Nagios Support Forum

Hi,

I am using Nagios to monitor my servers remotely. I can get hardware details like hard disk usage, RAM details etc. using nrpe on the monitoring server.
These are my questions:

1. How does nrpe get these values?
2. Which hardware device does nrpe use to get these values on the monitored server?
3. Can we get hard disk and RAM details without using nrpe on the monitored host?
4. Can we get hard disk and RAM details without using nrpe on the monitored host, without any utility installed on the monitored server?

It would be great if I can get help on this.

Thank you.

Hi,

I am using Nagios to monitor my servers remotely. I can get hardware details like hard disk usage, RAM details etc. using nrpe on the monitoring server.
These are my questions:

harshkhatore wrote:1. How does nrpe get these values? (I can get hardware details like hard disk usage, RAM details etc)

NRPE is responsible for Executing Nagios Plugins Remotely. In a nutshell, this means NRPE itself isn't doing anything more sophisticated than executing a Nagios plugin at the request of some remote machine (like a Nagios Core machine via check_nrpe). The components responsible for getting CPU usage, disk usage, all that stuff is the actual nagios plugins.

The main suite of nagios plugins is available here:
https://github.com/nagios-plugins/nagios-plugins/

Though there are also thousands upon thousands of community created plugins. Since these generally follow the Nagios plugin development guidelines, they can also be executed by NRPE (environmental restrictions notwithstanding).

harshkhatore wrote:2. Which hardware device does nrpe use to get these values on the monitored server?

I'm not sure what you mean by "hardware". NRPE uses some sort of networked connection and the general process pipeline in Linux.

harshkhatore wrote:3. Can we get hard disk and RAM details without using nrpe on the monitored host?

Sure, but you'll probably need an "agent" one way or another. If the "agent" isn't NRPE, it could be NCPA, Net-SNMP, NSClient++, just to name a few.

You could also use check_by_ssh to execute plugins via SSH. Your SSH server would act as an "agent" in this case.

harshkhatore wrote:4. Can we get hard disk and RAM details without using nrpe on the monitored host, without any utility installed on the monitored server?

It depends on what you mean by "utility". SSH is a utility. SNMP is a utility. TCP connections are brokered by a utility.

Most Linux systems don't offer, out of the box, a front-facing interface that tells "whoever wants to know" about the specific performance metrics of the machine. Which part of why agents like NRPE exist.

Thank you for the reply.
By utility I meant that can we get hardware details without installing Nagios nrpe on the server?
Also, which protocols can be used to do it, IPMI or SMTP to name a few.

Thank you.

First off, I'm moving this thread out of Nagios Core Development because it is not related to the ongoing development of Nagios Core.

In regards to your questions, without NRPE you would need something like SNMP or SSH (or even telnet) to monitor the internals of the machine. SMTP is an email-related protocol and is not really useful as a monitoring agent. IPMI could be used if the machine supports it.