Remote Monitoring

[matson-itops]

Need some community advice.

we need
a) have nagios do check from a public network (eg interweb at large. Confirm job public can reach site)
b) have nagios do check from private networks (eg remote offices. Confirm our internal networks can reach each other and internal resources)

do we need pollers, smaller slave nagios servers, smaller stand alone nagios server on each network?
how does network topology and routing play into this? (theoretically speaking)

[gregbeyer]

To check your (web) site is accessible to the public, you do not need Nagios outside your DMZ. I think your simplest solution is something like uptimerobot.com. You can do checks as simple as the site is up and responding. Or do more advanced checks like login testing JSON upload, and more. Some features are paid, so check what they offer both free and paid. Much easier and safer than running Nagios outside your DMZ.

For internal network to other internal network, use NRPE -- I think you want to do this:

https://serverfault.com/questions/41965 ... le-servers

There are many NRPE plugins besides check_http that you can use depending on what you want to test network-to-network. ssh, ftp, many more.

[swolf]

Hi @matson-itops,

Greg's perspective is interesting and I think his solution could work. Officially, our stance is that you would use the Nagios Cross-Platform Agent to perform checks from a remote server. NCPA can function either in a agent/"polled" capacity (taking commands directly from Nagios XI) or as an exporter (running in "passive" mode, and sending information back to NRDP). I wouldn't go for a subordinate nagios server until you had dozens-to-hundreds of checks that need to be run remotely.

Hopefully that helps, let me know if you need anything clarified.
-Sebastian

[snapier3]

Hi @matson-itops,

Greg's perspective is interesting and I think his solution could work. Officially, our stance is that you would use the Nagios Cross-Platform Agent to perform checks from a remote server. NCPA can function either in a agent/"polled" capacity (taking commands directly from Nagios XI) or as an exporter (running in "passive" mode, and sending information back to NRDP). I wouldn't go for a subordinate nagios server until you had dozens-to-hundreds of checks that need to be run remotely.

Hopefully that helps, let me know if you need anything clarified.
-Sebastian

I accomplished this (Windows Only) by using a modified version of the python wrapped Windows_SSH monitoring plugins that became available in the release of the Windows_SSH Wizard.

Originally I had one service that executed the ssh action via nagios against a target, this scripts collects and evaluates multiple metrics sending the results back idempotent services to NRDP.

check_sshr_nrdpr-flow.PNG

More recently I put together something like this using the core PowerShell in the og plugin.

NagiosXI --> Proxy NCPA | Executes Request via check_ncpa on list of responsible NCPA Targets --> Targets Responds with Passive Check Data to NRDP Server.

To do this I extracted the PowerShell from the Python wrapper in the plugin, deployed the collector plugin PS1 to all targets and I have an orchestration plugin that receives a list of target hosts where the collector plugin is executed using check_ncpa.

The plugin is available on GitHub and walks through the method and config if you want to play around with proxy checks.
https://github.com/SNapier/check_sshr_nrdpr

Happy Monitoring!

[matson-itops]

thank you kindly community for all the feedback

Appreciate the nudge in the right direction