[Wrf-users] run time error - collective abort of all ranks

[Jorge Alejandro Arevalo Borquez]

Hi,

do you could send rsl.out.0000 y rsl.error.0000 files?
why you are using DM and not SM? my experiencie indicate better performance
with SM in a system similiar which you describe.


Atentamente
Jorge Arévalo Bórquez


On Thu, Mar 12, 2009 at 3:18 PM, Gerardo De Jesus Montoya Gaviria <
gdmontoyag at unal.edu.co> wrote:

> I'm running WRF 3.0.1.1 in a Dell precision 690 machine with two Xeon
> 64bits dual-core processors, with ifort 10.1 and mpich2 for a domain 118x118
> points with dx=dy= 18.9km. The compilation after a lot of time (1 hour
> aprox.) finishes succesfully. the run finish with error message almost
> immediately after the comand: mpiexec -n 4 ./wrf.exe.
> We also used the comand: ulimit -s unlimited.  The erro message is:
>
> ************************************************************
> mpiexec -n 4 ./wrf.exe
>  starting wrf task            0  of            4
>  starting wrf task            1  of            4
>  starting wrf task            3  of            4
>  starting wrf task            2  of            4
> rank 2 in job 2  localhost.localdomain_53980   caused collective abort of
> all ranks
>  exit status of rank 2: killed by signal 11
> rank 1 in job 2  localhost.localdomain_53980   caused collective abort of
> all ranks
>  exit status of rank 1: killed by signal 11
> *************************************************************
> The configuration we are using is:
>
>  7.  Linux x86_64 i486 i586 i686, ifort compiler with icc  (dmpar)
>  Compile for nesting: 1=basic
>
> OS: Scientific Linux 5.1 64Bits
>
> We found similar problems posted in the User forum, but without a clear
> response. I'll greatly appreciate any help.
> Gerardo Montoya. Profesor titular, Universidad nacional de Colombia.
>
> _______________________________________________
> Wrf-users mailing list
> Wrf-users at ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20090313/eb859d56/attachment.html