Installing Standard LSF on a Subset of HP XC Nodes

# source LSF

. /opt/hptc/lsf/top/conf/profile.lsf.notxc

# valid hosts for standard LSF on this cluster

hosts="xc1 xc2 xc3 xc4 xc5 xc6"

hostname=`hostname`

valid=0

for i in $hosts; do

if [ "$hostname" = "$i" ]; then

valid=1

done

if [ "$valid" = "0" ]; then

exit 0

lsf_daemons "$1"

2. Save and exit the file. Then set permissions, create the appropriate softlink, and enable it:

# chmod 555 /opt/hptc/lsf/etc/slsf

# ln -s /opt/hptc/lsf/etc/slsf /etc/init.d/slsf

# chkconfig --add slsf

# chkconfig --list slsf

slsf 0:off 1:off 2:off 3:on 4:on 5:on 6:off

3. Edit /opt/hptc/systemimager/etc/chkconfig.map and add the following line to

enable this new "service" on all nodes in the cluster:

slsf 0:off 1:off 2:off 3:on 4:on 5:on 6:off

Adjust JOB_STARTER script for LSF-HPC for

SLURM

3. If the XC cluster version is earlier than v2.1 and LSF-HPC is configured with the recommended

JOB_STARTER script, make the following small change to the JOB_STARTER script. At the top of the

file change:

which srun > /dev/null 2> /dev/null

if [ "$?" != "0" ]; then

4. to the following:

if [ -z "$SLURM_JOBID" ]; then

5. This prevents the JOB_STARTER script from trying to invoke srun on the fat nodes.