title: Check CPU load average on systems using UCD SNMP Daemon
agents: snmp
author: Mathias Kettner <mk@mathias-kettner.de>
license: GPL
distribution: check_mk
description:
 This check measures the CPU load average via SNMP on systems running
 the UCD SNMP daemon. This daemon is mostly found on Linux systems
 and used by some hardware appliances (such as Riverbed Steelhead).

 The daemon sends a 1, 5 and 15 minute average of the CPU load.
 The 15 minute value is checked against a warning and a critical level.
 Since the SNMP MIB does not contain information about the number
 of CPUs this check assumes only one CPU to be present. Thatfore
 the warn/crit levels are interpreted as {absolute} values - not
 as values {per cpu} (like the in agent based load check).

examples:
 # Set CPU load levels for all SNMP devices to 10.0 and 20.0
 check_parameters += [
    ( (10.0, 20.0), [ "snmp" ], ALL_HOSTS, [ "CPU load" ] ),
 ]

perfdata:
 Three values: the load average of the last 1, 5 and 15 minutes.

inventory:
 One service will be created for each check supporting the UCD
 MIB. Currently the SNMP scan function will only consider hosts
 with the text {linux} in their system description. You can
 try an explicit inventory with {cmk --checks ucd_cpu_load -I HOSTNAME}
 if that is not the case.

[parameters]
warning (float): The 15 minute average load that triggers a warning
critical (float): The level for the critical state

[configuration]
cpuload_default_levels(float, float): The default levels for all CPU load checks. This is
  preset to {(5, 10)}.