On May 18, 2006, at 3:49 AM, Valter Dal Bo wrote:
> Thank you for your kind answer !
>
> How do you just "reply" to the posts in the list ?????
> Anyway.....let's use a workaround..... :-(
I just hit reply -- I'm not sure what you mean -- is there some
problem you are having?
> > Well, issuing the "lamboot" I get the following message:
> >
> > $ lamboot -v
> >
> > LAM 7.0.6/MPI 2 C++/ROMIO - Indiana University
> >
> > n-1<4268> ssi:boot:base:linear: booting n0 (localhost)
> > n-1<4268> ssi:boot:base:linear: finished
> >
> > The above means that the process exited without enabling node1 and
> > therefore it fails the initialization.
> > At first I thought it was due to the fact that rsh'ing I was getting
> > some messages in return:
>
> Actually, since you didn't give any host file to lamboot, it did
> exactly what it should. It defaulted to a hostfile of "localhost"
> and started a universe there. So at this point, LAM/MPI looks ok.
>
> Thing do not change even supplying a hostfile p.e.:
>
> file: hostfile
> redhat2 cpu=2
>
> then running a lamboot hostfile -d I get:
<snip>
> Issuing the same command on a RedHat9+LAM6.5.9 I get what follows:
>
> lamboot hostfile -d
>
> LAM 6.5.9/MPI 2 C++/ROMIO - Indiana University
>
<snip>
> Which looks quite a bit different, but I suppose the result would
> be the same; booting the 2 cpus.
Yes, between 6.5.9 and 7.0 we added a lot more debugging output to
make it easier to debug startup issues. But in both cases,
everything looks fine.
> The lamhalt also seem to be working fine, it is just when I try to
> run LS-DYNA a job using the "mpirun -np 2 executable -i=inputfile"
> that I get an error as reported:
>
> ----------------------------------------------------------------------
> -------
>
> It seems that there is no lamd running on the host .
>
> This indicates that the LAM/MPI runtime environment is not operating.
> The LAM/MPI runtime environment is necessary for MPI programs to run
> (the MPI program tired to invoke the "MPI_Init" function).
>
> Please run the "lamboot" command the start the LAM/MPI runtime
> environment. See the LAM/MPI documentation for how to invoke
> "lamboot" across multiple machines.
> ----------------------------------------------------------------------
> -------
<snip>
> Again, the /tmp/lam-catusr_at_redhat2/lam-debug-log.txt file says what
> I have already posted in my original message.
>
> Any idea ?
> I don't fancy compiling a 6.5.9 lam environment in order to run
> this program as I have several different other application which
> may use different versions of lam and I would prefer not to mess
> with the original RHEL installation if possible.
Ah, now I see the issue. And you diagnosed it correctly. You can
not use LAM 6.5.9 for the user application and 7.1 for the run-time
environment -- the versions must match. We try not to change our ABI
too much so that you don't have to recompile apps linked against
shared library versions of libmpi and liblam (you just use the new
shared libraries), but that's about the extent of our version
interoperability.
So if your copy of LS-DYNA is built against a static liblam, you need
to use 6.5.9 to run your jobs.
Brian
--
Brian Barrett
LAM/MPI developer and all around nice guy
Have a LAM/MPI day: http://www.lam-mpi.org/
|
