HP-UX Workload Manager User's Guide

Example configuration files
performance_goal.template
Chapter 9316
##################
# PRM Components #
##################
#
# prm structure
#
# First, we define the workload groups.
#
# For this example, we will assume two basic workload groups:
# finance and sales
# Each group only has one application that we will monitor.
#
prm {
groups = [finance:2], [sales:3];
apps = [finance]: [/opt/fin_app/count_money], # app for finance group
[sales]: [/opt/sales/do_ebiz]; # app for sales group
}
############################
# Service-Level Objectives #
############################
#
# This is an SLO for the finance group and is only “active” on weekdays.
# It has a response-time goal, desiring that finance queries should
# complete in less than 2 seconds. A data collector to provide these
# response-time metrics to HP-UX WLM is specified later in a tune
# structure (a tune structure is required for all metrics used in an
# slo structure).
#
# In the goal line of this SLO, even though “metric” is a keyword, it could
# also have been contained within the brackets (as a user-configurable item).
# This is because there are other types of goal statements (such as usage)
# besides a metric (performance-based) goal. However, as this template is
# focused on showing the use of performance goals, “metric” is used
# explicitly, leaving just the “what-am-I-measuring?” piece as configurable.
#
slo [finance_query] {
pri = [1]; # SLO of highest priority
mincpu = [20]; # minimum CPU allocation (%)
maxcpu = [50]; # maximum CPU allocation (%)
entity = PRM group [finance];