[R-sig-hpc] Addendum to Rmpi installation

[Lockwood, Glenn]

Ben,

It looks like your cluster has Intel MPI and MVAPICH2 installed, neither of which is supported by Rmpi.  There is a blog post (see http://leftcensored.skepsi.net/2011/03/28/using-r-and-snow-on-ohio-supercomputer-centers-glenn-cluster/) on how you can try to get Rmpi to install using MVAPICH2, and Intel MPI is based on MVAPICH2 so you might want to give it a try.

With that being said, we've never been able to get Rmpi with mvapich2 to actually work here.  It compiles and installs, but upon loading it, it throws all sorts of awful errors.  We just use OpenMPI instead, which works a lot more reliably with Rmpi.

Glenn


--
Glenn K. Lockwood, Ph.D.
SDSC User Services
glock at sdsc.edu
(858) 246-1075

On Sep 19, 2013, at 5:58 AM, Ben Weinstein <benweinstein2010 at gmail.com>
 wrote:

> Hi all,
> 
> After further sleuthing, it would appears that many installations point to
> this directory, wherein i see many potential versions of mpi (mvapich2?).
> Which are these are most appropriate for the RMPI configure arguments?
> 
> Thanks
> 
> Ben
> 
> 
> login4$ cd /opt/apps/intel13/
> login4$ ls
> 
> boost
> hdf5
> modulefiles
> mvapich2-x
> parsimonator
> upc
> clustalw2
> impi
> mrbayes
> nco
> phylip
> freeglut
>  impi_4_1
> mvapich2
> ncview
> quicktree
> grvy
> impi_4_1_0_018
> mvapich2_1_9
> netcdf
> R_mkl
> gsl
> metis
> mvapich2-hecura
> opencv
> udunits
> 
> -- 
> Ben Weinstein
> PhD Candidate
> Ecology and Evolution
> Stony Brook University
> 
> http://benweinstein.weebly.com/
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-hpc mailing list
> R-sig-hpc at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-hpc