Running CBP model with slurm

[rburghol]

• cd run_bhatt
• ./OneCommandWSM.csh [scenario]
  • Ex: cbp OneCommandWSM.csh CFBASE30Y20180615_vadeq
• monitor slurm job output (essentially what you would see in the old model if you ran from command prompt)
  • usage: tail -f tmp/[username]-slurm/[scenario]_[slurm run timestamp].out
  • ex: tail -f tmp/rob-slurm/CFBASE10Y20180615_2022-02-24-15-24-55.out
• sinfo
  • look at slurm node status
  • ex output:

    sinfo: error: NodeNames=deq2 Sockets=0 is invalid, reset to 1
    PARTITION AVAIL  TIMELIMIT  NODES  STATE NODELIST
    debug*       up   infinite      1  alloc deq2

• squeue
  • list the currently running slurm jobs
  • Ex:

    squeue: error: NodeNames=deq2 Sockets=0 is invalid, reset to 1
         JOBID PARTITION     NAME     USER ST       TIME  NODES NODELIST(REASON)
            25     debug CFBASE10      rob  R      14:26      1 deq2

• scancel [job id]
  • cancel a slurmjob (id obtained from squeue output)
  • Ex: scancel 25

[rburghol]

Need to generate order lists:

Set Permissions and Ownership for /scratch/gb604b on deq2 for rob recursively
../../config/seglists/JL6_6890_6990_order.riv: No such file or directory.