Advanced MPI scheduling: Difference between revisions
(Created page with "{{Draft}} See [https://slurm.schedmd.com/sbatch.html sbatch] documentation. Discuss some or all of: * -N, --nodes= * -n, --ntasks= * --ntasks-per-core= * --ntasks-per-node=...") |
No edit summary |
||
| Line 1: | Line 1: | ||
{{Draft}} | {{Draft}} | ||
Most users should submit MPI or distributed memory parallel jobs as illustrated | |||
at [[Running_jobs#MPI_job|Running jobs: MPI job]]. Simply request a number of | |||
processes with <code>--ntasks</code> or <code>-n</code> and trust the scheduler | |||
to allocate those processes in a way that balances the efficiency of your job | |||
with the overall efficiency of the cluster. | |||
If you need more detailed control over how your job is allocated, then read on | |||
to learn about SLURM's [https://slurm.schedmd.com/sbatch.html <code>sbatch</code>] | |||
command and how its numerous options constrain the placement of processes. | |||
<code>sbatch</code> options: | |||
* -N, --nodes= | * -N, --nodes= | ||
* -n, --ntasks= | * -n, --ntasks= | ||
| Line 18: | Line 27: | ||
* -m, --distribution=[arbitrary|<block|cyclic|plane] | * -m, --distribution=[arbitrary|<block|cyclic|plane] | ||
* --mem_bind= | * --mem_bind= | ||
=== Hybrid jobs: MPI and OpenMP, or MPI and threads === | |||
To come | |||
=== MPI and GPUs === | |||
To come | |||
=== Why srun instead of mpiexec or mpirun? === | |||
To come | |||
=== External links === | |||
* [https://slurm.schedmd.com/sbatch.html sbatch] documentation | |||
* [https://slurm.schedmd.com/srun.html srun] documentation | |||
* [https://www.open-mpi.org/faq/?category=slurm Open MPI] and SLURM | |||
Revision as of 12:36, 20 March 2017
This is not a complete article: This is a draft, a work in progress that is intended to be published into an article, which may or may not be ready for inclusion in the main wiki. It should not necessarily be considered factual or authoritative.
Most users should submit MPI or distributed memory parallel jobs as illustrated
at Running jobs: MPI job. Simply request a number of
processes with --ntasks or -n and trust the scheduler
to allocate those processes in a way that balances the efficiency of your job
with the overall efficiency of the cluster.
If you need more detailed control over how your job is allocated, then read on
to learn about SLURM's sbatch
command and how its numerous options constrain the placement of processes.
sbatch options:
- -N, --nodes=
- -n, --ntasks=
- --ntasks-per-core=
- --ntasks-per-node=
- --ntasks-per-socket=
- --tasks-per-node=
- --threads-per-core=
- -c, --cpus-per-task=
- --mincpus=
- --cores-per-socket=
- --mem= and --mem-per-cpu=
- --exclusive[=user|mcs]
- --hint=[compute_bound|memory_bound|multithread|nomultithread]
- -m, --distribution=[arbitrary|<block|cyclic|plane]
- --mem_bind=
Hybrid jobs: MPI and OpenMP, or MPI and threads[edit]
To come
MPI and GPUs[edit]
To come
Why srun instead of mpiexec or mpirun?[edit]
To come