nagios - monitoring network interface (linux)

what do you use for system interfaces though? i.e. linux hosts?

I'm currently running NGRE and I find any network stats.

for linux hosts themselves, like my email servers, I am still using SNMP. I have nagios run snmp query for interface eth0 on all my email servers so I can get bandwidth stats. Then I use pnp4nagios to make that data into graphs.

SNMP in my opinion is the best way to get interface statistics.

Trying to understand what nagios is suppose to use to monitor systems for cpu, memory? and why would nagios prefer to use NGRE for linux hosts versus just snmp. I hope you could provide some education.

sorry thats NRPE

nrpe would be a poor choice - monitoring performance stats is something snmp is good at (via the iftable) and that is available cross-platform - I am monitoring a bunch of linux, esx, solaris, cisco (switches/firewalls/routers) and storage (netapp) devices, all using iftable.

if you go with things like mrtg too, that's snmp.

In some ways SNMP can be complex to get rolling properly. NRPE for me provides an easy way to get information from a remote server when SNMP is too cumbersome or simply is not permitted due to network contstraints or other factors.

DaveHowe,

I don't know what this Iftable is - is that a plugin for Nagios? google is not showing anything solid.

Iftable I think simply means interface table. When running an SNMP query against a linux server, The bandwidth statistics will be for interfaces listed in the interface table; eth0, lo, etc, etc, etc:)