On Mon, Sep 29, 2008 at 3:05 PM, Luke Tierney <luke at stat.uiowa.edu> wrote:
Look at the configure output below: configure is finding mpi.h in
/usr/lib/mpich/include, soit is using mpich, not lam. You may be able
to tell Rmpi specifically where to look for LAM stuff or you may need
to uninstall mpich (almost surely that can be avoided but it may be
easier if you don't really need mpich).
OK - I don't mind if I am using lam or mpich, so which one should I use?
I am confused.
If it helps, I am using Ubuntu Hardy
Rainer
Best,
luke
On Mon, 29 Sep 2008, Rainer M Krug wrote:
On Mon, Sep 29, 2008 at 12:19 AM, Martin Morgan <mtmorgan at fhcrc.org>
wrote:
"Rainer M Krug" <r.m.krug at gmail.com> writes:
Hi
I am trying to utilize my dual core processor (and later a
High-performance clusters (HPC) ) by using the Rmpi, snow, snowfall,
... packages, but I am struggling at the beginning, i.e. to initialise
the "cluster" on my dual core computer. Whenever I try to initialize
it (via sfInit(parallel=TRUE, cpus=2) or mpi.spawn.Rslaves(nslaves=2)
), I get an error message:
sfInit(parallel=TRUE, cpus=2)
Forced parallel. Using session: XXXXXXXXR_rkrug_143706_092708
Error in mpi.comm.spawn(slave = mpitask, slavearg = args, nslaves =
count, :
MPI_Comm_spawn is not supported.
Error in sfInit(parallel = TRUE, cpus = 2) :
Starting of snow cluster failed! Error in mpi.comm.spawn(slave =
mpitask, slavearg = args, nslaves = count, :
MPI_Comm_spawn is not supported.
and
mpi.spawn.Rslaves(nslaves=2)
Error in mpi.spawn.Rslaves(nslaves = 2) :
You cannot use MPI_Comm_spawn API
This error comes from Rmpi,
head(mpi.spawn.Rslaves, 6)
1 function (Rscript = system.file("slavedaemon.R", package = "Rmpi"),
2 nslaves = mpi.universe.size(), root = 0, intercomm = 2, comm = 1,
3 hosts = NULL, needlog = TRUE, mapdrive = TRUE)
4 {
5 if (!is.loaded("mpi_comm_spawn"))
6 stop("You cannot use MPI_Comm_spawn API")
and occurs when the compiler variable MPI2 is undefined when your
package is installed. Likely this means that your mpi installation is
either old (unlikely?) or that your Rmpi installation failed to
properly detect the installed mpi version. It's difficult to know
which, without more information on how the mpi and Rmpi installations
went, at a minimum the result of R's sessionInfo() command and mpirun
--version but likely including the output of Rmpi's installation.
Version:
lamboot -V
LAM 7.1.2/MPI 2 C++/ROMIO - Indiana University
Arch: i486-pc-linux-gnu
Prefix: /usr/lib/lam
Configured by: buildd
Configured on: Sun Mar 23 08:07:16 UTC 2008
Configure host: rothera
SSI rpi: crtcp lamd sysv tcp usysv
Below find the session info and the info from the instalaltion of Rmpi.
In the install log from Rmpi, it says:
checking whether MPICH2 is declared... no
checking whether MPICH2 is declared... (cached) no
So what went wrong? Do I have to start the "cluster" on the dual-core
machine?
Can I set the compiler variable MPI2 manually?
_
platform i486-pc-linux-gnu
arch i486
os linux-gnu
system i486, linux-gnu
status
major 2
minor 7.2
year 2008
month 08
day 25
svn rev 46428
language R
version.string R version 2.7.2 (2008-08-25)
R version 2.7.2 (2008-08-25)
i486-pc-linux-gnu
locale:
LC_CTYPE=en_ZA.UTF-8;LC_NUMERIC=C;LC_TIME=en_ZA.UTF-8;LC_COLLATE=en_ZA.UTF-8;LC_MONETARY=C;LC_MESSAGES=en_ZA.UTF-8;LC_PAPER=en_ZA.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_ZA.UTF-8;LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] snow_0.3-3 Rmpi_0.5-5 snowfall_1.53
loaded via a namespace (and not attached):
[1] tools_2.7.2
Warning in install.packages(c("Rmpi"), dep = TRUE, repo =
"http://cbio.uct.ac.za/CRAN/") :
argument 'lib' is missing: using '/usr/local/lib/R/site-library'
trying URL 'http://cbio.uct.ac.za/CRAN/src/contrib/Rmpi_0.5-5.tar.gz'
Content type 'application/x-gzip' length 94643 bytes (92 Kb)
opened URL
==================================================
downloaded 92 Kb
* Installing *source* package 'Rmpi' ...
checking for gcc... gcc
checking for C compiler default output file name... a.out
checking whether the C compiler works... yes
checking whether we are cross compiling... no
checking for suffix of executables...
checking for suffix of object files... o
checking whether we are using the GNU C compiler... yes
checking whether gcc accepts -g... yes
checking for gcc option to accept ISO C89... none needed
I am here /usr/lib/mpich and it is MPICH
Try to find mpi.h ...
