<> = Submitting jobs = <> Whatever you read here may need to be adjusted to fit your specific case. Do not hesitate to ask for some help when needed. == Filesystems == Some but not all partitions are available to the compute nodes. Compute nodes will not be able to access any data from filesystems that are not listed here. * /work * /scratch * /home == Slurm partitions == There are currently 2 partitions, '''normal''' and '''bigmem'''. The '''normal''' partition is the default partition if you submit a job without precising which partition should be used. Your job will be placed in the normal partition (250GB). The normal partition has limited RAM of 250GB, in case you need more than that please use the '''bigmem''' partition. Use ''-p'' to specify needed partition. == Load necessary software == By default only some software will be available at login. To be able to use other software scripts you should first load them. The command '''module''' will help you to manage modules and their dependencies. To check which programs are loaded (ready to be used), use the command below. {{{ module list }}} Expected output is, {{{ Currently Loaded Modules: 1) autotools 2) prun/1.3 3) gnu8/8.3.0 4) openmpi3/3.1.4 5) ohpc }}} To check which programs are available (can be loaded). The same command can be used to search a specific package. {{{ module avail }}} Expected output is, {{{ --------------------------------------------------- /opt/ohpc/pub/moduledeps/gnu8-openmpi3 ---------------------------------------------------- adios/1.13.1 hypre/2.18.1 netcdf-cxx/4.3.1 petsc/3.12.0 py2-scipy/1.2.1 scorep/6.0 trilinos/12.14.1 boost/1.71.0 imb/2018.1 netcdf-fortran/4.5.2 phdf5/1.10.5 py3-mpi4py/3.0.1 sionlib/1.7.4 dimemas/5.4.1 mfem/4.0 netcdf/4.7.1 pnetcdf/1.12.0 py3-scipy/1.2.1 slepc/3.12.0 extrae/3.7.0 mpiP/3.4.1 omb/5.6.2 ptscotch/6.0.6 scalapack/2.0.2 superlu_dist/6.1.1 fftw/3.3.8 mumps/5.2.1 opencoarrays/2.8.0 py2-mpi4py/3.0.2 scalasca/2.5 tau/2.28 -------------------------------------------------------- /opt/ohpc/pub/moduledeps/gnu8 -------------------------------------------------------- hdf5/1.10.5 metis/5.1.0 mvapich2/2.3.2 openblas/0.3.7 pdtoolkit/3.25 py3-numpy/1.15.3 likwid/4.3.4 mpich/3.3.1 ocr/1.0.1 openmpi3/3.1.4 (L) py2-numpy/1.15.3 superlu/5.2.1 ------------------------------------------------------------- /tools/modulefiles -------------------------------------------------------------- MEGAHIT/1.2.9 ---------------------------------------------------------- /opt/ohpc/pub/modulefiles ---------------------------------------------------------- EasyBuild/3.9.4 clustershell/1.8.2 gnu7/7.3.0 llvm5/5.0.1 pmix/2.2.2 valgrind/3.15.0 autotools (L) cmake/3.15.4 gnu8/8.3.0 (L) ohpc (L) prun/1.3 (L) charliecloud/0.11 gnu/5.4.0 hwloc/2.1.0 papi/5.7.0 singularity/3.4.1 Where: L: Module is loaded Use "module spider" to find all possible modules. Use "module keyword key1 key2 ..." to search for all possible modules matching any of the "keys". }}} To search for a module, {{{ module avail <> #OR module spider <> }}} To load a module do, {{{ module load <> }}} Loading a module can be done following these 3 steps, 1. Locate the module, '''module avail''' 2. Check how to load it, '''module spider <>''' 3. Load your module using the instructions '''module load <>''' 4. Check that the module is loaded, '''module list''' 5. Enjoy!!! ;-) Read more about '''module''' usage [[https://lmod.readthedocs.io/en/latest/010_user.html|https://lmod.readthedocs.io/en/latest/010_user.html|target="_blank"]] == Prototype of batch script == This prototype should be in a script file, for example, my_first_script.sbatch {{{#!highlight bash #!/bin/bash #SBATCH -J test # Job name #SBATCH -o /work/<>/job.%j.out # Name of stdout output file (%j expands to jobId) #SBATCH -e /work/<>/job.%j.err # Name of stderr output file (%j expands to jobId) #SBATCH -n 16 # Total number of threads requested or total number of mpi tasks #SBATCH --mem=2000 # Maximum amount of memory in Mb the job can use #SBATCH -t 00:30:00 # Run time ([d-]hh:mm:ss) - 1.5 hours #SBATCH --mail-type=ALL #SBATCH --mail-user=your.email@wur.nl # Load your software/command module load CMD/version # Run your command CMD [OPTIONS] ARGUMENTS }}} To run a sbatch script use {{{ sbatch <