Basic principles of monitoring

Here are a few basic Centreon concepts:

• A host is any device that has an IP address and that one wishes to monitor. For example, a physical server, a virtual machine, a temperature probe, an IP camera, a printer or a storage space.
• A service is a check point, or indicator, to be monitored on a host. This can be the CPU usage rate, temperature, motion detection, bandwidth usage rate, disk I/O, and so on. A service can consist of one or several metrics.
• In order to collect each indicator value, monitoring plugins are used which are periodically executed by a collection engine called Centreon Engine.
• To be executed, a plugin needs a set of arguments that define, for example, which host to connect to or through which protocol. The plugin and its associated arguments form a command.

For example, to monitor a host with Centreon is to configure all the commands needed to measure the desired indicators, and then deploy that configuration to the collection engine so that these commands are run periodically.

Once hosts and services are monitored, they have a status in Centreon (e.g. OK, Warning, Critical...). You can keep track of any changes using the Resources Status page.

If an alert occurs (not-OK/not-UP status), contacts will be able to receive notifications, within set time periods.

In Centreon, monitoring is made easy by the following elements:

• Host templates and service templates, that allow you to define default values so as to speed up the creation of these objects.

• Monitoring Connectors, that provide ready-to-use host and service templates. These greatly simplify the configuration of hosts and services: for instance, all you have to do is to apply Monitoring Connector templates to a host for it to be monitored.

• The autodiscovery feature for hosts and services, that allows you to get a list of new hosts and services and to add them automatically to the list of monitored resources.