HP XC System 3.x Software - Manual (Page 99)

• LSF-HPC integrated with SLURM only runs daemons on one node within the HP XC system.
This node hosts an HP XC LSF Alias, which is an IP address and corresponding
host name
specifically established for LSF-HPC integrated with SLURM on HP XC to use. The HP XC
system is known by this HP XC LSF Alias within LSF.
Various LSF-HPC commands, such as lsid , lshosts, and bhosts, display HP XC LSF
Alias in their output. The default value of the HP XC LSF Alias, lsfhost.localdomain
is shown in the following examples:
$ lsid
Platform LSF HPC version number for SLURM, date stamp
Copyright 1992-2005 Platform Computing Corporation
My cluster name is hptclsf
My master name is lsfhost.localdomain
$ lshosts
HOST_NAME type model cpuf ncpus maxmem maxswp server RESOURCES
lsfhost.loc SLINUX6 Opteron8 60.0 8 2007M - Yes (slurm)
$ bhosts
HOST_NAME STATUS JL/U MAX NJOBS RUN SSUSP USUSP RSV
lsfhost.localdomai ok - 8 0 0 0 0 0
• All HP XC nodes are dynamically configured as “LSF Floating Client Hosts” so that you
can execute LSF-HPC commands from any HP XC node. When you do execute an LSF-HPC
command from an HP XC node, an entry in the output of the lshosts acknowledges the
node is licensed to run LSF-HPC commands.
In the following example, node n15 is configured as an LSF Client Host, not the LSF execution
host. This is shown in the output when you run lshosts command is run on that node:
The values for the type and model are UNKNOWN and the value for server is No.
$ lshosts
HOST_NAME type model cpuf ncpus maxmem maxswp server RESOURCES
lsfhost.loc SLINUX6 Opteron8 60.0 8 2007M - Yes (slurm)
$ ssh n15 lshosts
HOST_NAME type model cpuf ncpus maxmem maxswp server RESOURCES
lsfhost.loc SLINUX6 Opteron8 60.0 8 2007M - Yes (slurm)
n15 UNKNOWN UNKNOWN_ 1.0 - - - No ()
• The job-level run-time limits enforced by LSF-HPC integrated with SLURM are not supported.
• LSF-HPC integrated with SLURM does not support parallel or SLURM-based interactive
jobs in PTY mode (bsub -Is and bsub -Ip). However, after LSF-HPC dispatches a user
job on the HP XC system, you can use the srun or ssh command to access the job resources
directly accessible. For more information, see
“Working Interactively Within an Allocation” .
• LSF-HPC integrated with SLURM does not support user-account mapping and
system-account mapping.
• LSF-HPC integrated with SLURM does not support chunk jobs. If a job is submitted to chunk
queue, SLURM lets the job pend.
• LSF-HPC integrated with SLURM does not support topology-aware advanced reservation
scheduling.
10.4 Job Terminology
The following terms are used to describe jobs submitted to LSF-HPC integrated with SLURM:
Batch job A job submitted to LSF-HPC or SLURM that runs without any
I/O connection back to the terminal from which the job was
submitted. This job may run immediately, or it may run
10.4 Job Terminology 99