klinck@ccpo_odu_edu
New Member
I am a new user of the cesm code. I need to run standalone ocean cases to set up the ocean model for a southern ocean coupled calculation as part of an EaSM project.
I have the cesm1_0_4 code. (I tried to use the cesm1_1 code, but the generic machine files are not present). I have set up the compset X as described in the users guide. I am running the code on a linux computer with 4 dual core processers (a sunfire x2200). I am using mpi (mpich2). I have the intel compilers and mpi was compiled with the intel compilers.
The model build finishes without error. I have submitted the job with mpiexec using 6 processors.
mpiexec -n 6 -f $CASEROOT/charybdisfile ./ccsm.exe >&! ccsm.log.$LID
The run finishes with a segment fault (see part of the run log below). I can not find any information in the logs what process segment faults. I suspect that I have a problem with the distribution of processes so I have included the task distribution from the pes file. I am sending small slices of the output files to reduce the bulk, but I will send copies of whatever files might be useful. None of the individual model log files have any indication of an error.
I am hoping for suggestions of what to change, or at least where to find what went wrong.
Thanks in advance for your suggestions.
John
lines from env_mach_pes.xml (modified from the distribution with xmlchange)
ccsm.log
(deleted lines)
(seq_mct_drv) : Initialize atm component
(seq_mct_drv) : Initialize lnd component
(seq_mct_drv) : Initialize ocn component
(seq_mct_drv) : Initialize ice component
(seq_mct_drv) : Initialize glc component
=====================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= EXIT CODE: 139
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
=====================================================================================
APPLICATION TERMINATED WITH THE EXIT STRING: Segmentation fault (signal 11)
I have the cesm1_0_4 code. (I tried to use the cesm1_1 code, but the generic machine files are not present). I have set up the compset X as described in the users guide. I am running the code on a linux computer with 4 dual core processers (a sunfire x2200). I am using mpi (mpich2). I have the intel compilers and mpi was compiled with the intel compilers.
The model build finishes without error. I have submitted the job with mpiexec using 6 processors.
mpiexec -n 6 -f $CASEROOT/charybdisfile ./ccsm.exe >&! ccsm.log.$LID
The run finishes with a segment fault (see part of the run log below). I can not find any information in the logs what process segment faults. I suspect that I have a problem with the distribution of processes so I have included the task distribution from the pes file. I am sending small slices of the output files to reduce the bulk, but I will send copies of whatever files might be useful. None of the individual model log files have any indication of an error.
I am hoping for suggestions of what to change, or at least where to find what went wrong.
Thanks in advance for your suggestions.
John
lines from env_mach_pes.xml (modified from the distribution with xmlchange)
ccsm.log
(deleted lines)
(seq_mct_drv) : Initialize atm component
(seq_mct_drv) : Initialize lnd component
(seq_mct_drv) : Initialize ocn component
(seq_mct_drv) : Initialize ice component
(seq_mct_drv) : Initialize glc component
=====================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= EXIT CODE: 139
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
=====================================================================================
APPLICATION TERMINATED WITH THE EXIT STRING: Segmentation fault (signal 11)