Select an HP ProCurve switch/router (CPU/memory utilization) as the target host to create a monitor — the SNMP service should be up and running on it. Click New Monitor, then check SNMP Custom on the Favorites tab, click Next, and confirm the host. On the next page, click Select... to open the built-in SNMP MIB Browser and type hpmpCPUPctBusy into the Find box to locate it in the OID tree, selecting the specific row/instance you want to monitor since this is a table column, then select it and click OK. This object reports the percentage time the CPU was actively engaged during the last interval. On the monitor's Main parameters page you can set the target's SNMP port (default 161), credentials, polling interval, and other settings — see the SNMP Monitor help for details. On the State conditions and Alerting tabs, configure when the monitor should change state and trigger an alert; since this is a Gauge32-type OID, Value bounds is the most useful condition here — trigger an alert if utilization exceeds 85%, a common warning threshold before sustained high load affects performance. Click Finish to create the monitor; you can adjust any parameter later.