Loading

Enable stack monitoring for Elasticsearch on a self-managed cluster

The Elastic Stack monitoring features consist of two components: an agent that you install on each Elasticsearch and Logstash node, and a Monitoring UI in Kibana. The monitoring agent collects and indexes metrics from the nodes and you visualize the data through the Monitoring dashboards in Kibana. The agent can index data on the same Elasticsearch cluster, or send it to an external monitoring cluster.

To use the monitoring features with the security features enabled, you need to set up Kibana to work with the security features and create at least one user for the Monitoring UI. If you are using an external monitoring cluster, you also need to configure a user for the monitoring agent and configure the agent to use the appropriate credentials when communicating with the monitoring cluster.

You can collect monitoring and logging data in the following ways:

If you're building a production environment, then you should send monitoring data to a separate monitoring cluster so that historical data is available even when the nodes you are monitoring are not. To learn how to store monitoring data in a separate cluster, refer to Monitoring in a production environment.

You can also monitor your stack using legacy Elasticsearch monitoring features. This method is deprecated and should not be used. To learn more, refer to Legacy collection methods for self-managed Elasticsearch.

Warning

The Elasticsearch Monitoring plugin is deprecated and will be removed in 10.0. Choose one of the following replacements to monitor your Elasticsearch cluster:

  • AutoOps (recommended): Simplifies cluster management through performance recommendations, resource utilization visibility, and real-time issue detection with resolution paths.
  • Stack monitoring with Elastic Agent or Metricbeat: Collect logs and metrics from Elasticsearch and ship them to a monitoring cluster for visualization in Kibana.

To help you decide, refer to the AutoOps and Stack Monitoring comparison.