I am doing this by pulling the stats you mentioned (IO, CPU, latency) via SNMP with Icinga (a Nagios fork) and then graphing it with pnp4nagios. Pnp4
↧