Nagios, monitoring your life #1
Nagios is an open source host, service and network monitoring program. This is just a little definition which can be found at Nagios’ home page but this program can help a lot to any sysadmin, actually could do both help and bother if the sysadmin is who has to answer to problem reports. But let’s just be positive and see what it has to offer, I hope to make a good introduction with a little series of posts.
Introduction
Nagios works as a standalone process which reads some configuration files, actually it could be just a large file holding all the parameters/definitions/etc needed to monitor. The main file is known as nagios.cfg here can be defined the general behavior and also the files in which the configuration is split, the hierarchy could be as follows:
- nagios.cfg
- hosts.cfg
- hostgroups.cfg
- services.cfg
- servicegroups.cfg
- resources.cfg
Each file has its own well defined function, nagios.cfg is the main configuration file which includes the behavior of the process itself and points to the files in which you decide to split the settings.
hosts.cfg defines the hosts to be monitored, some settings including name, ip address, check intervals, notifications, etc.
hostgroups.cfg here we could define groups for hosts, this file is useful to group hosts by its locations, by clients, etc. it will one of the available views in the web environment.
services.cfg as the name would suggest this file will hold the service definitions.
servicegroup.cfg here we could group services, it is very useful on business depending on different kind of services. It is also another view in the nagios’ web environment.
resources.cfg this file will have definitions for resources to be used on the previous files, resources like plugins and event handler folder, database connection details (if we want to store the monitoring data on a database), etc.
Environment
This is the nagios setup I will explain on further posts, the nagios host is my NSLU2 from Linksys running an unofficial port of Debian for ARM big endian. This little baby will check for main services on thor and loki like HTTP, FTP, MySQL, SSH, etc. it will use the standard plugins, some of them will check remotely and some through an SSH connection using public key authentication to avoid the need of a password. The notifications will be mailed to the root at mydomain dot org which is an alias for my own user on thor (mail server), the most important services will also be notified via SMS through a useless Motorola V3 which has a broken screen but main functions remain working, the setup will prevent spending too much money with those SMSs.
Well I think this is it for the first post about Nagios, next one will come how to setup the whole environment, packages we will need, how to generate public keys with OpenSSH and how to use them with nagios plugins, etc.
