1) Login as root on all your host machine (Note: if your environment has configured to use lsf account id then use that)
2) Ensure LSF state is clean.
# /etc/rc3.d/S99lsf stop
Stopping the LSF subsystem
# sleep 20
# /etc/rc3.d/S99lsf status
Show status of the LSF subsystem
lim is stopped...
res is stopped...
sbatchd is stopped...
3) Start LSF on each node . Check status after 70 seconds before moving to next node.
# /etc/rc3.d/S99lsf start
# sleep 70
# /etc/rc3.d/S99lsf status
4) Check if bhosts shows atleast 2 nodes 'ok' and ProcessManager in EGO using commands
Login using your id on the host that has process manager and run following commands to verify whether the lsf is up:
. /lsf/conf/profile.lsf
lsid
bhosts
egosh service list
2) Ensure LSF state is clean.
# /etc/rc3.d/S99lsf stop
Stopping the LSF subsystem
# sleep 20
# /etc/rc3.d/S99lsf status
Show status of the LSF subsystem
lim is stopped...
res is stopped...
sbatchd is stopped...
3) Start LSF on each node . Check status after 70 seconds before moving to next node.
# /etc/rc3.d/S99lsf start
# sleep 70
# /etc/rc3.d/S99lsf status
4) Check if bhosts shows atleast 2 nodes 'ok' and ProcessManager in EGO using commands
Login using your id on the host that has process manager and run following commands to verify whether the lsf is up:
. /lsf/conf/profile.lsf
lsid
bhosts
egosh service list
Comments
Post a Comment