We are considering compiling openmpi with --with-pmi=/opt/slurm to
enable running mpi jobs with srun.
If we do this, will we have to recompile openmpi and/or programs built
with openmpi when we upgrade slurm? (If so, only for major upgrades, or
for minor upgrades as well?)
--
Regards,
Hi,
I have the case that OpenMPI was built against Slurm 14.03 (which provided
libslurm.so.27). Since upgrading to 14.11 I get errors
like:
[controller:35605] mca: base: component_find: unable to open
/opt/apps/openmpi/1.8.1/gcc/4.9/0/lib/openmpi/mca_ess_pmi:
libslurm.so.27: cannot open shared
No, you shouldn't have to do so - it's a dynamic library that gets picked
up at execution
On Thu, Apr 16, 2015 at 2:55 AM, Bjørn-Helge Mevik b.h.me...@usit.uio.no
wrote:
We are considering compiling openmpi with --with-pmi=/opt/slurm to
enable running mpi jobs with srun.
If we do this,
Hmmm...yeah, it sounds like Slurm changed it's library names and/or
dependencies. I'm afraid that you do indeed need to recompile OMPI in that
case. You probably need to rerun configure as well, just to be safe.
Sorry - outside OMPI's control :-/
On Thu, Apr 16, 2015 at 5:22 AM, Uwe Sauter
You have to start a separate munge daemon on the cluster that uses the key
being used by the slurmdbd. Then you put that munge's socket path in
AccountingStoragePass= in the cluster's slurm.conf. So one munge daemon
provides authentication for the cluster and another for communication
Hi Ralph,
beside the mentioned libslurm.so.28 there is also a libslurm.so pointing to the
same libslurm.so.28.0.0 file. Perhaps OpenMPI
could use this link instead of the versioned on?
File list in slurm/lib directory:
-rw-r--r-- 1 slurm slurm 68992 Mar 20 11:39 libpmi.a
-rwxr-xr-x 1 slurm
Dear Slurm developers,
before I set up Slurm with multi-cluster support, I would like to make
sure that it will be update-safe: It certainly will happen that I can
only update the Slurm installation on _one_ cluster at a time, maybe
with e.g. a week in between.
Is it a design principle that
To be clear, we aren't linking to libslurm at all. The issue is that libpmi
is linking to it, and we link to libpmi. So I think you have to recompile
to get the link dependencies correctly setup.
On Thu, Apr 16, 2015 at 5:32 AM, Uwe Sauter uwe.sauter...@gmail.com wrote:
Hi Ralph,
beside the
It has been our experience that it is necessary to rebuild OpenMPI
for each major slurm release, such as transitioning from Slurm
14.03.x to 14.11.x.
Andy
On 04/16/2015 07:49 AM, Ralph Castain
wrote:
Re: [slurm-dev] Re: Need for recompiling openmpi built with
--with-pmi?
Recompiling openmpi is sufficient, unless something else has changed in
openmpi that might require your programs to be rebuilt.
On 04/16/2015 09:17 AM, Bjørn-Helge Mevik wrote:
Andy Riebs andy.ri...@hp.com writes:
It has been our experience that it is necessary to rebuild OpenMPI
for
To all involved in this thread,
Thank you very much for your pointers and suggestions!
John DeSantis
2015-04-16 1:07 GMT-04:00 Christopher Samuel sam...@unimelb.edu.au:
On 16/04/15 14:43, Bill Barth wrote:
That's what I sent John off-list. Wasn't sure self-promotion was OK here.
I can't
Andy Riebs andy.ri...@hp.com writes:
It has been our experience that it is necessary to rebuild OpenMPI
for each major slurm release, such as transitioning from Slurm
14.03.x to 14.11.x.
Do you also need to recompile programs that are compiled with openmpi,
or is it enough to recompile
But if libpmi.so is provided by Slurm, why do I get the error messages? Does
OpenMPI statically link libpmi.a which then depends
on an older version of libslurm.so?
If OpenMPI dynamically links agains libpmi.so which itself either links against
libslurm.so or libslurm.so.28, shouldn't this
Yes Ulf, your understanding is correct. The DBD can talk to 2 previous
versions. As long as the DBD is updated first you should be fine.
On April 16, 2015 3:33:22 AM PDT, Ulf Markwardt ulf.markwa...@tu-dresden.de
wrote:
Dear Slurm developers,
before I set up Slurm with multi-cluster
14 matches
Mail list logo