HP XC System Software Administration Guide Version 3.2
B.2 Assumptions
• LSF-HPC with SLURM for SLURM was installed by the cluster_config process using
default values
• You have a proper Platform LSF license
• There is no need to communicate with an external LSF cluster (this can be done, but involves
additional procedures to prepare the external network connections).
B.3 Sample Case
Consider an HP XC system of 128 nodes consisting of:
• A head node with a host name of xc128
• 6 large SMP nodes (or fat nodes) with the host names xc[1-6]
• 122 thin nodes.
114 of the thin nodes are compute nodes and have host names of xc7-120
B.4 Instructions
1. Log into the head node of the HP XC system as superuser (root). Do not log in though the
cluster alias.
2. Change directory to /opt/hptc/lsf/top/conf.
3. Rename the existing setup files:
# mv profile.lsf profile.lsf.xc
# mv cshrc.lsf cshrc.lsf.xc
4. Obtain standard LSF and install it into the existing LSF tree.
These instructions assume that you are familiar with the procedures to install Standard LSF,
which consist of configuring the install.config file and running .
# ./lsfinstall -f install.config
Here are the required minimum settings for the installation configuration file (adjust as
necessary for your system):
LSF_TOP="/opt/hptc/lsf/top"
LSF_ADMINS="lsfadmin"
LSF_CLUSTER_NAME="hptclsf"
LSF_ADD_SERVERS="xc1 xc2 xc3 xc4 xc5 xc6"
Be sure to set the host names of your fat nodes in the LSF_ADD_SERVERS entry.
5. Correct the setup files
• The new installation of standard LSF will create new setup files. Rename these files:
# mv profile.lsf profile.lsf.notxc
# mv cshrc.lsf cshrc.lsf.notxc
• Now create soft links of the original files to their HP XC equivalents:
# ln -s profile.lsf.xc profile.lsf
# ln -s cshrc.lsf.xc cshrc.lsf
• The newly installed setup files for Standard LSF need to be edited so that the LSF
daemons can determine whether or not LSF is running on a SLURM based system.
If the HP XC file, /etc/hptc-release, is found, the setup files assume that the LSF
daemons are interfacing with SLURM.
However, this is not true for the fat nodes running Standard LSF. To bypass this, create
an identity file that exists only on the thin nodes, and have the Standard LSF setup files
look for this identity file.
The identity file will be /var/lsf/lsfslurm. Use the pdsh command to create this
file on all nodes except the fat nodes. For the sample case, the pdsh command resembles
this:
304 Installing Standard LSF on a Subset of Nodes