What does this error mean?

[Thomas Zeiser]

Hi Andrey,
hi Kevin,

On Wed, Jan 30, 2008 at 07:17:36PM +0300, Derbunovich, Andrey wrote:
> Kevin,
> 
> Thank you. Actually we never test our library with this mpiexec version.

we have Intel MPI 3.1.026 successfully running with torque-2.1.x
and Pete's mpiexec-0.82
(there are only sometimes timeout problem if quite large Infiniband
jobs start - but that's an other storry)

> Did you able to use this mpiexec with earlier versions of the Intel MPI
> Library?
> 
> Best regards,
> Andrey

> > > > > > mpiexec: read_keyvals: keyval 1 key kvsname val
> > 53001.jman-spawn-0.
> > > > > > mpiexec: read_keyvals: keyval 2 key key val DAPL_PROVIDER.
> > > > > > mpiexec: read_keyvals: keyval 3 key value val <NULL.

Kevin: what interconnect are you trying to use?

as "DAPL_PROVIDER" appears here, it looks for me that you use
Infiniband or Myrinet and not plain ethernet - thus giving
additional potential for failures.

Can you check if your job starts fine if you force Intel MPI to the
socks device by "export I_MPI_DEVICE=ssm".


thomas
-- 
Thomas Zeiser, HPC Services
Friedrich-Alexander-Universitaet Erlangen-Nuernberg
Regionales Rechenzentrum Erlangen (RRZE)
Martensstrasse 1, 91058 Erlangen, Germany
Tel.: +49 9131 85-28737, Fax: +49 9131 302941
thomas.zeiser at rrze.uni-erlangen.de
http://www.rrze.uni-erlangen.de/hpc/