hi,
thanks for the info.
we were no aware that you switch the executable name in the gui based on
the mpi settings.
we use a script template which ignores the gui mpi settings and take
instead the
mpiexec -np $NSLOTS XXX...
from the gridengine 2011 exec daemon.
cheers,
wolfgang
On 02/03/2014 01:12 PM, Sjors Scheres wrote:
> Hi Wolfgang,
> Make sure you are using the parallel executable relion_refine_mpi (and
> not the sequential executable in relion_refine). Also, make sure to
> use at least 3 MPI processes.
> HTH, S
>
> On 02/03/2014 11:53 AM, Wolfgang Lugmayr wrote:
>> hi,
>>
>> when using autorefine 3D we get the following error:
>> ERROR: Cannot split data into random halves without using MPI!
>> File: src/ml_optimiser.cpp line: 1266
>>
>> the program has been started with mpiexec as usual?
>>
>> we use openmpi 1.6.5 and relion 1.2
>>
>> the error comes with gcc and intel icc binaries
>>
>> gcc compilation:
>> CFLAGS="-O2" CXXFLAGS="-O2" \
>> ./INSTALL.sh -j $BUILD_CPUS 2>&1 | tee install.log
>>
>> icc compilation:
>> CXX=icpc CXXFLAGS="-O3" \
>> CC="icc" CFLAGS="-O3" \
>> F77=ifort FC=ifort \
>> ./INSTALL.sh -j $BUILD_CPUS 2>&1 | tee install.log
>>
>> cheers,
>> wolfgang
>>
>
--
Universitätsklinikum Hamburg-Eppendorf (UKE)
Institut für Struktur und Systembiologie
@ Institute of Molecular Biotechnology (IMBA)
Dr. Bohr-Gasse 3-7
1030 Vienna, Austria
Tel.: +43 (1) 790 44-4649
Email: [log in to unmask]
--
Besuchen Sie uns auf: www.uke.de
_____________________________________________________________________
Universitätsklinikum Hamburg-Eppendorf; Körperschaft des öffentlichen Rechts; Gerichtsstand: Hamburg
Vorstandsmitglieder: Prof. Dr. Christian Gerloff (Vertreter des Vorsitzenden), Prof. Dr. Dr. Uwe Koch-Gromus, Joachim Prölß, Rainer Schoppik
_____________________________________________________________________
SAVE PAPER - THINK BEFORE PRINTING
|