Elasticsearch¶

The elasticsearch probe monitors an Elasticsearch cluster via the REST JSON API, collecting cluster health, JVM memory, indexing and search throughput, and thread pool queue depths.

Quick start¶

# probes.d/10-elasticsearch.yaml — each file under probes.d/ is a YAML array of probes
- name: elasticsearch
  type: elasticsearch
  params:
    endpoint: http://localhost:9200

Parameters¶

Parameter	Must set	Default	Description
`endpoint`	In practice	`http://localhost:9200`	Base URL of the node; https:// for a secured cluster. Example: `https://es01:9200`
`username`	In practice	-	Basic-auth user; empty when security is disabled
`password`	In practice	-	Basic-auth password. A secret: reference it with `${secret:…}`, `${env:…}` or `${file:…}` rather than writing it in the file
`timeout`	No	`10`	HTTP request timeout in seconds
`interval`	No	`60`	Seconds between collections
`instance_name`	No	-	Stable identity override for this node

Metrics¶

Metric	Unit	Description
`senhub.elasticsearch.up`	1	1 when the cluster health endpoint is reachable
`elasticsearch.cluster.health`	1	Cluster health: 2 = green, 1 = yellow, 0 = red
`elasticsearch.cluster.nodes`	{node}	Total nodes in the cluster
`elasticsearch.cluster.data_nodes`	{node}	Data nodes in the cluster
`elasticsearch.cluster.shards.active`	{shard}	Active primary and replica shards
`elasticsearch.cluster.shards.unassigned`	{shard}	Unassigned shards
`elasticsearch.jvm.memory.heap.used`	By	JVM heap in use on the local node
`elasticsearch.indexing.operations.completed`	#	Indexing operations completed, tagged with `operation` (index, delete)
`elasticsearch.search.operations.completed`	#	Search operations completed, tagged with `operation` (query, fetch)
`elasticsearch.thread_pool.tasks.queued`	{task}	Tasks queued per thread pool, tagged with `thread_pool`

Operational notes¶

The probe queries /_cluster/health (cluster-level) and /_nodes/_local/stats (node-level).
For TLS-secured clusters, prefix the endpoint with https:// and ensure the certificate is valid or configure system trust.
Metrics align with the OpenTelemetry Collector contrib elasticsearchreceiver where naming equivalents exist.

Metric reference¶

Every metric this probe can emit. Metric is the OpenTelemetry name the OTLP, Prometheus and Zabbix outputs derive theirs from. Name is what a Nagios check and the API metrics= filter match. PRTG channel is the label PRTG shows, placeholders filled from the series' tags.

Metric	Name	PRTG channel	Unit	Description
`senhub.elasticsearch.up`	`senhub.elasticsearch.up`	Elasticsearch Up	#	1 when the cluster health endpoint is reachable, 0 otherwise
`elasticsearch.cluster.health`	`elasticsearch.cluster.health`	Elasticsearch Cluster Health	#	Cluster health status encoded as integer: green=2, yellow=1, red=0
`elasticsearch.cluster.nodes`	`elasticsearch.cluster.nodes`	Elasticsearch Cluster Nodes	#	Total number of nodes in the cluster
`elasticsearch.cluster.data_nodes`	`elasticsearch.cluster.data_nodes`	Elasticsearch Cluster Data Nodes	#	Number of data nodes in the cluster
`elasticsearch.cluster.shards.active`	`elasticsearch.cluster.shards.active`	Elasticsearch Active Shards	#	Number of active primary and replica shards
`elasticsearch.cluster.shards.unassigned`	`elasticsearch.cluster.shards.unassigned`	Elasticsearch Unassigned Shards	#	Number of shards not assigned to any node
`elasticsearch.cluster.shards.relocating`	`elasticsearch.cluster.shards.relocating`	Elasticsearch Relocating Shards	#	Number of shards being relocated between nodes
`elasticsearch.cluster.pending_tasks`	`elasticsearch.cluster.pending_tasks`	Elasticsearch Pending Tasks	#	Number of cluster-level changes not yet executed
`elasticsearch.jvm.memory.heap.used`	`elasticsearch.jvm.memory.heap.used`	Elasticsearch JVM Heap Used	B	JVM heap memory currently used by the node
`elasticsearch.jvm.memory.heap.max`	`elasticsearch.jvm.memory.heap.max`	Elasticsearch JVM Heap Max	B	Maximum JVM heap memory available to the node
`elasticsearch.jvm.gc.collections.count`	`elasticsearch.jvm.gc.collections.count`	Elasticsearch GC {collector} Collections	#	Number of GC collections for the given collector (young\|old)
`elasticsearch.jvm.gc.collections.elapsed`	`elasticsearch.jvm.gc.collections.elapsed`	Elasticsearch GC {collector} Time	ms	Cumulative wall-clock time spent in GC for the given collector
`elasticsearch.indexing.operations.completed`	`elasticsearch.indexing.operations.completed`	Elasticsearch Indexing {operation} Completed	#	Cumulative number of indexing operations completed (operation=index)
`elasticsearch.indexing.operations.time`	`elasticsearch.indexing.operations.time`	Elasticsearch Indexing {operation} Time	ms	Cumulative time spent in indexing operations
`elasticsearch.search.operations.completed`	`elasticsearch.search.operations.completed`	Elasticsearch Search {operation} Completed	#	Cumulative search operations (operation=query\|fetch)
`elasticsearch.search.operations.time`	`elasticsearch.search.operations.time`	Elasticsearch Search {operation} Time	ms	Cumulative time spent in search operations
`elasticsearch.process.cpu.usage`	`elasticsearch.process.cpu.usage`	Elasticsearch Process CPU Usage	%	CPU utilisation of the Elasticsearch process as a percentage (0-100)
`elasticsearch.os.memory.used`	`elasticsearch.os.memory.used`	Elasticsearch OS Memory Used	B	Physical memory used by the OS on the node
`elasticsearch.thread_pool.tasks.queued`	`elasticsearch.thread_pool.tasks.queued`	Elasticsearch Thread Pool {thread_pool} Queued	#	Number of tasks currently queued in the thread pool
`elasticsearch.thread_pool.tasks.completed`	`elasticsearch.thread_pool.tasks.completed`	Elasticsearch Thread Pool {thread_pool} Completed	#	Cumulative tasks completed by the thread pool
`elasticsearch.thread_pool.tasks.rejected`	`elasticsearch.thread_pool.tasks.rejected`	Elasticsearch Thread Pool {thread_pool} Rejected	#	Cumulative tasks rejected by the thread pool (queue full)