Re: [OMPI users] Segmentation fault with SLURM and non-local nodes

2011-02-06 Thread Michael Curtis
On 07/02/2011, at 12:36 PM, Michael Curtis wrote: > > On 04/02/2011, at 9:35 AM, Samuel K. Gutierrez wrote: > > Hi, > >> I just tried to reproduce the problem that you are experiencing and was >> unable to. >> >> SLURM 2.1.15 >> Open MPI 1.4.3 configured with: >> --with-platform=./contrib/p

Re: [OMPI users] Segmentation fault with SLURM and non-local nodes

2011-02-06 Thread Michael Curtis
On 04/02/2011, at 9:35 AM, Samuel K. Gutierrez wrote: Hi, > I just tried to reproduce the problem that you are experiencing and was > unable to. > > SLURM 2.1.15 > Open MPI 1.4.3 configured with: > --with-platform=./contrib/platform/lanl/tlcc/debug-nopanasas I compiled OpenMPI 1.4.3 (vanilla

Re: [OMPI users] calling a customized MPI_Allreduce with MPI_PACKED datatype

2011-02-06 Thread Massimo Cafaro
Jody, I have finally solved the problem. I was incorrectly passing 1 as the count argument for MPI_Allreduce, reasoning that I was passing to the reduction one buffer of type MPI_PACKED. Instead, I need to pass the actual size of the buffer for the reduction to work correctly. Thank you anyway

Re: [OMPI users] calling a customized MPI_Allreduce with MPI_PACKED datatype

2011-02-06 Thread Massimo Cafaro
hi Jody, yes, the buffer sizes are actually quite huge with regard to the actual data to be packed.unpacked. Massimo On Feb 6, 2011, at 10:44 PM, jody wrote: > Hi Massimo > > Just to make sure: usually the MPI_ERR_TUNCATE error is caused by > buffer sizes that are too small. > Can you verify

Re: [OMPI users] calling a customized MPI_Allreduce with MPI_PACKED datatype

2011-02-06 Thread jody
Hi Massimo Just to make sure: usually the MPI_ERR_TUNCATE error is caused by buffer sizes that are too small. Can you verify that the buffers you are using are large enough to hold the data they should receive? Jody On Sat, Feb 5, 2011 at 6:37 PM, Massimo Cafaro wrote: > Dear all, > > in one

Re: [OMPI users] Default hostfile not being used by mpirun

2011-02-06 Thread Barnet Wagman
Setting the orte_default_hostfile param in *$HOME/.openmpi/mca-params.conf* works (with $HOME set, of course), but for some reason setting it in the system conf file, *$prefix/etc/openmpi-mca-params.conf* does not. Using 'ompi_info --param ...', it appear that the the system conf file

Re: [OMPI users] Default hostfile not being used by mpirun

2011-02-06 Thread Barnet Wagman
Setting the orte_default_hostfile param in *$HOME/.openmpi/mca-params.conf* works (with $HOME set, of course), but for some reason setting it in the system conf file, *$prefix/etc/openmpi-mca-params.conf* does not. Using 'ompi_info --param ...', it appear that the the system conf file

Re: [OMPI users] Segmentation fault with SLURM and non-local nodes

2011-02-06 Thread Michael Curtis
On 04/02/2011, at 9:35 AM, Samuel K. Gutierrez wrote: > I just tried to reproduce the problem that you are experiencing and was > unable to. > > > SLURM 2.1.15 > Open MPI 1.4.3 configured with: > --with-platform=./contrib/platform/lanl/tlcc/debug-nopanasas > > I'll dig a bit further. Intere