INIT and MPI_FINALIZE). You can use the "lamexec" program to run
> non-MPI programs over the lambooted nodes.
> ---------------------------------------------------------------------------------------------
What is happening is exactly what the error message says -- at least one
of the programs exited without invoking MPI_INIT. mpirun is only geared
towards running MPI applications; it expects all of the executables that
it launches to either eventually invoke MPI_INIT or have one of its
children invoke MPI_INIT.
I'm guessing that you got a binary executable from LSTC, and you didn't
compile it from source with LAM/MPI 6.5.9? If so, you might want to check
the following things:
1. That the LSTC executable runs in your environment (expects the same
version of glibc, etc.)
2. That the LSTC executable was compiled against LAM/MPI 6.5.9.
--
{+} Jeff Squyres
{+} jsquyres_at_[hidden]
{+} http://www.lam-mpi.org/
|
