Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |  

This web mail archive is frozen.

This page is part of a frozen web archive of this mailing list.

You can still navigate around this archive, but know that no new mails have been added to it since July of 2016.

Click here to be taken to the new web archives of this list; it includes all the mails that are in this frozen archive plus all new mails that have been sent to the list since it was migrated to the new archives.

Subject: Re: [OMPI users] mpirun runs in serial even I set np to several processors
From: Gus Correa (gus_at_[hidden])
Date: 2014-04-14 15:51:24


On 04/14/2014 01:15 PM, Djordje Romanic wrote:
> Hi,
>
> I am trying to run WRF-ARW in parallel. This is configuration of my system:
> -------------------------------------------------
> Architecture: x86_64
> CPU op-mode(s): 32-bit, 64-bit
> Byte Order: Little Endian
> CPU(s): 4
> On-line CPU(s) list: 0-3
> Thread(s) per core: 1
> Core(s) per socket: 4
> Socket(s): 1
> NUMA node(s): 1
> Vendor ID: AuthenticAMD
> CPU family: 16
> Model: 2
> Stepping: 3
> CPU MHz: 1150.000
> BogoMIPS: 4587.84
> Virtualization: AMD-V
> L1d cache: 64K
> L1i cache: 64K
> L2 cache: 512K
> L3 cache: 2048K
> NUMA node0 CPU(s): 0-3
> -------------------------------------------------
>
> When I start wrf with mpirun -np 4 ./wrf.exe, I get this:
> -------------------------------------------------
> starting wrf task 0 of 1
> starting wrf task 0 of 1
> starting wrf task 0 of 1
> starting wrf task 0 of 1
> -------------------------------------------------
> This indicates that it is not using 4 processors, but 1.
>
> Any idea what might be the problem?
>
> Thanks,
> Djordje
>

Did you compile WRF with MPI enabled (i.e. serial,
not OpenMP, or openMP only, which is a different beast than OpenMPI =
note the trailing "I") ?

If you compiled serial, or OpenMP and did not set OMP_NUM_THREADS,
the mpirun command above will launch 4
separate/independent/repetitive/non-communicating processes
doing the same exact thing (and probably overwriting each other's
output, etc).

Did you compile and link to the OpenMPI libraries, or perhaps to
another MPI implementation?

I hope this helps,
Gus Correa