Hi,
yesterday I installed openmpi-1.9a1r27362 on Solaris and Linux and
I have a problem with mpiJava on Linux (openSUSE-Linux 12.1, x86_64).
linpc4 mpi_classfiles 104 javac HelloMainWithoutMPI.java
linpc4 mpi_classfiles 105 mpijavac HelloMainWithBarrier.java
linpc4 mpi_classfiles 106 mpijavac -showme
/usr/local/jdk1.7.0_07-64/bin/javac \
-cp ...:.:/usr/local/openmpi-1.9_64_cc/lib64/mpi.jar
It works with Java without MPI.
linpc4 mpi_classfiles 107 mpiexec java -cp $HOME/mpi_classfiles \
HelloMainWithoutMPI
Hello from linpc4.informatik.hs-fulda.de/193.174.26.225
It breaks with Java and MPI.
linpc4 mpi_classfiles 108 mpiexec java -cp $HOME/mpi_classfiles \
HelloMainWithBarrier
--------------------------------------------------------------------------
It looks like opal_init failed for some reason; your parallel process is
likely to abort. There are many reasons that a parallel process can
fail during opal_init; some of which are due to configuration or
environment problems. This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):
mca_base_open failed
--> Returned value -2 instead of OPAL_SUCCESS
--------------------------------------------------------------------------
--------------------------------------------------------------------------
It looks like orte_init failed for some reason; your parallel process is
likely to abort. There are many reasons that a parallel process can
fail during orte_init; some of which are due to configuration or
environment problems. This failure appears to be an internal failure;
here's some additional information (which may only be relevant to an
Open MPI developer):
opal_init failed
--> Returned value Out of resource (-2) instead of ORTE_SUCCESS
--------------------------------------------------------------------------
--------------------------------------------------------------------------
It looks like MPI_INIT failed for some reason; your parallel process is
likely to abort. There are many reasons that a parallel process can
fail during MPI_INIT; some of which are due to configuration or environment
problems. This failure appears to be an internal failure; here's some
additional information (which may only be relevant to an Open MPI
developer):
ompi_mpi_init: orte_init failed
--> Returned "Out of resource" (-2) instead of "Success" (0)
--------------------------------------------------------------------------
*** An error occurred in MPI_Init
*** on a NULL communicator
*** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
*** and potentially your MPI job)
[linpc4:15332] Local abort before MPI_INIT completed successfully; not able to
aggregate error messages, and not able to guarantee that all other processes were
killed!
-------------------------------------------------------
Primary job terminated normally, but 1 process returned
a non-zero exit code.. Per user-direction, the job has been aborted.
-------------------------------------------------------
--------------------------------------------------------------------------
mpiexec detected that one or more processes exited with non-zero status, thus
causing
the job to be terminated. The first process to do so was:
Process name: [[58875,1],0]
Exit code: 1
--------------------------------------------------------------------------
I configured with the following command.
../openmpi-1.9a1r27362/configure --prefix=/usr/local/openmpi-1.9_64_cc \
--libdir=/usr/local/openmpi-1.9_64_cc/lib64 \
--with-jdk-bindir=/usr/local/jdk1.7.0_07-64/bin \
--with-jdk-headers=/usr/local/jdk1.7.0_07-64/include \
JAVA_HOME=/usr/local/jdk1.7.0_07-64 \
LDFLAGS="-m64" \
CC="cc" CXX="CC" FC="f95" \
CFLAGS="-m64" CXXFLAGS="-m64 -library=stlport4" FCFLAGS="-m64" \
CPP="cpp" CXXCPP="cpp" \
CPPFLAGS="" CXXCPPFLAGS="" \
C_INCL_PATH="" C_INCLUDE_PATH="" CPLUS_INCLUDE_PATH="" \
OBJC_INCLUDE_PATH="" OPENMPI_HOME="" \
--enable-cxx-exceptions \
--enable-mpi-java \
--enable-heterogeneous \
--enable-opal-multi-threads \
--enable-mpi-thread-multiple \
--with-threads=posix \
--with-hwloc=internal \
--without-verbs \
--without-udapl \
--with-wrapper-cflags=-m64 \
--enable-debug \
|& tee log.configure.$SYSTEM_ENV.$MACHINE_ENV.64_cc
It works fine on Solaris machines as long as the hosts belong to the
same kind (Sparc or x86_64).
tyr mpi_classfiles 194 mpiexec -host sunpc0,sunpc1,sunpc4 \
java -cp $HOME/mpi_classfiles HelloMainWithBarrier
Process 1 of 3 running on sunpc1
Process 2 of 3 running on sunpc4.informatik.hs-fulda.de
Process 0 of 3 running on sunpc0
sunpc4 fd1026 107 mpiexec -host tyr,rs0,rs1 \
java -cp $HOME/mpi_classfiles HelloMainWithBarrier
Process 1 of 3 running on rs0.informatik.hs-fulda.de
Process 2 of 3 running on rs1.informatik.hs-fulda.de
Process 0 of 3 running on tyr.informatik.hs-fulda.de
It breaks if the hosts belong to both kinds of machines.
sunpc4 fd1026 106 mpiexec -host tyr,rs0,sunpc1 \
java -cp $HOME/mpi_classfiles HelloMainWithBarrier
[rs0.informatik.hs-fulda.de:7718] *** An error occurred in MPI_Comm_dup
[rs0.informatik.hs-fulda.de:7718] *** reported by process [565116929,1]
[rs0.informatik.hs-fulda.de:7718] *** on communicator MPI_COMM_WORLD
[rs0.informatik.hs-fulda.de:7718] *** MPI_ERR_INTERN: internal error
[rs0.informatik.hs-fulda.de:7718] *** MPI_ERRORS_ARE_FATAL (processes
in this communicator will now abort,
[rs0.informatik.hs-fulda.de:7718] *** and potentially your MPI job)
[sunpc4.informatik.hs-fulda.de:07900] 1 more process has sent help
message help-mpi-errors.txt / mpi_errors_are_fatal
[sunpc4.informatik.hs-fulda.de:07900] Set MCA parameter
"orte_base_help_aggregate" to 0 to see all help / error messages
Please let me know if I can provide anything else to track these errors.
Thank you very much for any help in advance.
Kind regards
Siegmar
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}