Found in /usr/lib/mpich/include
Try to find libmpi.so or libmpich.a
Found libmpich in /usr/lib/mpich/lib
##########################################
checking whether MPICH2 is declared... no
checking whether MPICH2 is declared... (cached) no
##########################################
checking for openpty in -lutil... yes
checking for main in -lpthread... yes
configure: creating ./config.status
config.status: creating src/Makevars
** libs
gcc -std=gnu99 -I/usr/share/R/include -DPACKAGE_NAME=\"\"
-DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\"
-DPACKAGE_BUGREPORT=\"\" -DHAVE_DECL_MPICH2=0 -DHAVE_DECL_MPICH2=0
-I/usr/lib/mpich/include -DMPICH -fPIC -fpic -g -O2 -c
conversion.c -o conversion.o
gcc -std=gnu99 -I/usr/share/R/include -DPACKAGE_NAME=\"\"
-DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\"
-DPACKAGE_BUGREPORT=\"\" -DHAVE_DECL_MPICH2=0 -DHAVE_DECL_MPICH2=0
-I/usr/lib/mpich/include -DMPICH -fPIC -fpic -g -O2 -c
internal.c -o internal.o
gcc -std=gnu99 -I/usr/share/R/include -DPACKAGE_NAME=\"\"
-DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\"
-DPACKAGE_BUGREPORT=\"\" -DHAVE_DECL_MPICH2=0 -DHAVE_DECL_MPICH2=0
-I/usr/lib/mpich/include -DMPICH -fPIC -fpic -g -O2 -c
RegQuery.c -o RegQuery.o
gcc -std=gnu99 -I/usr/share/R/include -DPACKAGE_NAME=\"\"
-DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\"
-DPACKAGE_BUGREPORT=\"\" -DHAVE_DECL_MPICH2=0 -DHAVE_DECL_MPICH2=0
-I/usr/lib/mpich/include -DMPICH -fPIC -fpic -g -O2 -c Rmpi.c -o
Rmpi.o
gcc -std=gnu99 -shared -o Rmpi.so conversion.o internal.o RegQuery.o
Rmpi.o -L/usr/lib/mpich/lib -lmpich -lutil -lpthread -fPIC
-L/usr/lib/R/lib -lR
** R
** demo
** inst
** preparing package for lazy loading
** help
Building/Updating help pages for package 'Rmpi'
Formats: text html latex example
hosts text html latex
internal text html latex
mpi.abort text html latex
mpi.apply text html latex example
mpi.barrier text html latex
mpi.bcast text html latex
mpi.bcast.Robj text html latex
mpi.bcast.cmd text html latex
mpi.cart.coords text html latex example
mpi.cart.create text html latex example
mpi.cart.get text html latex example
mpi.cart.rank text html latex example
mpi.cart.shift text html latex example
mpi.cartdim.get text html latex example
mpi.comm text html latex example
mpi.comm.disconnect text html latex
mpi.comm.free text html latex
mpi.comm.inter text html latex
mpi.comm.set.errhandler text html latex
mpi.comm.spawn text html latex
mpi.const text html latex
mpi.dims.create text html latex example
mpi.exit text html latex
mpi.finalize text html latex
mpi.gather text html latex example
mpi.gather.Robj text html latex example
mpi.get.count text html latex
mpi.get.processor.name text html latex
mpi.get.sourcetag text html latex
mpi.info text html latex
mpi.init.sprng text html latex
mpi.intercomm.merge text html latex
mpi.parSim text html latex
mpi.parapply text html latex example
mpi.probe text html latex
mpi.realloc text html latex
mpi.reduce text html latex
mpi.remote.exec text html latex example
mpi.scatter text html latex example
mpi.scatter.Robj text html latex example
mpi.send text html latex example
mpi.send.Robj text html latex
mpi.sendrecv text html latex example
mpi.setup.rng text html latex
mpi.spawn.Rslaves text html latex example
mpi.universe.size text html latex
mpi.wait text html latex
** building package indices ...
* DONE (Rmpi)
The downloaded packages are in
/tmp/RtmpdX6zlZ/downloaded_packages
Dirk Eddelbuettel
(https://stat.ethz.ch/pipermail/r-devel/2008-September/050665.html)
suggested that snow's makeSOCKcluster is an easier starting point for
single computer 'clusters' or other configurations where significant
system administration is not desired -- these should work without
additional software on most systems, even if more limiting in the long
term (in my opinion). See, e.g., the examples on the help page for
clusterApply for basic operation. A bit oddly, apparently the
'snowfall' package restricts snow's functionality to mpi clusters. So
you might start with snow directly.
Martin
I followed the pdf "Developing parallel programs using snowfall" by
Jochen Knaus,. installed the relevant libraries and programs, but it
does not work.
I am stuck.
Any help appreciated,
Rainer
--
Rainer M. Krug, PhD (Conservation Ecology, SUN), MSc (Conservation
Biology, UCT), Dipl. Phys. (Germany)
Centre of Excellence for Invasion Biology
Faculty of Science
Natural Sciences Building
Private Bag X1
University of Stellenbosch
Matieland 7602
South Africa