some files not found, setting default values in POP2 model

[yaozhixiong@foxmail_com]

Hi, I am trying to install cesm1_2_0 on a new machine. I crated a new case with compset 'X'. The case could be built successfully.Because the machine can only use lsf to submit jobs, I modified some places. I comment out PBS or BSUB, and add APP_NAME=intelg_small  #  queuename
NP=16                             # total number of nodes I also replace  #mpiexec -n 16 $EXEROOT/cesm.exe >&! cesm.log.$LID #mpirun -np 16 $EXEROOT/cesm.exe >&! cesm.log.$LID byRUN="/home-gg/users/nscc670_YZX/CESM/cesm120/caseoutput/testx5/bld/cesm.exe"Details can to seen in attachment testx5.run. After I submitted the job using 'bsub testx5.run', something wrong happened, although it seems that the model began to initialize. The error information says that:File drv_in not found, setting default valuesNo cpl_modelio.nml found, using defaults for pio settingsNo atm_modelio.nml found, using defaults for pio settingsNo lnd_modelio.nml found, using defaults for pio settingsNo ocn_modelio.nml found, using defaults for pio settingsNo ice_modelio.nml found, using defaults for pio settingsNo glc_modelio.nml found, using defaults for pio settingsNo rof_modelio.nml found, using defaults for pio settingsNo wav_modelio.nml found, using defaults for pio settingsforrtl: No such file or directory
seq_flds_mod: read seq_cplflds_inparm namelist from: drv_in
forrtl: severe (29): file not found, unit 90, file /home-gg/users/nscc670_YZX/CESM/cesm120/caseoutput/testx5/drv_inDetails can be seen in attachment. All of these unfound files are in RUNDIR, /home-gg/users/nscc670_YZX/CESM/cesm120/caseoutput/testx5/run. If I copy 'drv_in' into directory testx5, then model still wrong, and another file will not be found again. And the unfound file is still in RUNDIR. So, I think there must be some paths I should modify. Then model will find the files in RUNDIR. Any help would be greatly appreciated. Thanks.zhixiong

 

[jedwards]

For an example of using the LSF queing system look at the yellowstone machine definition.   You need mpirun.lsf in place of mpiexec, you also need to submit using stdio:bsub < testx5.run

 

[yaozhixiong@foxmail_com]

Thanks for your reply.I modified 2 parts in file testx5.run.part 1:#!/bin/csh -f#===============================================================================# USERDEFINED# This is where the batch submission is set.  The above code computes# the total number of tasks, nodes, and other things that can be useful# here.  Use PBS, BSUB, or whatever the local environment supports.#=============================================================================== ##PBS -N testx5##PBS -q batch##PBS -l nodes=2:ppn=8##PBS -l walltime=00:59:00##PBS -r n##PBS -j oe##PBS -S /bin/csh -V ##BSUB -l nodes=2:ppn=8:walltime=00:59:00##BSUB -q batch###BSUB -k eo###BSUB -J testx5###BSUB -W 00:59:00 #BSUB -J testx5#BSUB -q intelg_small#BSUB -W 1:00limit coredumpsize 1000000limit stacksize unlimited part 2:    mpirun.lsf $EXEROOT/cesm.exe >&! cesm.log.$LID
#mpiexec -n 16 $EXEROOT/cesm.exe >&! cesm.log.$LID  #mpirun -np 16 $EXEROOT/cesm.exe >&! cesm.log.$LID  Then I submitted the jobbsub < testx5.runBut it cannot be submitted. It says,Error: #!/bin/csh is not a regular file or doesn't exist or is an empty file.Request aborted by esub. Job not submitted.
Thank youzhixiong 

 

[jedwards]

This appears to be a more basic problem with submitting jobs on your system - have you successfully run a simple test such as the mpi hello program in the cesm users guide?