Nagios Support Forum

It was natural to me that the ability to monitor NAT devices (Linux) from many locations on one central server (good old RMM), would be a given on Nagios, but that's seemingly not quite the case. The NREP agent needs a direct connection from the server to the device being monitored. Is the NCPA agent able to establish an active connection to a central server (as opposed to the other way around)? I found some basic and vague install instructions using this documentation https://assets.nagios.com/downloads/ncp ... g_NCPA.pdf, but I for the life of me cant any decent info/documentation describing how one would start using it. Everything i found refers to configuring the server to point to the agent (the server having direct reachability to the agent).

My objective is simple. I would like to monitor remote linux devices behind firewalls from one central server. I would like to be able to configure the monitoring from one central server. I would like to monitor all the basics (cpu/ram/disk) as well as application monitoring (Apache/Postgresql). Can i achieve this using Nagios (without needing to invest the rest of my life to get it to work)?

I apologize that you're running into issues setting this up. Does your client have the ability to reach the server? If that is the case, you could use passive checks. https://assets.nagios.com/downloads/ncp ... ssive.html. NRDP may also work for you: https://assets.nagios.com/downloads/nag ... erview.pdf.

Hi HSmith,

Thanks for your response. So say i go with the NRDR (NSCA?) agent. How do i now monitor the box, starting with the basics (cpu/ram/disk/nic). The documentation describes manually running send_nrdp.php to send info about someservice. Where do i take it from here, and how do i get it to automatically send me all info (hopefully not a cron job but rather a live running service)?

And how about application monitoring, and perhaps netflow info from the box?

Typically yes, we run this sort of thing on a cron, but if you have a lot of checks you can have a single cron that handles the logic of running those checks, so you don't have 100 entries in your cron config. We provide the script to get the data to Nagios, and many people have differing opinions on how best to run it, but our recommendation is to keep it simple with a cron.

Regarding netflow, we actually have a product called Nagios Network Analyzer that was built to handle just that sort of thing: https://www.nagios.com/products/nagios- ... -analyzer/

It has a fully-functional 60-day trial just like all of our products, so I would suggest taking it for a spin and seeing if it will work for your needs.

Hi,

Could you possibly point me to the script and/or documentation describing how, and more importantly, "what" I could monitor using NSCA. As I mentioned, I would like the everything basics (cpu/ram/disk/nic), and application monitoring. Is the data sent to Nagios encrypted? I found some info for WIndows, but not Linux.

Can I collect netflow from a Linux server, and can I send it to a central Nagios server? Can I configure netflow to also include URL info (for http traffic)?

Personally I would use NRDP over NSCA as it is web-based (so encrypted if SSL is enabled in the XI server) and easier to work with since it doesn't require a binary to be compiled:

https://assets.nagios.com/downloads/nag ... erview.pdf

As for *what* can be monitored, just about anything as long as there is a plugin for it. Basically you write a wrapper script that does two things: First it runs a plugin to check something (like CPU) and captures the output and exit code, then it sends this information to Nagios via the send_nrdp.sh bash script (example given in the doc). Once you have the wrapper set up the way you need it, you can just about swap in any plugin you need. Tons of them can be found on our Exchange:

https://exchange.nagios.org/

As for the netflow, Network Analyzer will be able to collect that data, however I don't think it includes specific URLs (limitation of the netflow protocol).