I will add to what Terry said by mentioning
that the MPI implementation has no awareness of ordinary POSIX or Fortran
disk I/O routines. It cannot help on those.
Any automated help the MPI implementation
can provide would only apply to MPI_File_xxx disk I/O. These
are implemented by the MPI library.
It is possible for MPI-IO to be implemented
in a way that lets a single process or the set of process on a node act
as the disk i/O agents for the entire job but someone else will need to
tell you if OpenMPI can do this, I think OpenMPI built on the ROMIO
MPI-IO implementation and based on my outdated knowledge of ROMIO, I would
be a bit surprised if it has his option.
Dick Treumann - MPI Team
IBM Systems & Technology Group
Dept X2ZA / MS P963 -- 2455 South Road -- Poughkeepsie, NY 12601
Tele (845) 433-7846 Fax (845) 433-8363
From:
Terry Frankcombe <terry@chem.gu.se>
To:
Open MPI Users <users@open-mpi.org>
Date:
09/29/2010 09:50 PM
Subject:
Re: [OMPI users] a question about [MPI]IO
on systems without
network filesystem
Sent by:
users-bounces@open-mpi.org
Hi Paul
I think you should clarify whether you mean you want you application to
send all it's data back to a particular rank, which then does all IO (in
which case the answer is any MPI implementation can do this... it's a
matter of how you code the app), or if you want the application to know
nothing about it, but have the system somehow intercept all IO and make
it magically appear at a particular node (much harder).
On Wed, 2010-09-29 at 11:34 +0200, Paul Kapinos wrote:
> Dear OpenMPI developer,
>
> We have a question about the possibility to use MPI IO (and possible
> regular I/O) on clusters which does *not* have a common filesystem
> (network filesystem) on all nodes.
>
> A common filesystem is mainly NOT a hard precondition to use OpenMPI:
> http://www.open-mpi.org/faq/?category=running#do-i-need-a-common-filesystem
>
>
> Say, we have a (diskless? equipped with very small disks?) cluster,
on
> which only one node have access to a filesystem.
>
> Is it possible to configure/run OpenMPI in a such way, that only _one_
> process (e.g. master) performs real disk I/O, and other processes
sends
> the data to the master which works as an agent?
>
> Of course this would impacts the performance, because all data must
be
> send over network, and the master may became a bottleneck. But is
such
> scenario - IO of all processes bundled to one process - practicable
at all?
>
>
> Best wishes
> Paul
>
>
>
> _______________________________________________
> users mailing list
> users@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users