Ok, so here’s the deal: you want to monitor all your network infrastructure with Nagios (or something else, but these instructions are geared towards Nagios specifically), but you’re new to the program and so you’re starting with the little things first, like collecting system up-time on your switches. The problem is that you’re just seeing service timeouts where you would like to see that up-time. Fear not! I have done the leg-work (or click-work) and present you with the answer to your problem in two parts.
The first part is to get your switch set up to use SNMP version 3 (much more secure than 1 or 2c). Log in to your switch and run the following
At this point, you’ll have to set up an initial user. Don’t sweat it–you’ll be deleting this user in a minute, so you can just enter junk information here.
snmpv3 user "username" auth sha "password" priv des "password"
snmpv3 group operatorauth user "username" sec-model ver3
no snmpv3 user initial
Now SNMP is enabled on your switch, you can move over to the Nagios end of things.
Note: I’m using NConf to configure Nagios, and these instructions assume that you’re doing the same. Wherever your
snmp_check command is configured, set it up to execute with the following parameters:
snmp_check!-P 3 -a sha -U [username as set above] -A [password as set above] -L authNoPriv -o 126.96.36.199.188.8.131.52.0
Now just reload the Nagios service, and the next time a service check is run for up-time, you should get a nice number telling you how many days the switch in question has been up.
Giving credit where it’s due, the SNMP configuration information came from here (where you can also find information on how to configure Cisco switches).