原英文版地址: https://www.elastic.co/guide/en/elasticsearch/reference/7.7/monitoring-production.html, 原文档版权归 www.elastic.co 所有
本地英文版地址: ../en/monitoring-production.html
重要: 此版本不会发布额外的bug修复或文档更新。最新信息请参考 当前版本文档。
Elasticsearch Guide [7.7] » Monitor a cluster » Monitoring in a production environment
« How monitoring works Collecting Elasticsearch monitoring data with Metricbeat »

Monitoring in a production environmentedit

In production, 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. For example, you can use Metricbeat to ship monitoring data about Kibana, Elasticsearch, Logstash, and Beats to the monitoring cluster. If you are sending your data to the ESMS, see Elastic Stack Monitoring Service.

Metricbeat is the recommended method for collecting and shipping monitoring data to a monitoring cluster.

If you have previously configured legacy collection methods, you should migrate to using Metricbeat collection. Use either Metricbeat collection or legacy collection methods; do not use both.

Learn more about Collecting monitoring data with Metricbeat.

If you have at least a Gold Subscription, using a dedicated monitoring cluster also enables you to monitor multiple clusters from a central location.

To store monitoring data in a separate cluster:

  1. Set up the Elasticsearch cluster you want to use as the monitoring cluster. For example, you might set up a two host cluster with the nodes es-mon-1 and es-mon-2.

    • To monitor an Elasticsearch 7.x cluster, you must run Elasticsearch 7.x on the monitoring cluster.
    • There must be at least one ingest node in the monitoring cluster; it does not need to be a dedicated ingest node.

    1. (Optional) Verify that the collection of monitoring data is disabled on the monitoring cluster. By default, the xpack.monitoring.collection.enabled setting is false.

      For example, you can use the following APIs to review and change this setting:

      GET _cluster/settings

PUT _cluster/settings
{
  "persistent": {
    "xpack.monitoring.collection.enabled": false
  }
}

    2. If the Elasticsearch security features are enabled on the monitoring cluster, create users that can send and retrieve monitoring data.

      If you plan to use Kibana to view monitoring data, username and password credentials must be valid on both the Kibana server and the monitoring cluster.

      • If you plan to use Metricbeat to collect data about Elasticsearch or Kibana, create a user that has the remote_monitoring_collector built-in role and a user that has the remote_monitoring_agent built-in role. Alternatively, use the remote_monitoring_user built-in user.

      • If you plan to use HTTP exporters to route data through your production cluster, create a user that has the remote_monitoring_agent built-in role.

        For example, the following request creates a remote_monitor user that has the remote_monitoring_agent role:

        POST /_security/user/remote_monitor
{
  "password" : "changeme",
  "roles" : [ "remote_monitoring_agent"],
  "full_name" : "Internal Agent For Remote Monitoring"
}

        Alternatively, use the remote_monitoring_user built-in user.

  2. Configure your production cluster to collect data and send it to the monitoring cluster:

  3. (Optional) Configure Logstash to collect data and send it to the monitoring cluster.

  4. (Optional) Configure the Beats to collect data and send it to the monitoring cluster.

  5. (Optional) Configure Kibana to collect data and send it to the monitoring cluster:

  6. (Optional) Create a dedicated Kibana instance for monitoring, rather than using a single Kibana instance to access both your production cluster and monitoring cluster.

    1. (Optional) Disable the collection of monitoring data in this Kibana instance. Set the xpack.monitoring.kibana.collection.enabled setting to false in the kibana.yml file. For more information about this setting, see Monitoring settings in Kibana.
  7. Configure Kibana to retrieve and display the monitoring data.
« How monitoring works Collecting Elasticsearch monitoring data with Metricbeat »