After you configure the Elasticsearch host, PATROL for Elasticsearch enables you to automatically collect and monitor the Cluster, Node, and Index performance metrics. The configuration parameters of the Cluster provide information about the Elasticsearch Version, Lucene version, DataNodes, Master Nodes, and other details. For information about how to configure an Elasticsearch host for monitoring see:
PATROL for Elasticsearch monitors the performance metrics of the Elasticsearch Cluster, Nodes, JVM, Indices, and Shards. It provides ample metrics for each of these entities to fulfill all your monitoring requirements.
Elasticsearch KM monitors your cluster health each minute. The Cluster Status attribute indicates the overall health of your cluster.
You can set alerts on the Cluster Status attribute so that you get notified when the cluster health is at risk.
Key Performance Indicator (KPI) metrics
The KM provides Number of documents deleted, Field data cache size, Field data cache evictions, and other attributes apart from the KPI metrics.
For the complete list of attributes available for Clusters, see the Cluster monitor type.
Cluster performance report
In Elasticsearch, a Node is a single instance of Elasticsearch. It usually runs one instance per machine.
Node monitoring can be used to:
Key Performance Indicator (KPI) metrics
Elasticsearch KM monitors the Average response time of a query, Field data cache evictions, Query cache count, and other attributes apart from the KPI metrics.
For the complete list of attributes available for Nodes, see the Elasticsearch Node monitor type.
Node performance report
In Elasticsearch, an index is a collection of documents that have similar characteristics.
Index monitoring helps you to:
Determine if the average response time of the queries remains high. Queries can be redefined and metrics can be compared to measure the efficiency of queries.
Determine the rate at which indices are refreshed with new information
Determine the number of documents indexed per millisecond
Key Performance Indicator (KPI) metrics
Elasticseach KM monitors the availability of each Index, Average response time of a query, Field data cache evictions, Query cache evictions, and other attributes apart from the KPI metrics.
For the complete list of attributes available for Index, see the Elasticsearch Index monitor type.
Index performance report
Elasticsearch KM provides Heap Utilization attribute that displays the JVM heap utilized by the node. This helps you to ensure that your system does not run out of memory. The KM also provides Heap size used (in GB) and Maximum heap size configured (in GB) for the Node.
Key Performance Indicator (KPI) metrics for JVM
For the complete list of attributes available for JVM, see the JVM monitor type.
JVM performance report
Shards are individual instances of a Lucene index. Shard allocation helps us understand if the Elasticsearch application is using optimum number of Shards.
The KM reports number of active, initializing, relocating, and unassigned shards.
Key Performance Indicator (KPI) metrics
For the complete list of attributes available for Shards, see the Elasticsearch Shards monitor type.
Shards performance report
Elasticsearch KM facilitates filtering of indices by including or excluding them (based on name) from monitoring.
For information about configuring filters, see Configuring the Elasticsearch monitor profile.
Elasticsearch KM automatically selects the next running node to collect data, if configured node fails. This assures high availability of performance metrics.