[GE users] SGE+OpenMPI: ERROR: A daemon on node xyz failed to start as expected
Azhar Ali Shah
aas_lakyari at yahoo.com
Thu Jul 3 18:50:59 BST 2008
Hi Reuti and Joe,
Having tried different installations and playing with different nodes I have come to know that if on any node I run a MPICH2 compiled porgram using Open MPI, I get following message:
[taramel:04844] *** An error occurred in MPI_Send
[taramel:04844] *** on communicator MPI_COMM_WORLD
[taramel:04844] *** MPI_ERR_RANK: invalid rank
[taramel:04844] *** MPI_ERRORS_ARE_FATAL (goodbye)
Now if I uninstall the Open MPI using make uninstall and make clean, restart the system and install the Open MPI again, compile the program with Open MPI and then use the mpirun with Open MPI even then I get same error.
Though Reuti pointed out to followup error but I could not undertand how to get rid of it. This has left me wondering without sucess. Any thoughts please?
PS: I have checked the .rhost file for firewall related issues but it all seems OK.
--- On Wed, 7/2/08, Joe Landman <landman at scalableinformatics.com> wrote:
From: Joe Landman <landman at scalableinformatics.com>
Subject: Re: [GE users] SGE+OpenMPI: ERROR: A daemon on node xyz failed to start as expected
To: users at gridengine.sunsource.net
Date: Wednesday, July 2, 2008, 1:43 PM
Azhar Ali Shah wrote:
> I configured OpenMPI using --prefix and --with-sge parameters
> followed by 'make all install'
> Given the above message, should I 'recompile Open MPI with the
> configure option --enable-heterogeneous' as my environemnt is so?
I would simply suggest disabling progress threads (and threading in
general in MPI) unless you have a really good reason why you need them.
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics LLC,
email: landman at scalableinformatics.com
web : http://www.scalableinformatics.com
phone: +1 734 786 8423
fax : +1 866 888 3112
cell : +1 734 612 4615
To unsubscribe, e-mail: users-unsubscribe at gridengine.sunsource.net
For additional commands, e-mail: users-help at gridengine.sunsource.net
More information about the gridengine-users