HP XC System Software Administration Guide Version 3.0

Thresholds in LSF-HPC-SLURM Interplay
When the HP XC system starts, some computer nodes may take a while to boot. If LSF-HPC starts to report
the current number of processors before the system stabilizes, the smaller jobs that are already queued are
scheduled. It may be better to run a larger job requesting more processors.
Use the following parameters to control when LSF-HPC starts to report total usable processors on HP XC
system:
DescriptionEnvironment Variable
The threshold parameter is the percentage of total usable
processors in LSF partition.
The default value is 80.
LSF_HPC_NCPU_THRESHOLD=threshold
This is the upper limit for the number of processors that changed
since the last checking cycle.
The default value is zero (0).
LSF_HPC_NCPU_INCREMENT=increment
The incr_cycles parameter is the minimum number of
consecutive cycles in which the number of processors changed
does not exceed LSF_HPC_NCPU_INCREMENT.
The default value is 1. LSF-HPC examines total usable processors
every 2 minutes.
LSF_HPC_NCPU_INCR_CYCLES=incr_cycles
This setting defines how two thresholds are combined. The only
valid values are and and or.
The default is or.
LSF_HPC_NCPU_COND= or | and
Configuring an External Virtual Host Name for LSF-HPC on HP XC Systems
An external virtual host name for LSF-HPC on an HP XC system needs to be accessed from the external
network. This access could be required if the HP XC system is added to an existing LSF cluster, or if the HP
XC system is 'Multi-Clustered' with another LSF cluster. See the LSF documentation for more details on LSF
Multi-Clusters.
Perform the following steps to configure an external virtual host name:
1. If LSF-HPC failover is enabled, ensure that each node with the resource management role has an external
network connection. Run the following command to confirm this:
# shownode roles --role resource_management external
resource_management: n[127-128]
external: n[125-128]
resource_management: n[127-128]
external: n[125-128]
2. Obtain a virtual IP and corresponding host name, and ensure that they are not already in use:
# ping -i 2 -c 2 xclsf
PING xclsf (10.10.123.1): 56 data bytes
--- xclsf ping statistics ---
2 packets transmitted, 0 packets received, 100% packet loss
# ping -i 2 -c 2 10.10.123.1
PING 10.10.123.1 (10.10.123.1): 56 data bytes
--- 10.10.123.1 ping statistics ---
2 packets transmitted, 0 packets received, 100% packet loss
3. Ensure that no jobs are running. For more information, see “Draining Nodes” (page 113)
4. Shut down LSF:
# controllsf stop
5. Change the virtual host name to the new one and confirm:
Example 135