On Fri, 8 Apr 2016, Harriet Li wrote:
> I think the problem might have been that I followed the standard
> petsc installation and used --download-mpich, which seemed to have
> installed mpich somewhere that petsc would use it but that wouldn't
> show up with 'whereis'.
Oh, certainly! That would
~$ ls -l $(whereis mpirun)
ls: cannot access mpirun:: No such file or directory
lrwxrwxrwx 1 root root 24 Jul 14 2015 /usr/bin/mpirun ->
/etc/alternatives/mpirun
lrwxrwxrwx 1 root root 7 Dec 28 2013 /usr/bin/mpirun.openmpi -> orterun
lrwxrwxrwx 1 root root 24 Jul 14 2015 /usr/bin/X11/mpirun ->
On Thu, 7 Apr 2016, Harriet Li wrote:
> The only MPI's I could find are:
>
> $ whereis mpirun
> mpirun: /usr/bin/mpirun.openmpi /usr/bin/mpirun /usr/bin/X11/mpirun.openmpi
> /usr/bin/X11/mpirun /usr/share/man/man1/mpirun.1.gz
>
> $ whereis mpicxx
> mpicxx: /usr/bin/mpicxx /usr/bin/mpicxx.openmpi
It certainly counts. You have several which mean you need to be very
careful about the MPI you are picking up during compilation versus the one
you are using at runtime. You might consider using a "modules" management
system where you have exactly one loaded in your PATH, compile paths, link
paths
On Thu, Apr 7, 2016 at 3:57 PM, Cody Permann wrote:
> This generally means your compile environment and runtime environment
> aren't the same. Do you have multiple MPIs hanging around? Are your paths
> sane?
>
Heh, we just noticed this in the training as well:
Finite Element Types="LAGRANG
The only MPI's I could find are:
$ whereis mpirun
mpirun: /usr/bin/mpirun.openmpi /usr/bin/mpirun /usr/bin/X11/mpirun.openmpi
/usr/bin/X11/mpirun /usr/share/man/man1/mpirun.1.gz
$ whereis mpicxx
mpicxx: /usr/bin/mpicxx /usr/bin/mpicxx.openmpi /usr/bin/X11/mpicxx
/usr/bin/X11/mpicxx.openmpi /usr/s
This generally means your compile environment and runtime environment
aren't the same. Do you have multiple MPIs hanging around? Are your paths
sane?
On Thu, Apr 7, 2016 at 3:55 PM Harriet Li wrote:
> Hello,
>
> I'm trying to run example 4 under Introduction in parallel. It seems to be
> running
Hello,
I'm trying to run example 4 under Introduction in parallel. It seems to be
running in serial, but on multiple processors, instead of splitting up the
work.
I've attached the config.log file for my libmesh compilation as well as the
output when I try to run the example using
mpirun -np 2 .
