Apache Solr¶

The solr probe monitors Apache Solr via the native metrics API (/solr/admin/metrics?wt=json&group=all) and the core status endpoint (/solr/admin/cores?action=STATUS), reporting JVM heap and threads, node-level request and cache counters, and per-core document count and index size.

Quick start¶

# probes.d/10-solr.yaml — each file under probes.d/ is a YAML array of probes
- name: solr
  type: solr
  params:
    endpoint: http://localhost:8983

Parameters¶

Parameter	Must set	Default	Description
`endpoint`	In practice	`http://localhost:8983`	Base URL of the node, without the /solr path. Example: `http://solr01:8983`
`jolokia_url`	No	-	Alternative spelling of endpoint kept for consistency with the JVM probes; its scheme, host and port replace endpoint, the path is dropped
`timeout`	No	`10`	HTTP request timeout in seconds
`interval`	No	`60`	Seconds between collections
`instance_name`	No	-	Stable identity override for this node

jolokia_url does not switch the probe to Jolokia: the native metrics API is always what it reads. It is accepted so a configuration written for the JVM probes keeps working, and only its scheme, host and port are used.

Metrics¶

Metric	Unit	Description
`senhub.solr.up`	1	1 when the Solr admin metrics endpoint responded
`jvm.memory.heap.used`	By	JVM heap memory used by the Solr process
`jvm.memory.heap.max`	By	JVM maximum heap size
`jvm.threads.count`	{thread}	Current live JVM thread count
`solr.requests.count`	{request}	Requests processed by the node
`solr.errors.count`	{error}	Request errors on the node
`solr.requests.time`	s	Time spent handling QUERY requests (cumulative)
`solr.cache.inserts`	#	Query result cache inserts (cumulative)
`solr.cache.hits`	{hit}	Cache hits
`solr.document.count`	{document}	Number of indexed documents per core, tagged with `core`
`solr.index.size`	By	Index size on disk per core

Operational notes¶

No authentication is required by default. If Solr is configured with Basic Auth, the probe does not yet support credentials — use the unauthenticated path or a local loopback address.
For SolrCloud, point the probe at one node; cluster-wide aggregates are not covered in this release.

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.solr.up`	`senhub.solr.up`	Solr Up	#	1 when the Solr admin metrics endpoint responded successfully, 0 otherwise
`jvm.memory.heap.used`	`jvm.memory.heap.used`	JVM Heap Used	B	JVM heap memory currently used by the Solr process
`jvm.threads.count`	`jvm.threads.count`	JVM Thread Count	#	Number of live threads in the Solr JVM
`solr.requests.count`	`solr.requests.count`	Solr Requests	#	Cumulative number of requests handled by QUERY handlers
`solr.requests.time`	`solr.requests.time`	Solr Request Time	ms	Cumulative time spent handling QUERY requests (meanMs * count, ms)
`solr.errors.count`	`solr.errors.count`	Solr Errors	#	Cumulative number of errors across all QUERY handlers
`solr.cache.hits`	`solr.cache.hits`	Solr Cache Hits	#	Cumulative query result cache hits
`solr.cache.inserts`	`solr.cache.inserts`	Solr Cache Inserts	#	Cumulative query result cache inserts
`solr.document.count`	`solr.document.count`	Solr {core} Document Count	#	Number of indexed documents in the core
`solr.index.size`	`solr.index.size`	Solr {core} Index Size	B	On-disk index size in bytes for the core