Reputation: 31
I want to create metric that show the current CPU Usage on a host. Metric that I want is like on the Ganglia (gweb) How I can build that?
Upvotes: 0
Views: 870
Reputation: 4027
If the host is Linux/UNIX you could install NRPE on the monitored host, then use check_nrpe
to remotely run the check_cpu
plugin.
If the host is windows, you could install NC_Net on the monitored host, then use check_nt
to query the CPU usage.
If the host is SNMP-capable, you could use check_snmp
to query the CPU OID , either 1.3.6.1.4.1.2021.11.11.0
(NET-SNMP) or 1.3.6.1.4.1.9.2.1.58.0
(HOST-MIB)
If the host is a VMware guest, then you need to query the VirtualCentre. Check on monitoringexchange.org for the check_vmware
plugin.
This will allow Nagios to alert based on CPU usage thresholds. To obtain graphs and so on (as in Ganglia) you will need to add something like pnp4nagios
to graph the perfstats.
Upvotes: 3