Monitoring With Nagios

Monitoring is very important for the uptime of your servers and applications. Once a monitoring program finds a problem it sends alerts and optionally tries to fix the problem. While monitoring software varies from platform to platform and protocols used Nagios stands up among all rivals.

Nagios is Open Source, free and probably the most widely used software for IT monitoring. You can hardly imagine a scenario in which Nagios cannot be used - from  monitoring your Windows Active Directory servers to your network switches, Nagios will be always the perfect solution. 

One of the strongest Nagios features is its SNMP support for communication with the monitored nodes. SNMP is easy to understand and configure, widely supported and very lightweight and fast. SNMP daemons / agents are available for any platform and in every major Linux distribution repository. For example, in RedHat based operating systems you can find it in the main repository under the name:

net-snmp.i686 : A collection of SNMP protocol tools and libraries

Another great feature of Nagios is its large community. There you will not only get help with any kind of Nagios related problems but you will also find thousands of Nagios extensions - anything from Monitoring MySQL replication to Microsoft Exchange server queue.

Furthermore, Nagios is very flexible and its configuration covers various options such as:
  • Objects - objects are involved in every aspect of Nagios configuration. If are not acquainted with the objects idea such as in object oriented programming it may take a while before falling in love with the principles of re-usability and inheritance. While this may seem unnecessary for monitoring a few hosts this can save you a lot of time and confusion for larger environments.
  • Passive cheks - allow remote hosts to send notifications upon internal events;
  • Very extensive configuration options - anything you can think of and wish to configure, Nagios has it. From notification periods (you may not wish to be waken up in 3:00 AM that your printer is not responding) to dependencies (when a router fails it is obvious that all nodes behind it will seem down).
Last but not least Nagios may not be the easiest and most suitable solution for very specific needs. As in real life nothing is really absolutely universal and perfect. That's why other monitoring systems (ZenOS, Solarwinds) exist and you may wish to check them to make sure you have made the best choice for your specific needs and skills.

blog comments powered by Disqus