Intel® Fortran Compiler
Build applications that can scale for the future with optimized code designed for Intel® Xeon® and compatible processors.

smeagol error

pankajd
Beginner
736 Views
hi i get following error when i submit job

0: [cli_0]: aborting job:
0: Fatal error in MPI_Comm_group: Invalid communicator, error stack:
0: MPI_Comm_group(148): MPI_Comm_group(comm=0x0, group=0x7fff0b7d40e4) failed
0: MPI_Comm_group(72).: Invalid communicator
1: [cli_1]: aborting job:
1: Fatal error in MPI_Comm_group: Invalid communicator, error stack:
1: MPI_Comm_group(148): MPI_Comm_group(comm=0x0, group=0x7fff744785e4) failed
1: MPI_Comm_group(72).: Invalid communicatorsiesta: System type = bulk
0:
0: * Maximum dynamic memory allocated = 2 MB
0:
0: siesta: ===============================
0: Begin CG move = 0
0: ===============================
0:
0: iodm: Reading Density Matrix from files
0:
0: InitMesh: MESH = 48 x 48 x 48 = 110592
0: InitMesh: Mesh cutoff (required, used) = 40.000 53.991 Ry
0:
0: * Maximum dynamic memory allocated = 1is 7 MB
rank 1 in job 1 yn149.cn.yuva.param_46209 caused collective abort of all ranks
exit status of rank 1: killed by signal 9
----
my command is

/opt/intel/mpi/bin64/mpirun --totalnum=$NPROCS --file=$PBS_NODEFILE --rsh=/usr/bin/ssh -1 --ordered --verbose -l -machinefile $PBS_NODEFILE -np $NPROCS /home/
gjkl/tfh/bhb/smeagol.1.0b/Src/smeagol < sih.fdf > sih.smeagol.out

0 Kudos
1 Reply
Ron_Green
Moderator
736 Views
Quoting - pankajd
hi i get following error when i submit job

0: [cli_0]: aborting job:
0: Fatal error in MPI_Comm_group: Invalid communicator, error stack:
0: MPI_Comm_group(148): MPI_Comm_group(comm=0x0, group=0x7fff0b7d40e4) failed
0: MPI_Comm_group(72).: Invalid communicator
1: [cli_1]: aborting job:
1: Fatal error in MPI_Comm_group: Invalid communicator, error stack:
1: MPI_Comm_group(148): MPI_Comm_group(comm=0x0, group=0x7fff744785e4) failed
1: MPI_Comm_group(72).: Invalid communicatorsiesta: System type = bulk
0:
0: * Maximum dynamic memory allocated = 2 MB
0:
0: siesta: ===============================
0: Begin CG move = 0
0: ===============================
0:
0: iodm: Reading Density Matrix from files
0:
0: InitMesh: MESH = 48 x 48 x 48 = 110592
0: InitMesh: Mesh cutoff (required, used) = 40.000 53.991 Ry
0:
0: * Maximum dynamic memory allocated = 1is 7 MB
rank 1 in job 1 yn149.cn.yuva.param_46209 caused collective abort of all ranks
exit status of rank 1: killed by signal 9
----
my command is

/opt/intel/mpi/bin64/mpirun --totalnum=$NPROCS --file=$PBS_NODEFILE --rsh=/usr/bin/ssh -1 --ordered --verbose -l -machinefile $PBS_NODEFILE -np $NPROCS /home/
gjkl/tfh/bhb/smeagol.1.0b/Src/smeagol < sih.fdf > sih.smeagol.out


There are many possible problems, it is difficult to say. What is clear is that the call to MPI_COMM_GROUP does not like the first argument you are giving it, which is the communicator. Typically MPI_COMM_WORLD, is this the communicator you are using? If so, does your code have

INCLUDE 'mpif.h'

so that MPI_COMM_WORLD is defined properly? You may consider adding
implicit none

to determine if your communicator is defined properly.

ron
0 Kudos
Reply