How do the data collectors in the Prometheus system function?

The Prometheus system’s data collector helps users understand the performance and health of the system by displaying and storing monitoring data. It operates on the following basic principle:

1. Source of data: The Prometheus data collector regularly fetches metric data from monitored target systems, which can be servers, applications, databases, and so on.
2. Target discovery: Prometheus automatically identifies and recognizes the targets that need to be monitored through service discovery or static configuration, and begins collecting data regularly.
3. Data fetching: The Prometheus collector will retrieve metric data from the target system at predefined intervals, typically in the form of text-based metric data exposed at specific endpoints on the target system.
4. Storage: The collector will store the collected metric data in a local time series database for later querying and analysis.
5. The query language in Prometheus, called PromQL, is similar to SQL and allows users to query and analyze monitoring data stored in the database.
6. Visualization: Users can utilize graphical interfaces provided by Prometheus or tools like Grafana to display and visualize monitoring data, allowing for a more intuitive understanding of the system’s performance status and trends.

In general, the Prometheus data collector regularly fetches metric data from target systems and stores it in a local database to help users monitor and analyze system performance.