English | 简体中文 | 繁體中文 | Русский язык | Français | Español | Português | Deutsch | 日本語 | 한국어 | Italiano | بالعربية

Elasticsearch Monitoring

To monitor the health of the cluster, the monitoring feature collects metrics from each node and stores them in the Elasticsearch index. All settings related to monitoring in Elasticsearch must be in the elasticsearch.yml file for each node, or, if possible, in the dynamic cluster settings.

To start monitoring, we need to check the cluster settings, which can be done in the following ways:

GET _cluster/settings
{
   "persistent" : { },
   "transient" : { }
}

Each component in the stack is responsible for monitoring itself and then forwarding these documents to the Elasticsearch production cluster for routing and indexing (storage). The routing and indexing process in Elasticsearch is handled by the so-called collectors and exporters.

Collector

The collector runs once per collection interval to obtain data from the public API of the Elasticsearch it is monitoring. After data collection is completed, the data is batched and sent to the monitoring cluster for export.

Each data type collects only one collector. Each collector can create zero or more monitored documents.

Exporter

Exporters route data collected from any Elasticsearch Stack source to the monitoring cluster. Multiple exporters can be configured, but the conventional and default settings use a single exporter. Exporters can be configured at the node level and cluster level.

There are two types of exporters in Elasticsearch-

  • local −This exporter routes data back to the same cluster

  • http −The preferred exporter, which can route data to any supported Elasticsearch cluster accessible via HTTP.

Before the export program can route monitoring data, they must be configured with certain Elasticsearch resources. These resources include templates and ingestion pipelines