Technologies/Elasticsearch/elasticsearch.node.disk.io.read
ElasticsearchElasticsearchMetric

elasticsearch.node.disk.io.read

Disk read operations
Dimensions:None
Available on:DynatraceDynatrace (2)PrometheusPrometheus (1)DatadogDatadog (1)OpenTelemetryOpenTelemetry (1)

Summary

Measures disk read I/O operations performed by the node, tracking the volume of data read from storage. High read rates during search indicate that queries are accessing data not cached in memory, suggesting inadequate filesystem cache (see "Filesystem Cache Starvation" insight). Sudden read spikes may indicate inefficient queries forcing disk access or cache warming operations after node restarts.

Interface Metrics (5)
DynatraceDynatrace
Amount of read operations per second
Dimensions:None
PrometheusPrometheus
Count of disk read operations
Dimensions:None
DatadogDatadog
The total number of reads from the file store [v1.0+].
Dimensions:None
OpenTelemetryOpenTelemetry
The total number of kilobytes read across all file stores for this node.
Dimensions:None
DynatraceDynatrace
Total kilobytes read from disk
Dimensions:None
Sources
elasticsearch_filesystem_io_stats_device_read_operations_countwww.dynatrace.com
filesystem_io_stats_device_read_operations_countgithub.com
elasticsearch.fs.total.disk_readsgithub.com
elasticsearch.node.disk.io.readgithub.com

Technical Annotations (6)

Configuration Parameters (1)
on_disk_rescorerecommended: true
For dense_vector fields when vector data exceeds available RAM
Technical References (5)
TSDBconceptsynthetic _idcomponentdoc values skippercomponentkNN rescoringconceptdense_vectorcomponent
Related Insights (3)
Document GET Operation Latencyinfo

High latency for document GET operations indicates potential issues with routing, shard location, or storage performance when retrieving specific documents by ID.

▸
TSDB synthetic _id format changes require re-indexing for performance benefitswarning
▸
On-disk vector rescoring avoids memory pressure in low-RAM environmentsinfo
▸