SLURM views the LSF-HPC system as one large computer with many resources available to run jobs.

SLURM does not provide the same amount of information that can be obtained via standard LSF. But on HP XC systems, where the compute nodes have the same architecture and are expected to be allocated solely through LSF on a per-processor or per-node basis, the information provided by SLURM is sufficient and allows the LSF-HPC design to be more scalable and generate less overhead on the compute nodes.

Integrating LSF-HPC with SLURM on HP XC systems also provides you with a parallel launch command to distribute and manage parallel tasks efficiently. The SLURM srun command offers much flexibility in requesting topological requirements across an HP XC system, such as requesting contiguous nodes or executing only one task per node, or requesting nodes with specific features. This flexibility is preserved in LSF-HPC through the external SLURM scheduler; this is discussed in more detail in Section 7.1.2.

In an HP XC system, only one node runs LSF-HPC, but all nodes in the are configured as LSF-HPC Client Hosts; this means that every node is able to access LSF-HPC. You can submit jobs from any node in the HP XC system.

See Section 7.1.5 and the lsf_diff(1) manpage for more information on the subtle differences between standard LSF and LSF-HPC. Differences described in HP XC System Software documentation take precedence over descriptions in the LSF documentation from Platform Computing Corporation.

7.1.2 Topology Support

LSF-HPC contains topology support when requesting resources for a job. This topology support is available through LSF’s standard external scheduler feature, which makes use of a SLURM external scheduler provided with LSF-HPC on HP XC System Software systems.

Section 1.4.3 describes the interaction of SLURM and LSF-HPC on HP XC System Software systems.

You can apply LSF-HPC’s external scheduler functionality with the bsub command and in LSF queue configurations. See the LSF bqueues(1) command for more information on determining how the available queues are configured on HP XC System Software systems. The format of the LSF bsub command with the external SLURM scheduler option is:

bsub -ext"SLURM[slurm-arguments]"[bsub-options]jobname [job-options]

The slurm-argumentsparameter can be one or more of the following srun options, separated by semicolons:

•nodes=min [-max]

•mincpus= ncpus

•mem=value in Megabytes

•tmp=value in Megabytes

•constraint=’feature’

•nodelist= list-of-nodes

•exclude= list-of-nodes

•contiguous=yes

mem=value tmp=value constraint=value nodelist=list-of-nodesexclude=list-of-nodescontiguous=yes

The srun(1) manpage provides details on these options and their arguments.

7-2Using LSF