HP-UX Workload Manager User's Guide

Example configuration files
entitlement_per_process.wlm
Chapter 9 293
}
# Grant 5 shares to servers for every active httpd process.
# Never allow the allocation to fall below 10 shares, nor to
# rise above 90% of the CPU resources.
#
slo servers_proportional {
pri = 1;
mincpu = 10;
maxcpu = 90;
entity = PRM group servers;
cpushares = 5 more per metric apache.active_procs ;
}
# Any CPU resources that remain after satisfying the above SLO is given to
# the OTHERS group by default. You can change this default using the
# distribute_excess keyword. For more information on this keyword, see
# the wlmconf(4) manpage.
#
# Collect the glance metric APP_ACTIVE_PROC as a measure
# of how many httpd processes are actually busy.
#
# NOTE: If the number of active processes is too volatile, consider using the
# smooth(1M) tool to calculate a running average, smoothing out the extreme
# values. smooth is shown below, commented out.
#
tune apache.active_procs {
coll_argv =
wlmrcvdc
# smooth
glance_prm APP_ACTIVE_PROC servers ;
}
For information on the cpushares keyword, see “Specifying a
shares-per-metric allocation request (optional)” on page 205.