Managing Serviceguard Extension for SAP Version B.05.10, December 2012

The following example WLM configuration file illustrates how to favor dialog to batch processing.
prm {
groups = OTHERS : 1,
Batch : 2,
Dialog : 3,
SAP_other: 4;
users = <sid>adm: SAP_other, ora<sid>: SAP_other;
#
# utilize the data provided by sapdisp.mon to identify batch and dialog
#
procmap = Batch : /opt/wlm/toolkits/sap/bin/wlmsapmap -f
/etc/cmcluster/<SID>/wlmprocmap.<SID>_<INSTNAME><INR>_<HOSTNAME> -t BTC,
Dialog :
/opt/wlm/toolkits/sap/bin/wlmsapmap -f /etc/cmcluster/P03/wlmprocmap.P03_DVEBMGS00_cutst144 -t DIA;}
#
# definition of Service Level Objectives
#
slo s_Dialog {
pri = 1;
entity = PRM group Dialog;
mincpu = 1;
maxcpu = 600;
goal = usage _CPU;
condition = metric sap_db_active;
}
slo s_Batch {
pri = 2;
entity = PRM group Batch;
mincpu = 1;
maxcpu = 600;
goal = usage _CPU;
condition = metric sap_db_active;
}
The following example WLM configuration file ties the CPU core shares guaranteed for dialog
processing to the number of dialog processes running as of the current SAP operation mode. In
times of high load, the overall dialog processing power of the instance is guaranteed to be at least
25% of a core multiplied by the number of dialog work processes running.
#
# Uses absolute CPU units so that 100 shares == 1 CPU.
#
tune {
wlm_interval = 5;
absolute_cpu_units = 1;
}
#
# Create a workload group for the dialog processes and use the wlmsapmap
# utility to move all the dialog processes to this workload group.
#
prm {
groups = OTHERS : 1, dialog : 2;
procmap = dialog : /opt/wlm/toolkits/sap/bin/wlmsapmap
-f /etc/cmcluster/C11/wlmprocmap.C11_D01_node1 -t DIA; }
#
# Request 25 shares per dialog job.
#
#
slo s_dialog {
pri = 1;
entity = PRM group dialog
cpushares = 25 shares total per metric m_dialog; }
#
# Report the number of active processes in the dialog workload group.
#
tune m_dialog {
coll_argv = glance_prm APP_ACTIVE_PROC dialog; }
Optional Step: OS795
Legacy Package Configuration 103