Nagios

Marc O’Gorman

marc@og.co.nz

18 May 2017

1 Intro

With the current trend of moving hardware onto a remote cloud system, there is a need to allow remote monitoring of the VMs (Virtual Machines) running in the cloud. Nagios is such an application. It can be configured to alert a remote administrator about actual or impending failures on remote machines. We rely on Nagios to alert us at O’Gorman Computer Consultants on behalf of our clients whose business computer systems are in the cloud.

2 What is Nagios

Nagios is a system service checking system. The developers explain its name as a recursive acronym meaning Nagios Aint Gonna Insist On Sainthood. It is free and open source and runs on any Unix like system or even (God forbid) on MicroSoft systems.
It started out named SATAN (Security Administrator’s Tool for Analysing Networks) but this was objectionable to many potential users and the name was changed to NetSaint. But then owing to trademark infringement issues it was renamed to Nagios. The agios portion of the name is Greek for saint and should be pronounced ajios or ayos rather than with a hard g.
Nagios relies on a running network with computers running checks on themselves and reporting back to a Nagios server.
The server then displays these in graphical format normally with green depicting a running server and services. Yellow meaning the server or service needs attention and red meaning that the server or service is critical and is down.
The service checks can be either active or passive. The server probes these checks via nrpe and gets a response or the server does a check on this service by checking the server for this service. e.g. smtp check on a mail server.
Normally for its host alive check it pings the server. If the server responds to the ping then it is regarded as being up. This can be changed to an ssh check to see if the secure shell service is running when the server is behind a firewall that doesn’t respond to ping checks.

3 NRPE

Most standard checks are reliant on a Nagios Remote Plugin Executor or NRPE.
These run on a server and allow the Nagios Host to read the output provided it is one of the allowed hosts in the allowed_hosts parameter.
Standard checks are :
Other Nagios checks can be utilised for specific servers and enabled

4 Nagios Alerts

Nagios can be configured to alert contacts normally by email but it can also be configured to alert by SMS if you have a provider set up to accept emails and send to cell phones.
The Nagios alerts can also be configured to only alert once if a server is down and not alert for the child services that depend on the server.