Creating an SNMP OID to monitor
In this recipe, we'll learn how to configure a Net-SNMP snmpd server on a Linux server to return the output of a command in an SNMP OID. This can be useful as an alternative to NRPE monitoring for information that is not otherwise available in a checkable network service, so that Nagios Core can check it via its standard check_snmp method.
As an example, this can be a very good way of monitoring hardware devices, such as RAID arrays on remote servers, where command-line diagnostic tools are available to report a status as a number or string, but they only work locally and don't otherwise include any information in an SNMP MIB tree.
Getting ready
The host we intend to check should be running a Net-SNMP snmpd server that allows full read access to the MIB tree for a specified community string, such as public. This SNMP server should be capable of using the exec directive in its configuration to return the output of a command as the value of an SNMP OID when requested...
