Agentless monitoring of bare-metal servers

[hmsdefender]

I am responsible for setting up bare-metal monitoring for an IaaS datacenter (hundreds of servers). I will have a Linux based host on which Nagios Core, or other monitoring software can be installed, but I may only communicate with the client servers at the bare-metal level (ping, fan, temperature, and other hardware features). I can only rely on the functionality available on a target server in its bare-metal state because the software installed on those servers (OSs, hypervisors, etc.) will change as often as the customers using the hardware change. This means I am limited to the functionality available on a given server before software is installed (management protocols such as iLO, DRAC, etc - whatever the hardware supports for getting the status of its internal components, vendor info, model info, serial number, etc.). I have scripts whereby I can query the server hardware for information, but I can't figure out how to set that up under Nagios Core.

Can anyone point me to documentation or examples that explains how this might be accomplished using Nagios Core? The few articles I found regarding "agentless" monitoring using Nagios Core actually used agents, albeit non-Nagios agents like the check_MK_agent. Outside of firmware/BIOS updates, I cannot install any software on the servers.

[scottwilkerson]

What you are going to need to do, is turn your script into bona-fide nagios plugins where they will return a string of text and proper exit code.

Then you can use these scripts on your nagios server to reach out to and get the data you need.

Here is the Nagios Plugin Development Guidelines
https://nagios-plugins.org/doc/guidelines.html