Skip to main content

Exporters

Updated Nov 20, 2022 ·

Overview​

Prometheus collects metrics by sending HTTP requests to the /metrics endpoint of each target. The endpoint can also be changed and Prometheus can be configured to use a different path other than /metrics Note that most systems don't expose metrics on an HTTP endpoint. For these instances, we can install exporters on the targets which:

  • Collects metrics from the service
  • Converts metrics to a format expected by Prometheus
  • Exposes /metrics endpoint so Prometheus can scrape the data.

It can monitor various systems, including:

  • Websites running on Windows
  • Batch processes on Linux
  • Servers on both Windows and Linux

Many exporters are developed by the community or the Prometheus project. They are available for various use cases, including:

  • Databases
  • Message queues
  • Cloud services
  • Hardware, etc.

Setting up a Node Exporter​

Exporters act as intermediaries between the target nodes and the Prometheus server. It is installed on the target nodoe and it translates system or application data into Prometheus-compatible formats. To setup a node exporter on a Linux node, please see Setting up Exporters.

Types of Exporters​

Some common exporters include:

  • Node Exporter

    • Monitors system-level metrics like CPU, memory, and disk usage.
    • Collects data from Linux-based systems.
    • Provides essential server health metrics.

  • Blackbox Exporter

    • Probes services over HTTP, HTTPS, DNS, TCP, and ICMP.
    • Monitors the availability of external services.
    • Customizable for specific endpoints.

  • MySQL Exporter

    • Collects MySQL database metrics.
    • Monitors query times, replication, and connections.
    • Helps track MySQL performance.

  • JMX Exporter

    • Exposes metrics from Java applications via JMX.
    • Monitors JVM performance and garbage collection.
    • Used for Java-based applications.

  • Exporter for Cloud Services

    • Collects metrics from cloud services like AWS, Azure, and GCP.
    • Tracks cloud resource usage and performance.
    • Monitors both cloud and on-premise resources.