4. LSF-HPC prepares the user environment for the job on the LSF-HPC execution host
node and dispatches the job with the job_starter.sh script. This user environment
includes standard LSF environment variables and two SLURM-specific environment
variables: SLURM_JOBID and SLURM_NPROCS.
SLURM_JOBID is the SLURM job ID of the job. Note that this is no t the same as the
LSF jobID.
SLURM_NPROCS is the number of processors allocated.
These environment variables are intended for use by the user’s job, whether it is exp licitly
(user scripts may use these variables as necessary) or implicitly (any srun commands in
the user’s job can use these variables to d eterm ine its allocation of resources).
The value for SLURM_NPROCS is 4 and the SLURM_JOBID is 53 in t his example.
5. The user job myscript begins execution on com pu te node n1.
Thefirstlineinmyscript is the hostname command. It executes l ocally and returns
the name of node, n1.
6. The second line in the myscript script is the srun hostname command. The
srun command in myscript inherits SLURM_JOBID and SLURM_NPROCS from
the environment and executes the hostname command on each compute node in the
allocation.
7. The output of the hostname tasks (n1, n2, n3,andn4). is aggregated back to the srun
launch command (shown as dashed lines in Figure 7-1), and is ultimately returned to the
srun commandinthejobstarterscript, where it is collected by LSF-HPC.
Thelastlineinmyscript is the mpirun -srun ./hellompi command. The srun
command inside the mpirun command in myscript inherits the SLURM_JOBID and
SLURM_NPROCS environment variables from the environment and executes hellompi on
each compute node in the allocation.
The output of the hellompi tasks is aggregated back to the srun launch command where it
is collected by LSF-HPC.
The command executes on the allocated compute nodes n1, n2, n3,andn4.
When the job finishes, LSF-HPC cancels the SLURM allocation, which frees the compute
nodes for use by another job.
7.1.5 Differences Between LSF on HP XC and Stan dard LSF
LSF for the H P XC environment supports all the standard features and functions th at standard
LSF supports, except for those items described in this s ection, in S ection 7.1.6, and in the HP
XC release notes for LSF.
• The external scheduler option fo r HP XC provides additional capabilities at the job level and
queue level by allowing the inclusion of several S LURM opt ions in the LSF command lin e.
• LSF does not collect maxswap, ndisks, r15s, r1m, r15m, ut, pg, io, tmp, swp and
mem load indices from each application node. lshosts and lsload commands will
display "-" for all of these items.
• LSF-enforced job-level r un-time limits are not supported.
• Except run-time (wall clock ) and total number of CPUs, LSF cannot report any other job
accounting information.
• LSF does not support parallel or SLURM-based interactive j obs in PT Y mode (bsub -Is
and bsub -Ip).
• LSF does not support user-account mapping and system-account mapping.
7-6 Using LSF
Komentarze do niniejszej Instrukcji