What do Zabbix hosts utilize to collect metrics?

In Zabbix, hosts utilize items to collect metrics. Items are the fundamental building blocks in Zabbix that define how data is collected from a host. Each item corresponds to a specific metric or piece of data that needs to be monitored, such as CPU load, memory usage, disk space, network traffic, and many others. By configuring items, Zabbix knows what data to gather, how to gather it (like polling or using a trap), and the data type (numerical, string, etc.).

When you create an item for a host, you specify several parameters, including the key that indicates what metric you want to monitor, the type of data being collected, and the update interval. This setup allows Zabbix to continuously gather data from various monitored hosts, making it possible to analyze performance trends or detect anomalies.

In the context of the other options, data sources refer to the origin of data and are not a specific component within Zabbix itself. Triggers are conditions set to react when item values exceed predefined thresholds, while events are records of what occurs when triggers are activated. Therefore, these elements work together in Zabbix monitoring, but items are specifically responsible for the direct collection of metrics from hosts.