Skip to main content
Version: 3.0.0

Events Ingester Telemetry

This document describes the OpenTelemetry metrics and assets for the Events Ingester.

For general setup instructions, see the OpenTelemetry Configuration guide.

Telemetry Assets​

Metrics Reference​

Naming note​

In Prometheus, metric names are normalized with underscores (e.g., events_ingester_events_received), and counters typically include the _total suffix.

Metrics Table​

MetricTypeUnitDescriptionPromQL example
events_ingester.events.receivedCountercountNumber of events accepted by router input path.sum(rate(events_ingester_events_received_total[5m]))
events_ingester.events.dispatchedCountercountNumber of events forwarded to the ingester handler.sum(rate(events_ingester_events_dispatched_total[5m]))
events_ingester.events.droppedCountercountDropped events in router/ingester paths.sum by (reason) (rate(events_ingester_events_dropped_total[5m]))
events_ingester.composition.lookup.duration_secondsHistogramsecondsTime spent resolving composition ID for an involved object.histogram_quantile(0.95, sum by (le) (rate(events_ingester_composition_lookup_duration_seconds_bucket[5m])))
events_ingester.records.build.failureCountercountFailures while building DB records from events.sum by (reason) (increase(events_ingester_records_build_failure_total[1h]))
events_ingester.batch.flush.duration_secondsHistogramsecondsDuration of a batch flush cycle.histogram_quantile(0.95, sum by (le) (rate(events_ingester_batch_flush_duration_seconds_bucket[5m])))
events_ingester.batch.flush.sizeHistogramrecordsNumber of records per flush.histogram_quantile(0.95, sum by (le) (rate(events_ingester_batch_flush_size_bucket[5m])))
events_ingester.db.insert.rowsCounterrowsNumber of rows inserted by successful batch writes.sum(rate(events_ingester_db_insert_rows_total[5m]))
events_ingester.db.insert.failureCountercountBatch insert failures.sum by (type) (increase(events_ingester_db_insert_failure_total[1h]))
events_ingester.queue.depthGaugecountIn-memory queue buffered job count.max(events_ingester_queue_depth)
events_ingester.record_channel.depthGaugecountIn-memory record channel buffered item count.max(events_ingester_record_channel_depth)